diff options
101 files changed, 4817 insertions, 2449 deletions
diff --git a/Makefile.am b/Makefile.am index a0fba116..1c30a6ff 100644 --- a/Makefile.am +++ b/Makefile.am @@ -22,5 +22,5 @@ SUBDIRS = \ EXTRA_DIST = corpus tests python/pkg python/src python/tests python/examples compound-split environment AUTOMAKE_OPTIONS = foreign ACLOCAL_AMFLAGS = -I m4 -AM_CPPFLAGS = -D_GLIBCXX_PARALLEL +AM_CPPFLAGS = -D_GLIBCXX_PARALLEL -march=native -mtune=native -O2 -pipe -fomit-frame-pointer -Wall @@ -1,24 +1,35 @@ `cdec` is a research platform for machine translation and similar structured prediction problems. -## Installation +## Building from a downloaded archive -Build `cdec`: +Instructions: - autoreconf -ifv ./configure make ./tests/run-system-tests.pl -You will need the following libraries / tools: +You will need the following software: -- [Autoconf / Automake / Libtool](http://www.gnu.org/software/autoconf/) - - Older versions of GNU autotools may not work properly. - [Boost C++ libraries (version 1.44 or later)](http://www.boost.org/) - If you build your own boost, you _must install it_ using `bjam install`. - Older versions of Boost _may_ work, but problems have been reported with command line option parsing on some platforms with older versions. - [GNU Flex](http://flex.sourceforge.net/) +## Building from a git clone + +In addition to the standard `cdec` third party requirements, you will additionally need the following software: + +- [Autoconf / Automake / Libtool](http://www.gnu.org/software/autoconf/) + - Older versions of GNU autotools may not work properly. + +Instructions: + + autoreconf -ifv + ./configure + make + ./tests/run-system-tests.pl + ## Further information -[For more information, refer to the cdec documentation](http://www.cdec-decoder.org) +[For more information, refer to the `cdec` documentation](http://www.cdec-decoder.org) diff --git a/configure.ac b/configure.ac index 59224e0c..eb09676e 100644 --- a/configure.ac +++ b/configure.ac @@ -1,4 +1,4 @@ -AC_INIT([cdec],[2013-01-20]) +AC_INIT([cdec],[2013-03-08]) AC_CONFIG_SRCDIR([decoder/cdec.cc]) AM_INIT_AUTOMAKE AC_CONFIG_HEADERS(config.h) @@ -214,6 +214,7 @@ AC_CONFIG_FILES([training/pro/Makefile]) AC_CONFIG_FILES([training/rampion/Makefile]) AC_CONFIG_FILES([training/minrisk/Makefile]) AC_CONFIG_FILES([training/mira/Makefile]) +AC_CONFIG_FILES([training/latent_svm/Makefile]) AC_CONFIG_FILES([training/dtrain/Makefile]) # external feature function example code diff --git a/corpus/lowercase.pl b/corpus/lowercase.pl index 688e493b..9fd91dac 100755 --- a/corpus/lowercase.pl +++ b/corpus/lowercase.pl @@ -2,7 +2,7 @@ use strict; binmode(STDIN,":utf8"); binmode(STDOUT,":utf8"); -while(<>) { +while(<STDIN>) { $_ = lc $_; print; } diff --git a/corpus/paste-files.pl b/corpus/paste-files.pl index 4cb424ad..ef2cd937 100755 --- a/corpus/paste-files.pl +++ b/corpus/paste-files.pl @@ -34,6 +34,7 @@ while(1) { $done = 1; last; } + $r =~ s/\r//g; chomp $r; if ($r =~ /\|\|\|/) { $r = ''; diff --git a/corpus/support/quote-norm.pl b/corpus/support/quote-norm.pl index e4e5055e..b104e73c 100755 --- a/corpus/support/quote-norm.pl +++ b/corpus/support/quote-norm.pl @@ -11,6 +11,8 @@ while(<STDIN>) { s/&\s*squot\s*;/'/gi; s/&\s*quot\s*;/"/gi; s/&\s*amp\s*;/&/gi; + s/&\s*nbsp\s*;/&/gi; + s/&\s*#\s*160\s*;/ /gi; s/ (\d\d): (\d\d)/ $1:$2/g; s/[\x{20a0}]\x{20ac}]/ EUR /g; s/[\x{00A3}]/ GBP /g; @@ -20,6 +22,7 @@ while(<STDIN>) { s/&\s*#45\s*;/--/g; s/�c/--/g; s/ ,,/ "/g; + s/„/"/g; s/``/"/g; s/''/"/g; s/[「」]/"/g; diff --git a/corpus/support/token_list b/corpus/support/token_list index d470cb22..43dd80d9 100644 --- a/corpus/support/token_list +++ b/corpus/support/token_list @@ -37,6 +37,8 @@ tel. 10. ##################### abbreviation: words that contain period. +EE.UU. +ee.uu. U.A.E Ala. Ph.D. @@ -292,3 +294,155 @@ Z. т.н. т.ч. н.э. +# Swahili +A.D. +Afr. +A.G. +agh. +A.H. +A.M. +a.s. +B.A. +B.C. +Bi. +B.J. +B.K. +B.O.M. +Brig. +Bro. +bt. +bw. +Bw. +Cap. +C.C. +cCM. +C.I.A. +cit. +C.M.S. +Co. +Corp. +C.S.Sp. +C.W. +D.C. +Dk. +Dkt. +Dk.B. +Dr. +E.C. +e.g. +E.M. +E.n. +etc. +Feb. +F.F.U. +F.M. +Fr. +F.W. +I.C.O. +i.e. +I.L.C. +Inc. +Jan. +J.F. +Jr. +J.S. +J.V.W.A. +K.A.R. +K.A.U. +K.C.M.C. +K.k. +K.K. +k.m. +km. +K.m. +K.N.C.U. +K.O. +K.S. +Ksh. +kt. +kumb. +k.v. +kv. +L.G. +ltd. +Ltd. +M.A. +M.D. +mf. +Mh. +Mhe. +mil. +m.m. +M.m. +Mm. +M.M. +Mr. +Mrs. +M.S. +Mt. +Mw. +M.W. +Mwl. +na. +Na. +N.F. +N.J. +n.k. +nk. +n.k.w. +N.N. +Nov. +O.C.D. +op. +P.C. +Phd. +Ph.D. +P.J. +P.o. +P.O. +P.O.P. +P.P.F. +Prof. +P.s. +P.S. +Q.C. +Rd. +s.a.w. +S.A.W. +S.D. +Sept. +sh. +Sh. +SH. +shs. +Shs. +S.J. +S.L. +S.L.P. +S.s. +S.S. +St. +s.w. +s.w.T. +taz. +Taz. +T.C. +T.E.C. +T.L.P. +T.O.H.S. +Tsh. +T.V. +tz. +uk. +Uk. +U.M.C.A. +U.N. +U.S. +Ush. +U.W.T. +Viii. +Vol. +V.T.C. +W.H. +yamb. +Y.M.C.A. diff --git a/corpus/support/tokenizer.pl b/corpus/support/tokenizer.pl index 0350a894..acc537fb 100755 --- a/corpus/support/tokenizer.pl +++ b/corpus/support/tokenizer.pl @@ -226,7 +226,7 @@ sub proc_token { } ## step 1: check the most common case - if($token =~ /^[a-z0-9\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}]+$/i){ + if($token =~ /^[a-z0-9\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}\p{Devanagari}]+$/i){ ### most common cases return $token; } @@ -246,7 +246,7 @@ sub proc_token { ## number return $token; } - if($token =~ /^(@|#)[A-Za-z0-9_\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}]+.*$/){ + if($token =~ /^(@|#)[A-Za-z0-9_\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}\p{Devanagari}]+.*$/){ ## twitter hashtag or address return proc_rightpunc($token); } @@ -277,7 +277,7 @@ sub proc_token { } #my $t1 = '[\x{0600}-\x{06ff}a-z\d\_\.\-]'; - my $t1 = '[a-z\d\_\-\.\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}]'; + my $t1 = '[a-z\d\_\-\.\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}\p{Devanagari}]'; if($token =~ /^\/(($t1)+\/)+($t1)+\/?$/i){ ### /nls/p/.... return $token; @@ -361,7 +361,7 @@ sub deep_proc_token { } ##### step 0: if it mades up of all puncts, remove one punct at a time. - if($line !~ /[\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}a-zA-Z\d]/){ + if($line !~ /[\p{Cyrillic}\p{Greek}\p{Hebrew}\p{Han}\p{Arabic}\p{Devanagari}a-zA-Z\d]/){ if($line =~ /^(\!+|\@+|\++|\=+|\*+|\<+|\>+|\|+|\?+|\.+|\-+|\_+|\&+)$/){ ## ++ @@@@ !!! .... return $line; diff --git a/corpus/tokenize-anything.sh b/corpus/tokenize-anything.sh index 1a24193d..028992cf 100755 --- a/corpus/tokenize-anything.sh +++ b/corpus/tokenize-anything.sh @@ -9,5 +9,5 @@ $SUPPORT/utf8-normalize.sh | sed -e 's/ al - / al-/g' | $SUPPORT/fix-contract.pl | sed -e 's/^ //' | sed -e 's/ $//' | - perl -e 'while(<>){s/(\d+)(\.+)$/$1 ./;print;}' + perl -e 'while(<>){s/(\d+)(\.+)$/$1 ./; s/(\d+)(\.+) \|\|\|/$1 . |||/; print;}' diff --git a/decoder/Makefile.am b/decoder/Makefile.am index 6499b38b..82b50f19 100644 --- a/decoder/Makefile.am +++ b/decoder/Makefile.am @@ -60,6 +60,7 @@ libcdec_a_SOURCES = \ ff_rules.h \ ff_ruleshape.h \ ff_sample_fsa.h \ + ff_source_path.h \ ff_source_syntax.h \ ff_spans.h \ ff_tagger.h \ @@ -140,6 +141,7 @@ libcdec_a_SOURCES = \ ff_wordalign.cc \ ff_csplit.cc \ ff_tagger.cc \ + ff_source_path.cc \ ff_source_syntax.cc \ ff_bleu.cc \ ff_factory.cc \ diff --git a/decoder/cdec_ff.cc b/decoder/cdec_ff.cc index 3ab0f9f6..0bf441d4 100644 --- a/decoder/cdec_ff.cc +++ b/decoder/cdec_ff.cc @@ -14,6 +14,7 @@ #include "ff_rules.h" #include "ff_ruleshape.h" #include "ff_bleu.h" +#include "ff_source_path.h" #include "ff_source_syntax.h" #include "ff_register.h" #include "ff_charset.h" @@ -70,6 +71,7 @@ void register_feature_functions() { ff_registry.Register("InputIndicator", new FFFactory<InputIndicator>); ff_registry.Register("LexicalTranslationTrigger", new FFFactory<LexicalTranslationTrigger>); ff_registry.Register("WordPairFeatures", new FFFactory<WordPairFeatures>); + ff_registry.Register("SourcePathFeatures", new FFFactory<SourcePathFeatures>); ff_registry.Register("WordSet", new FFFactory<WordSet>); ff_registry.Register("Dwarf", new FFFactory<Dwarf>); ff_registry.Register("External", new FFFactory<ExternalFeature>); diff --git a/decoder/ff_klm.cc b/decoder/ff_klm.cc index fefa90bd..c8ca917a 100644 --- a/decoder/ff_klm.cc +++ b/decoder/ff_klm.cc @@ -1,6 +1,7 @@ #include "ff_klm.h" #include <cstring> +#include <cstdlib> #include <iostream> #include <boost/scoped_ptr.hpp> @@ -151,8 +152,9 @@ template <class Model> class BoundaryRuleScore { template <class Model> class KLanguageModelImpl { public: - double LookupWords(const TRule& rule, const vector<const void*>& ant_states, double* oovs, void* remnant) { + double LookupWords(const TRule& rule, const vector<const void*>& ant_states, double* oovs, double* emit, void* remnant) { *oovs = 0; + *emit = 0; const vector<WordID>& e = rule.e(); BoundaryRuleScore<Model> ruleScore(*ngram_, *static_cast<BoundaryAnnotatedState*>(remnant)); unsigned i = 0; @@ -169,8 +171,9 @@ class KLanguageModelImpl { if (e[i] <= 0) { ruleScore.NonTerminal(*static_cast<const BoundaryAnnotatedState*>(ant_states[-e[i]])); } else { - const WordID cdec_word_or_class = ClassifyWordIfNecessary(e[i]); // in future, - // maybe handle emission + float ep = 0.f; + const WordID cdec_word_or_class = ClassifyWordIfNecessary(e[i], &ep); + if (ep) { *emit += ep; } const lm::WordIndex cur_word = MapWord(cdec_word_or_class); // map to LM's id if (cur_word == 0) (*oovs) += 1.0; ruleScore.Terminal(cur_word); @@ -205,12 +208,14 @@ class KLanguageModelImpl { // if this is not a class-based LM, returns w untransformed, // otherwise returns a word class mapping of w, // returns TD::Convert("<unk>") if there is no mapping for w - WordID ClassifyWordIfNecessary(WordID w) const { + WordID ClassifyWordIfNecessary(WordID w, float* emitp) const { if (word2class_map_.empty()) return w; if (w >= word2class_map_.size()) return kCDEC_UNK; - else - return word2class_map_[w]; + else { + *emitp = word2class_map_[w].second; + return word2class_map_[w].first; + } } // converts to cdec word id's to KenLM's id space, OOVs and <unk> end up at 0 @@ -256,32 +261,32 @@ class KLanguageModelImpl { int lc = 0; if (!SILENT) cerr << " Loading word classes from " << file << " ...\n"; - AddWordToClassMapping_(TD::Convert("<s>"), TD::Convert("<s>")); - AddWordToClassMapping_(TD::Convert("</s>"), TD::Convert("</s>")); - while(in) { - getline(in, line); - if (!in) continue; + AddWordToClassMapping_(TD::Convert("<s>"), TD::Convert("<s>"), 0.0); + AddWordToClassMapping_(TD::Convert("</s>"), TD::Convert("</s>"), 0.0); + while(getline(in, line)) { dummy.clear(); TD::ConvertSentence(line, &dummy); ++lc; - if (dummy.size() != 2) { + if (dummy.size() != 3) { + cerr << " Class map file expects: CLASS WORD logp(WORD|CLASS)\n"; cerr << " Format error in " << file << ", line " << lc << ": " << line << endl; abort(); } - AddWordToClassMapping_(dummy[0], dummy[1]); + AddWordToClassMapping_(dummy[1], dummy[0], strtof(TD::Convert(dummy[2]).c_str(), NULL)); } } - void AddWordToClassMapping_(WordID word, WordID cls) { + void AddWordToClassMapping_(WordID word, WordID cls, float emit) { if (word2class_map_.size() <= word) { - word2class_map_.resize((word + 10) * 1.1, kCDEC_UNK); + word2class_map_.resize((word + 10) * 1.1, pair<WordID,float>(kCDEC_UNK,0.f)); assert(word2class_map_.size() > word); } - if(word2class_map_[word] != kCDEC_UNK) { + if(word2class_map_[word].first != kCDEC_UNK) { cerr << "Multiple classes for symbol " << TD::Convert(word) << endl; abort(); } - word2class_map_[word] = cls; + word2class_map_[word].first = cls; + word2class_map_[word].second = emit; } ~KLanguageModelImpl() { @@ -304,7 +309,9 @@ class KLanguageModelImpl { int order_; vector<lm::WordIndex> cdec2klm_map_; - vector<WordID> word2class_map_; // if this is a class-based LM, this is the word->class mapping + vector<pair<WordID,float> > word2class_map_; // if this is a class-based LM, + // .first is the word->class mapping + // .second is the emission log probability }; template <class Model> @@ -322,6 +329,7 @@ KLanguageModel<Model>::KLanguageModel(const string& param) { } fid_ = FD::Convert(featname); oov_fid_ = FD::Convert(featname+"_OOV"); + emit_fid_ = FD::Convert(featname+"_Emit"); // cerr << "FID: " << oov_fid_ << endl; SetStateSize(pimpl_->ReserveStateSize()); } @@ -340,9 +348,12 @@ void KLanguageModel<Model>::TraversalFeaturesImpl(const SentenceMetadata& /* sme void* state) const { double est = 0; double oovs = 0; - features->set_value(fid_, pimpl_->LookupWords(*edge.rule_, ant_states, &oovs, state)); + double emit = 0; + features->set_value(fid_, pimpl_->LookupWords(*edge.rule_, ant_states, &oovs, &emit, state)); if (oovs && oov_fid_) features->set_value(oov_fid_, oovs); + if (emit && emit_fid_) + features->set_value(emit_fid_, emit); } template <class Model> diff --git a/decoder/ff_klm.h b/decoder/ff_klm.h index b5ceffd0..db4032f7 100644 --- a/decoder/ff_klm.h +++ b/decoder/ff_klm.h @@ -28,8 +28,9 @@ class KLanguageModel : public FeatureFunction { SparseVector<double>* estimated_features, void* out_context) const; private: - int fid_; // conceptually const; mutable only to simplify constructor - int oov_fid_; // will be zero if extra OOV feature is not configured by decoder + int fid_; // LanguageModel + int oov_fid_; // LanguageModel_OOV + int emit_fid_; // LanguageModel_Emit [only used for class-based LMs] KLanguageModelImpl<Model>* pimpl_; }; diff --git a/decoder/ff_ngrams.cc b/decoder/ff_ngrams.cc index 9c13fdbb..d337b28b 100644 --- a/decoder/ff_ngrams.cc +++ b/decoder/ff_ngrams.cc @@ -60,7 +60,7 @@ namespace { } } -static bool ParseArgs(string const& in, bool* explicit_markers, unsigned* order, vector<string>& prefixes, string& target_separator) { +static bool ParseArgs(string const& in, bool* explicit_markers, unsigned* order, vector<string>& prefixes, string& target_separator, string* cluster_file) { vector<string> const& argv=SplitOnWhitespace(in); *explicit_markers = false; *order = 3; @@ -103,6 +103,10 @@ static bool ParseArgs(string const& in, bool* explicit_markers, unsigned* order, LMSPEC_NEXTARG; prefixes[5] = *i; break; + case 'c': + LMSPEC_NEXTARG; + *cluster_file = *i; + break; case 'S': LMSPEC_NEXTARG; target_separator = *i; @@ -124,6 +128,7 @@ usage: << "NgramFeatures Usage: \n" << " feature_function=NgramFeatures filename.lm [-x] [-o <order>] \n" + << " [-c <cluster-file>]\n" << " [-U <unigram-prefix>] [-B <bigram-prefix>][-T <trigram-prefix>]\n" << " [-4 <4-gram-prefix>] [-5 <5-gram-prefix>] [-S <separator>]\n\n" @@ -203,6 +208,12 @@ class NgramDetectorImpl { SetFlag(flag, HAS_FULL_CONTEXT, state); } + WordID MapToClusterIfNecessary(WordID w) const { + if (cluster_map.size() == 0) return w; + if (w >= cluster_map.size()) return kCDEC_UNK; + return cluster_map[w]; + } + void FireFeatures(const State<5>& state, WordID cur, SparseVector<double>* feats) { FidTree* ft = &fidroot_; int n = 0; @@ -285,7 +296,7 @@ class NgramDetectorImpl { context_complete = true; } } else { // handle terminal - const WordID cur_word = e[j]; + const WordID cur_word = MapToClusterIfNecessary(e[j]); SparseVector<double> p; if (cur_word == kSOS_) { state = BeginSentenceState(); @@ -348,9 +359,52 @@ class NgramDetectorImpl { } } + void ReadClusterFile(const string& clusters) { + ReadFile rf(clusters); + istream& in = *rf.stream(); + string line; + int lc = 0; + string cluster; + string word; + while(getline(in, line)) { + ++lc; + if (line.size() == 0) continue; + if (line[0] == '#') continue; + unsigned cend = 1; + while((line[cend] != ' ' && line[cend] != '\t') && cend < line.size()) { + ++cend; + } + if (cend == line.size()) { + cerr << "Line " << lc << " in " << clusters << " malformed: " << line << endl; + abort(); + } + unsigned wbeg = cend + 1; + while((line[wbeg] == ' ' || line[wbeg] == '\t') && wbeg < line.size()) { + ++wbeg; + } + if (wbeg == line.size()) { + cerr << "Line " << lc << " in " << clusters << " malformed: " << line << endl; + abort(); + } + unsigned wend = wbeg + 1; + while((line[wend] != ' ' && line[wend] != '\t') && wend < line.size()) { + ++wend; + } + const WordID clusterid = TD::Convert(line.substr(0, cend)); + const WordID wordid = TD::Convert(line.substr(wbeg, wend - wbeg)); + if (wordid >= cluster_map.size()) + cluster_map.resize(wordid + 10, kCDEC_UNK); + cluster_map[wordid] = clusterid; + } + cluster_map[kSOS_] = kSOS_; + cluster_map[kEOS_] = kEOS_; + } + + vector<WordID> cluster_map; + public: explicit NgramDetectorImpl(bool explicit_markers, unsigned order, - vector<string>& prefixes, string& target_separator) : + vector<string>& prefixes, string& target_separator, const string& clusters) : kCDEC_UNK(TD::Convert("<unk>")) , add_sos_eos_(!explicit_markers) { order_ = order; @@ -369,6 +423,9 @@ class NgramDetectorImpl { dummy_rule_.reset(new TRule("[DUMMY] ||| [BOS] [DUMMY] ||| [1] [2] </s> ||| X=0")); kSOS_ = TD::Convert("<s>"); kEOS_ = TD::Convert("</s>"); + + if (clusters.size()) + ReadClusterFile(clusters); } ~NgramDetectorImpl() { @@ -409,9 +466,10 @@ NgramDetector::NgramDetector(const string& param) { vector<string> prefixes; bool explicit_markers = false; unsigned order = 3; - ParseArgs(param, &explicit_markers, &order, prefixes, target_separator); + string clusters; + ParseArgs(param, &explicit_markers, &order, prefixes, target_separator, &clusters); pimpl_ = new NgramDetectorImpl(explicit_markers, order, prefixes, - target_separator); + target_separator, clusters); SetStateSize(pimpl_->ReserveStateSize()); } diff --git a/decoder/ff_rules.cc b/decoder/ff_rules.cc index 6716d3da..410e083c 100644 --- a/decoder/ff_rules.cc +++ b/decoder/ff_rules.cc @@ -107,7 +107,12 @@ void RuleSourceBigramFeatures::TraversalFeaturesImpl(const SentenceMetadata& sme (*features) += it->second; } -RuleTargetBigramFeatures::RuleTargetBigramFeatures(const std::string& param) { +RuleTargetBigramFeatures::RuleTargetBigramFeatures(const std::string& param) : inds(1000) { + for (unsigned i = 0; i < inds.size(); ++i) { + ostringstream os; + os << (i + 1); + inds[i] = os.str(); + } } void RuleTargetBigramFeatures::PrepareForInput(const SentenceMetadata& smeta) { @@ -126,11 +131,18 @@ void RuleTargetBigramFeatures::TraversalFeaturesImpl(const SentenceMetadata& sme it = rule2_feats_.insert(make_pair(&rule, SparseVector<double>())).first; SparseVector<double>& f = it->second; string prev = "<r>"; + vector<WordID> nt_types(rule.Arity()); + unsigned ntc = 0; + for (int i = 0; i < rule.f_.size(); ++i) + if (rule.f_[i] < 0) nt_types[ntc++] = -rule.f_[i]; for (int i = 0; i < rule.e_.size(); ++i) { WordID w = rule.e_[i]; - if (w < 0) w = -w; - if (w == 0) return; - const string& cur = TD::Convert(w); + string cur; + if (w > 0) { + cur = TD::Convert(w); + } else { + cur = TD::Convert(nt_types[-w]) + inds[-w]; + } ostringstream os; os << "RBT:" << prev << '_' << cur; const int fid = FD::Convert(Escape(os.str())); diff --git a/decoder/ff_rules.h b/decoder/ff_rules.h index b100ec34..f210dc65 100644 --- a/decoder/ff_rules.h +++ b/decoder/ff_rules.h @@ -51,6 +51,7 @@ class RuleTargetBigramFeatures : public FeatureFunction { void* context) const; virtual void PrepareForInput(const SentenceMetadata& smeta); private: + std::vector<std::string> inds; mutable std::map<const TRule*, SparseVector<double> > rule2_feats_; }; diff --git a/decoder/ff_source_path.cc b/decoder/ff_source_path.cc new file mode 100644 index 00000000..2a3bee2e --- /dev/null +++ b/decoder/ff_source_path.cc @@ -0,0 +1,42 @@ +#include "ff_source_path.h" + +#include "hg.h" + +using namespace std; + +SourcePathFeatures::SourcePathFeatures(const string& param) : FeatureFunction(sizeof(int)) {} + +void SourcePathFeatures::FireBigramFeature(WordID prev, WordID cur, SparseVector<double>* features) const { + int& fid = bigram_fids[prev][cur]; + if (!fid) fid = FD::Convert("SB:"+TD::Convert(prev) + "_" + TD::Convert(cur)); + if (fid) features->add_value(fid, 1.0); +} + +void SourcePathFeatures::FireUnigramFeature(WordID cur, SparseVector<double>* features) const { + int& fid = unigram_fids[cur]; + if (!fid) fid = FD::Convert("SU:" + TD::Convert(cur)); + if (fid) features->add_value(fid, 1.0); +} + +void SourcePathFeatures::TraversalFeaturesImpl(const SentenceMetadata& smeta, + const HG::Edge& edge, + const vector<const void*>& ant_contexts, + SparseVector<double>* features, + SparseVector<double>* estimated_features, + void* context) const { + WordID* res = reinterpret_cast<WordID*>(context); + const vector<int>& f = edge.rule_->f(); + int prev = 0; + unsigned ntc = 0; + for (unsigned i = 0; i < f.size(); ++i) { + int cur = f[i]; + if (cur < 0) + cur = *reinterpret_cast<const WordID*>(ant_contexts[ntc++]); + else + FireUnigramFeature(cur, features); + if (prev) FireBigramFeature(prev, cur, features); + prev = cur; + } + *res = prev; +} + diff --git a/decoder/ff_source_path.h b/decoder/ff_source_path.h new file mode 100644 index 00000000..03126412 --- /dev/null +++ b/decoder/ff_source_path.h @@ -0,0 +1,26 @@ +#ifndef _FF_SOURCE_PATH_H_ +#define _FF_SOURCE_PATH_H_ + +#include <vector> +#include <map> +#include "ff.h" + +class SourcePathFeatures : public FeatureFunction { + public: + SourcePathFeatures(const std::string& param); + protected: + virtual void TraversalFeaturesImpl(const SentenceMetadata& smeta, + const HG::Edge& edge, + const std::vector<const void*>& ant_contexts, + SparseVector<double>* features, + SparseVector<double>* estimated_features, + void* context) const; + + private: + void FireBigramFeature(WordID prev, WordID cur, SparseVector<double>* features) const; + void FireUnigramFeature(WordID cur, SparseVector<double>* features) const; + mutable std::map<WordID, std::map<WordID, int> > bigram_fids; + mutable std::map<WordID, int> unigram_fids; +}; + +#endif diff --git a/environment/LocalConfig.pm b/environment/LocalConfig.pm index b9549c6e..f7c3b1c7 100644 --- a/environment/LocalConfig.pm +++ b/environment/LocalConfig.pm @@ -12,6 +12,7 @@ my $host = domainname; # keys are: HOST_REGEXP, MERTMem, QSubQueue, QSubMemFlag, QSubExtraFlags my $CCONFIG = { + 'StarCluster' => { 'HOST_REGEXP' => qr/compute-\d+\.internal$/, 'JobControl' => 'qsub', @@ -33,7 +34,7 @@ my $CCONFIG = { #'QSubQueue' => '-q long', }, 'UMIACS' => { - 'HOST_REGEXP' => qr/^d.*\.umiacs\.umd\.edu$/, + 'HOST_REGEXP' => qr/^(n|s|d).*\.umiacs\.umd\.edu$/, 'JobControl' => 'qsub', 'QSubMemFlag' => '-l pmem=', 'QSubQueue' => '-q batch', @@ -67,6 +68,12 @@ my $CCONFIG = { 'JobControl' => 'fork', 'DefaultJobs' => 12, }, + 'cluster.cl.uni-heidelberg.de' => { + 'HOST_REGEXP' => qr/node25/, + 'JobControl' => 'qsub', + 'QSubMemFlag' => '-l h_vmem=', + 'DefaultJobs' => 13, + }, 'LOCAL' => { # LOCAL must be last in the list!!! 'HOST_REGEXP' => qr//, 'QSubMemFlag' => ' ', diff --git a/extractor/data_array.cc b/extractor/data_array.cc index 481abb80..203fe219 100644 --- a/extractor/data_array.cc +++ b/extractor/data_array.cc @@ -119,6 +119,7 @@ int DataArray::GetSentenceId(int position) const { } void DataArray::WriteBinary(const fs::path& filepath) const { + std::cerr << "File: " << filepath.string() << std::endl; WriteBinary(fopen(filepath.string().c_str(), "w")); } diff --git a/extractor/suffix_array.cc b/extractor/suffix_array.cc index 9988b1a2..65b2d581 100644 --- a/extractor/suffix_array.cc +++ b/extractor/suffix_array.cc @@ -1,5 +1,6 @@ #include "suffix_array.h" +#include <cassert> #include <chrono> #include <iostream> #include <string> @@ -186,7 +187,8 @@ shared_ptr<DataArray> SuffixArray::GetData() const { } void SuffixArray::WriteBinary(const fs::path& filepath) const { - FILE* file = fopen(filepath.string().c_str(), "r"); + FILE* file = fopen(filepath.string().c_str(), "w"); + assert(file); data_array->WriteBinary(file); int size = suffix_array.size(); diff --git a/mteval/ns.cc b/mteval/ns.cc index d8214558..b64d4798 100644 --- a/mteval/ns.cc +++ b/mteval/ns.cc @@ -61,7 +61,7 @@ string EvaluationMetric::DetailedScore(const SufficientStats& stats) const { return os.str(); } -enum BleuType { IBM, Koehn, NIST }; +enum BleuType { IBM, Koehn, NIST, QCRI }; template <unsigned int N = 4u, BleuType BrevityType = IBM> struct BleuSegmentEvaluator : public SegmentEvaluator { BleuSegmentEvaluator(const vector<vector<WordID> >& refs, const EvaluationMetric* em) : evaluation_metric(em) { @@ -91,7 +91,7 @@ struct BleuSegmentEvaluator : public SegmentEvaluator { float& ref_len = out->fields[2*N + 1]; hyp_len = hyp.size(); ref_len = lengths_[0]; - if (lengths_.size() > 1 && BrevityType == IBM) { + if (lengths_.size() > 1 && (BrevityType == IBM || BrevityType == QCRI)) { float bestd = 2000000; float hl = hyp.size(); float bl = -1; @@ -186,7 +186,7 @@ struct BleuSegmentEvaluator : public SegmentEvaluator { template <unsigned int N = 4u, BleuType BrevityType = IBM> struct BleuMetric : public EvaluationMetric { - BleuMetric() : EvaluationMetric(BrevityType == IBM ? "IBM_BLEU" : (BrevityType == Koehn ? "KOEHN_BLEU" : "NIST_BLEU")) {} + BleuMetric() : EvaluationMetric(BrevityType == IBM ? "IBM_BLEU" : (BrevityType == Koehn ? "KOEHN_BLEU" : (BrevityType == NIST ? "NIST_BLEU" : "QCRI_BLEU"))) {} unsigned SufficientStatisticsVectorSize() const { return N*2 + 2; } boost::shared_ptr<SegmentEvaluator> CreateSegmentEvaluator(const vector<vector<WordID> >& refs) const { return boost::shared_ptr<SegmentEvaluator>(new BleuSegmentEvaluator<N,BrevityType>(refs, this)); @@ -194,26 +194,37 @@ struct BleuMetric : public EvaluationMetric { float ComputeBreakdown(const SufficientStats& stats, float* bp, vector<float>* out) const { if (out) { out->clear(); } float log_bleu = 0; + float log_bleu_adj = 0; // for QCRI int count = 0; + float alpha = BrevityType == QCRI ? 1 : 0.01; for (int i = 0; i < N; ++i) { if (stats.fields[i+N] > 0) { float cor_count = stats.fields[i]; // correct_ngram_hit_counts[i]; // smooth bleu - if (!cor_count) { cor_count = 0.01; } + if (!cor_count) { cor_count = alpha; } float lprec = log(cor_count) - log(stats.fields[i+N]); // log(hyp_ngram_counts[i]); if (out) out->push_back(exp(lprec)); log_bleu += lprec; + if (BrevityType == QCRI) + log_bleu_adj += log(alpha) - log(stats.fields[i+N] + alpha); ++count; } } log_bleu /= count; + log_bleu_adj /= count; float lbp = 0.0; const float& hyp_len = stats.fields[2*N]; const float& ref_len = stats.fields[2*N + 1]; - if (hyp_len < ref_len) - lbp = (hyp_len - ref_len) / hyp_len; + if (hyp_len < ref_len) { + if (BrevityType == QCRI) + lbp = (hyp_len - ref_len - alpha) / hyp_len; + else + lbp = (hyp_len - ref_len) / hyp_len; + } log_bleu += lbp; if (bp) *bp = exp(lbp); + if (BrevityType == QCRI) + return exp(log_bleu) - exp(lbp + log_bleu_adj); return exp(log_bleu); } string DetailedScore(const SufficientStats& stats) const { @@ -253,6 +264,8 @@ EvaluationMetric* EvaluationMetric::Instance(const string& imetric_id) { m = new BleuMetric<4, NIST>; } else if (metric_id == "KOEHN_BLEU") { m = new BleuMetric<4, Koehn>; + } else if (metric_id == "QCRI_BLEU") { + m = new BleuMetric<4, QCRI>; } else if (metric_id == "SSK") { m = new SSKMetric; } else if (metric_id == "TER") { diff --git a/python/pkg/cdec/sa/extract.py b/python/pkg/cdec/sa/extract.py index dc72c18c..b6502c52 100644 --- a/python/pkg/cdec/sa/extract.py +++ b/python/pkg/cdec/sa/extract.py @@ -1,23 +1,26 @@ #!/usr/bin/env python import sys import os +import re +import gzip import argparse import logging -import re -import multiprocessing as mp import signal +import multiprocessing as mp import cdec.sa from cdec.sa._sa import monitor_cpu extractor, prefix = None, None -online = False +online, compress = False, False -def make_extractor(config, grammars, features): - global extractor, prefix, online +def make_extractor(args): + global extractor, prefix, online, compress signal.signal(signal.SIGINT, signal.SIG_IGN) # Let parent process catch Ctrl+C - load_features(features) - extractor = cdec.sa.GrammarExtractor(config, online) - prefix = grammars + load_features(args.features) + extractor = cdec.sa.GrammarExtractor(args.config, online) + prefix = args.grammars + online = args.online + compress = args.compress def load_features(features): for featdef in features: @@ -28,7 +31,7 @@ def load_features(features): sys.path.remove(prefix) def extract(inp): - global extractor, prefix, online + global extractor, prefix, online, compress i, sentence = inp sentence = sentence[:-1] fields = re.split('\s*\|\|\|\s*', sentence) @@ -37,7 +40,7 @@ def extract(inp): if online: if len(fields) < 3: sys.stderr.write('Error: online mode requires references and alignments.' - ' Not adding sentence to training data: {0}\n'.format(sentence)) + ' Not adding sentence to training data: {}\n'.format(sentence)) sentence = fields[0] else: sentence, reference, alignment = fields[0:3] @@ -47,15 +50,17 @@ def extract(inp): if len(fields) > 1: sentence = fields[0] suffix = ' ||| ' + ' ||| '.join(fields[1:]) - grammar_file = os.path.join(prefix, 'grammar.{0}'.format(i)) - with open(grammar_file, 'w') as output: + + grammar_file = os.path.join(prefix, 'grammar.'+str(i)) + if compress: grammar_file += '.gz' + with (gzip.open if compress else open)(grammar_file, 'w') as output: for rule in extractor.grammar(sentence): output.write(str(rule)+'\n') # Add training instance _after_ extracting grammars if online: extractor.add_instance(sentence, reference, alignment) grammar_file = os.path.abspath(grammar_file) - return '<seg grammar="{0}" id="{1}"> {2} </seg>{3}'.format(grammar_file, i, sentence, suffix) + return '<seg grammar="{}" id="{}">{}</seg>{}'.format(grammar_file, i, sentence, suffix) def main(): global online @@ -71,15 +76,17 @@ def main(): help='number of sentences / chunk') parser.add_argument('-f', '--features', nargs='*', default=[], help='additional feature definitions') - parser.add_argument('-o', '--online', action='store_true', default=False, + parser.add_argument('-o', '--online', action='store_true', help='online grammar extraction') + parser.add_argument('-z', '--compress', action='store_true', + help='compress grammars with gzip') args = parser.parse_args() if not os.path.exists(args.grammars): os.mkdir(args.grammars) for featdef in args.features: if not featdef.endswith('.py'): - sys.stderr.write('Error: feature definition file <{0}>' + sys.stderr.write('Error: feature definition file <{}>' ' should be a python module\n'.format(featdef)) sys.exit(1) @@ -88,14 +95,14 @@ def main(): start_time = monitor_cpu() if args.jobs > 1: logging.info('Starting %d workers; chunk size: %d', args.jobs, args.chunksize) - pool = mp.Pool(args.jobs, make_extractor, (args.config, args.grammars, args.features)) + pool = mp.Pool(args.jobs, make_extractor, (args,)) try: for output in pool.imap(extract, enumerate(sys.stdin), args.chunksize): print(output) except KeyboardInterrupt: pool.terminate() else: - make_extractor(args.config, args.grammars, args.features) + make_extractor(args) for output in map(extract, enumerate(sys.stdin)): print(output) diff --git a/python/pkg/cdec/sa/online_extractor.py b/python/pkg/cdec/sa/online_extractor.py deleted file mode 100755 index 03a46b3b..00000000 --- a/python/pkg/cdec/sa/online_extractor.py +++ /dev/null @@ -1,337 +0,0 @@ -#!/usr/bin/env python - -import collections, sys - -import cdec.configobj - -CAT = '[X]' # Default non-terminal -MAX_SIZE = 15 # Max span of a grammar rule (source) -MAX_LEN = 5 # Max number of terminals and non-terminals in a rule (source) -MAX_NT = 2 # Max number of non-terminals in a rule -MIN_GAP = 1 # Min number of terminals between non-terminals (source) - -# Spans are _inclusive_ on both ends [i, j] -# TODO: Replace all of this with bit vectors? -def span_check(vec, i, j): - k = i - while k <= j: - if vec[k]: - return False - k += 1 - return True - -def span_flip(vec, i, j): - k = i - while k <= j: - vec[k] = ~vec[k] - k += 1 - -# Next non-terminal -def next_nt(nt): - if not nt: - return 1 - return nt[-1][0] + 1 - -class NonTerminal: - def __init__(self, index): - self.index = index - def __str__(self): - return '[X,{0}]'.format(self.index) - -def fmt_rule(f_sym, e_sym, links): - a_str = ' '.join('{0}-{1}'.format(i, j) for (i, j) in links) - return '[X] ||| {0} ||| {1} ||| {2}'.format(' '.join(str(sym) for sym in f_sym), - ' '.join(str(sym) for sym in e_sym), - a_str) - -class OnlineGrammarExtractor: - - def __init__(self, config=None): - if isinstance(config, str) or isinstance(config, unicode): - if not os.path.exists(config): - raise IOError('cannot read configuration from {0}'.format(config)) - config = cdec.configobj.ConfigObj(config, unrepr=True) - elif not config: - config = collections.defaultdict(lambda: None) - self.category = CAT - self.max_size = MAX_SIZE - self.max_length = config['max_len'] or MAX_LEN - self.max_nonterminals = config['max_nt'] or MAX_NT - self.min_gap_size = MIN_GAP - # Hard coded: require at least one aligned word - # Hard coded: require tight phrases - - # Phrase counts - self.phrases_f = collections.defaultdict(lambda: 0) - self.phrases_e = collections.defaultdict(lambda: 0) - self.phrases_fe = collections.defaultdict(lambda: collections.defaultdict(lambda: 0)) - - # Bilexical counts - self.bilex_f = collections.defaultdict(lambda: 0) - self.bilex_e = collections.defaultdict(lambda: 0) - self.bilex_fe = collections.defaultdict(lambda: collections.defaultdict(lambda: 0)) - - # Aggregate bilexical counts - def aggr_bilex(self, f_words, e_words): - - for e_w in e_words: - self.bilex_e[e_w] += 1 - - for f_w in f_words: - self.bilex_f[f_w] += 1 - for e_w in e_words: - self.bilex_fe[f_w][e_w] += 1 - - # Aggregate stats from a training instance: - # Extract hierarchical phrase pairs - # Update bilexical counts - def add_instance(self, f_words, e_words, alignment): - - # Bilexical counts - self.aggr_bilex(f_words, e_words) - - # Phrase pairs extracted from this instance - phrases = set() - - f_len = len(f_words) - e_len = len(e_words) - - # Pre-compute alignment info - al = [[] for i in range(f_len)] - al_span = [[f_len + 1, -1] for i in range(f_len)] - for (f, e) in alignment: - al[f].append(e) - al_span[f][0] = min(al_span[f][0], e) - al_span[f][1] = max(al_span[f][1], e) - - # Target side word coverage - # TODO: Does Cython do bit vectors? - cover = [0] * e_len - - # Extract all possible hierarchical phrases starting at a source index - # f_ i and j are current, e_ i and j are previous - def extract(f_i, f_j, e_i, e_j, wc, links, nt, nt_open): - # Phrase extraction limits - if wc + len(nt) > self.max_length or (f_j + 1) > f_len or \ - (f_j - f_i) + 1 > self.max_size: - return - # Unaligned word - if not al[f_j]: - # Open non-terminal: extend - if nt_open: - nt[-1][2] += 1 - extract(f_i, f_j + 1, e_i, e_j, wc, links, nt, True) - nt[-1][2] -= 1 - # No open non-terminal: extend with word - else: - extract(f_i, f_j + 1, e_i, e_j, wc + 1, links, nt, False) - return - # Aligned word - link_i = al_span[f_j][0] - link_j = al_span[f_j][1] - new_e_i = min(link_i, e_i) - new_e_j = max(link_j, e_j) - # Open non-terminal: close, extract, extend - if nt_open: - # Close non-terminal, checking for collisions - old_last_nt = nt[-1][:] - nt[-1][2] = f_j - if link_i < nt[-1][3]: - if not span_check(cover, link_i, nt[-1][3] - 1): - nt[-1] = old_last_nt - return - span_flip(cover, link_i, nt[-1][3] - 1) - nt[-1][3] = link_i - if link_j > nt[-1][4]: - if not span_check(cover, nt[-1][4] + 1, link_j): - nt[-1] = old_last_nt - return - span_flip(cover, nt[-1][4] + 1, link_j) - nt[-1][4] = link_j - for rule in self.form_rules(f_i, new_e_i, f_words[f_i:f_j + 1], e_words[new_e_i:new_e_j + 1], nt, links): - phrases.add(rule) - extract(f_i, f_j + 1, new_e_i, new_e_j, wc, links, nt, False) - nt[-1] = old_last_nt - if link_i < nt[-1][3]: - span_flip(cover, link_i, nt[-1][3] - 1) - if link_j > nt[-1][4]: - span_flip(cover, nt[-1][4] + 1, link_j) - return - # No open non-terminal - # Extract, extend with word - collision = False - for link in al[f_j]: - if cover[link]: - collision = True - # Collisions block extraction and extension, but may be okay for - # continuing non-terminals - if not collision: - plus_links = [] - for link in al[f_j]: - plus_links.append((f_j, link)) - cover[link] = ~cover[link] - links.append(plus_links) - for rule in self.form_rules(f_i, new_e_i, f_words[f_i:f_j + 1], e_words[new_e_i:new_e_j + 1], nt, links): - phrases.add(rule) - extract(f_i, f_j + 1, new_e_i, new_e_j, wc + 1, links, nt, False) - links.pop() - for link in al[f_j]: - cover[link] = ~cover[link] - # Try to add a word to a (closed) non-terminal, extract, extend - if nt and nt[-1][2] == f_j - 1: - # Add to non-terminal, checking for collisions - old_last_nt = nt[-1][:] - nt[-1][2] = f_j - if link_i < nt[-1][3]: - if not span_check(cover, link_i, nt[-1][3] - 1): - nt[-1] = old_last_nt - return - span_flip(cover, link_i, nt[-1][3] - 1) - nt[-1][3] = link_i - if link_j > nt[-1][4]: - if not span_check(cover, nt[-1][4] + 1, link_j): - nt[-1] = old_last_nt - return - span_flip(cover, nt[-1][4] + 1, link_j) - nt[-1][4] = link_j - # Require at least one word in phrase - if links: - for rule in self.form_rules(f_i, new_e_i, f_words[f_i:f_j + 1], e_words[new_e_i:new_e_j + 1], nt, links): - phrases.add(rule) - extract(f_i, f_j + 1, new_e_i, new_e_j, wc, links, nt, False) - nt[-1] = old_last_nt - if new_e_i < nt[-1][3]: - span_flip(cover, link_i, nt[-1][3] - 1) - if link_j > nt[-1][4]: - span_flip(cover, nt[-1][4] + 1, link_j) - # Try to start a new non-terminal, extract, extend - if (not nt or f_j - nt[-1][2] > 1) and len(nt) < self.max_nonterminals: - # Check for collisions - if not span_check(cover, link_i, link_j): - return - span_flip(cover, link_i, link_j) - nt.append([next_nt(nt), f_j, f_j, link_i, link_j]) - # Require at least one word in phrase - if links: - for rule in self.form_rules(f_i, new_e_i, f_words[f_i:f_j + 1], e_words[new_e_i:new_e_j + 1], nt, links): - phrases.add(rule) - extract(f_i, f_j + 1, new_e_i, new_e_j, wc, links, nt, False) - nt.pop() - span_flip(cover, link_i, link_j) - # TODO: try adding NT to start, end, both - # check: one aligned word on boundary that is not part of a NT - - # Try to extract phrases from every f index - f_i = 0 - while f_i < f_len: - # Skip if phrases won't be tight on left side - if not al[f_i]: - f_i += 1 - continue - extract(f_i, f_i, f_len + 1, -1, 1, [], [], False) - f_i += 1 - - for rule in sorted(phrases): - print rule - - # Create a rule from source, target, non-terminals, and alignments - def form_rules(self, f_i, e_i, f_span, e_span, nt, al): - - # This could be more efficient but is unlikely to be the bottleneck - - rules = [] - - nt_inv = sorted(nt, cmp=lambda x, y: cmp(x[3], y[3])) - - f_sym = f_span[:] - off = f_i - for next_nt in nt: - nt_len = (next_nt[2] - next_nt[1]) + 1 - i = 0 - while i < nt_len: - f_sym.pop(next_nt[1] - off) - i += 1 - f_sym.insert(next_nt[1] - off, NonTerminal(next_nt[0])) - off += (nt_len - 1) - - e_sym = e_span[:] - off = e_i - for next_nt in nt_inv: - nt_len = (next_nt[4] - next_nt[3]) + 1 - i = 0 - while i < nt_len: - e_sym.pop(next_nt[3] - off) - i += 1 - e_sym.insert(next_nt[3] - off, NonTerminal(next_nt[0])) - off += (nt_len - 1) - - # Adjusting alignment links takes some doing - links = [list(link) for sub in al for link in sub] - links_len = len(links) - nt_len = len(nt) - nt_i = 0 - off = f_i - i = 0 - while i < links_len: - while nt_i < nt_len and links[i][0] > nt[nt_i][1]: - off += (nt[nt_i][2] - nt[nt_i][1]) - nt_i += 1 - links[i][0] -= off - i += 1 - nt_i = 0 - off = e_i - i = 0 - while i < links_len: - while nt_i < nt_len and links[i][1] > nt_inv[nt_i][3]: - off += (nt_inv[nt_i][4] - nt_inv[nt_i][3]) - nt_i += 1 - links[i][1] -= off - i += 1 - - # Rule - rules.append(fmt_rule(f_sym, e_sym, links)) - if len(f_sym) >= self.max_length or len(nt) >= self.max_nonterminals: - return rules - last_index = nt[-1][0] if nt else 0 - # Rule [X] - if not nt or not isinstance(f_sym[-1], NonTerminal): - f_sym.append(NonTerminal(last_index + 1)) - e_sym.append(NonTerminal(last_index + 1)) - rules.append(fmt_rule(f_sym, e_sym, links)) - f_sym.pop() - e_sym.pop() - # [X] Rule - if not nt or not isinstance(f_sym[0], NonTerminal): - for sym in f_sym: - if isinstance(sym, NonTerminal): - sym.index += 1 - for sym in e_sym: - if isinstance(sym, NonTerminal): - sym.index += 1 - for link in links: - link[0] += 1 - link[1] += 1 - f_sym.insert(0, NonTerminal(1)) - e_sym.insert(0, NonTerminal(1)) - rules.append(fmt_rule(f_sym, e_sym, links)) - if len(f_sym) >= self.max_length or len(nt) + 1 >= self.max_nonterminals: - return rules - # [X] Rule [X] - if not nt or not isinstance(f_sym[-1], NonTerminal): - f_sym.append(NonTerminal(last_index + 2)) - e_sym.append(NonTerminal(last_index + 2)) - rules.append(fmt_rule(f_sym, e_sym, links)) - return rules - -def main(argv): - - extractor = OnlineGrammarExtractor() - - for line in sys.stdin: - print >> sys.stderr, line.strip() - f_words, e_words, a_str = (x.split() for x in line.split('|||')) - alignment = sorted(tuple(int(y) for y in x.split('-')) for x in a_str) - extractor.add_instance(f_words, e_words, alignment) - -if __name__ == '__main__': - main(sys.argv) diff --git a/python/pkg/cdec/score.py b/python/pkg/cdec/score.py index 22257774..829dfdfd 100644 --- a/python/pkg/cdec/score.py +++ b/python/pkg/cdec/score.py @@ -1 +1 @@ -from _cdec import BLEU, TER, CER, Metric +from _cdec import BLEU, TER, CER, SSK, QCRI, Metric diff --git a/python/src/_cdec.cpp b/python/src/_cdec.cpp index 38cf04a9..4b950f01 100644 --- a/python/src/_cdec.cpp +++ b/python/src/_cdec.cpp @@ -1,4 +1,4 @@ -/* Generated by Cython 0.17.1 on Tue Feb 19 16:54:40 2013 */ +/* Generated by Cython 0.18 on Tue Apr 23 18:41:37 2013 */ #define PY_SSIZE_T_CLEAN #include "Python.h" @@ -53,12 +53,15 @@ (PyErr_Format(PyExc_TypeError, \ "expected index value, got %.200s", Py_TYPE(o)->tp_name), \ (PyObject*)0)) - #define PyIndex_Check(o) (PyNumber_Check(o) && !PyFloat_Check(o) && !PyComplex_Check(o)) + #define __Pyx_PyIndex_Check(o) (PyNumber_Check(o) && !PyFloat_Check(o) && \ + !PyComplex_Check(o)) + #define PyIndex_Check __Pyx_PyIndex_Check #define PyErr_WarnEx(category, message, stacklevel) PyErr_Warn(category, message) #define __PYX_BUILD_PY_SSIZE_T "i" #else #define __PYX_BUILD_PY_SSIZE_T "n" #define CYTHON_FORMAT_SSIZE_T "z" + #define __Pyx_PyIndex_Check PyIndex_Check #endif #if PY_VERSION_HEX < 0x02060000 #define Py_REFCNT(ob) (((PyObject*)(ob))->ob_refcnt) @@ -295,8 +298,6 @@ #define CYTHON_WITHOUT_ASSERTIONS #endif - -/* inline attribute */ #ifndef CYTHON_INLINE #if defined(__GNUC__) #define CYTHON_INLINE __inline__ @@ -308,8 +309,6 @@ #define CYTHON_INLINE #endif #endif - -/* unused attribute */ #ifndef CYTHON_UNUSED # if defined(__GNUC__) # if !(defined(__cplusplus)) || (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 4)) @@ -323,24 +322,17 @@ # define CYTHON_UNUSED # endif #endif - typedef struct {PyObject **p; char *s; const long n; const char* encoding; const char is_unicode; const char is_str; const char intern; } __Pyx_StringTabEntry; /*proto*/ - -/* Type Conversion Predeclarations */ - #define __Pyx_PyBytes_FromUString(s) PyBytes_FromString((char*)s) #define __Pyx_PyBytes_AsUString(s) ((unsigned char*) PyBytes_AsString(s)) - #define __Pyx_Owned_Py_None(b) (Py_INCREF(Py_None), Py_None) #define __Pyx_PyBool_FromLong(b) ((b) ? (Py_INCREF(Py_True), Py_True) : (Py_INCREF(Py_False), Py_False)) static CYTHON_INLINE int __Pyx_PyObject_IsTrue(PyObject*); static CYTHON_INLINE PyObject* __Pyx_PyNumber_Int(PyObject* x); - static CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject*); static CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t); static CYTHON_INLINE size_t __Pyx_PyInt_AsSize_t(PyObject*); - #if CYTHON_COMPILING_IN_CPYTHON #define __pyx_PyFloat_AsDouble(x) (PyFloat_CheckExact(x) ? PyFloat_AS_DOUBLE(x) : PyFloat_AsDouble(x)) #else @@ -348,6 +340,7 @@ static CYTHON_INLINE size_t __Pyx_PyInt_AsSize_t(PyObject*); #endif #define __pyx_PyFloat_AsFloat(x) ((float) __pyx_PyFloat_AsDouble(x)) + #ifdef __GNUC__ /* Test for GCC > 2.95 */ #if __GNUC__ > 2 || (__GNUC__ == 2 && (__GNUC_MINOR__ > 95)) @@ -448,7 +441,7 @@ struct __pyx_opt_args_5_cdec_as_str { char *error_msg; }; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":25 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":25 * cdef void read_handle(self, FILE* f) * * cdef class FeatureVector: # <<<<<<<<<<<<<< @@ -475,7 +468,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_24___init__ { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":21 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":21 * return '[%s]' % self.cat * * cdef class NTRef: # <<<<<<<<<<<<<< @@ -488,7 +481,7 @@ struct __pyx_obj_5_cdec_NTRef { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":121 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":121 * return CandidateSet(self) * * cdef class Scorer: # <<<<<<<<<<<<<< @@ -502,7 +495,7 @@ struct __pyx_obj_5_cdec_Scorer { }; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":12 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":12 * cdef void read_handle(self, FILE* f) * * cdef class IntList: # <<<<<<<<<<<<<< @@ -519,7 +512,7 @@ struct __pyx_obj_4cdec_2sa_3_sa_IntList { }; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":29 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":29 * cdef FloatList values * * cdef class Phrase: # <<<<<<<<<<<<<< @@ -536,7 +529,7 @@ struct __pyx_obj_4cdec_2sa_3_sa_Phrase { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":90 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":90 * return candidate * * def __iter__(self): # <<<<<<<<<<<<<< @@ -552,7 +545,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_22___iter__ { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":193 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":193 * super(MRule, self).__init__(lhs, rhs, e, scores, None) * * cdef class Grammar: # <<<<<<<<<<<<<< @@ -565,7 +558,7 @@ struct __pyx_obj_5_cdec_Grammar { }; -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":63 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":63 * def todot(self): * """lattice.todot() -> Representation of the lattice in GraphViz dot format.""" * def lines(): # <<<<<<<<<<<<<< @@ -588,7 +581,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_20_lines { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":108 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":108 * del hypos * * def sample_trees(self, unsigned n): # <<<<<<<<<<<<<< @@ -606,7 +599,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_12_sample_trees { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":161 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":161 * * property edges: * def __get__(self): # <<<<<<<<<<<<<< @@ -622,7 +615,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_13___get__ { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":5 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":5 * import cdec.sa._sa as _sa * * def _phrase(phrase): # <<<<<<<<<<<<<< @@ -635,7 +628,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_2__phrase { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":65 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":65 * return result * * cdef class CandidateSet: # <<<<<<<<<<<<<< @@ -650,7 +643,7 @@ struct __pyx_obj_5_cdec_CandidateSet { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":167 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":167 * * property nodes: * def __get__(self): # <<<<<<<<<<<<<< @@ -666,7 +659,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_14___get__ { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":49 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":49 * return TRule(lhs, f, e, scores, a) * * cdef class TRule: # <<<<<<<<<<<<<< @@ -679,7 +672,7 @@ struct __pyx_obj_5_cdec_TRule { }; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":35 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":35 * cdef public int chunklen(self, int k) * * cdef class Rule: # <<<<<<<<<<<<<< @@ -697,7 +690,7 @@ struct __pyx_obj_4cdec_2sa_3_sa_Rule { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":177 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":177 * _phrase(self.f), _phrase(self.e), scores) * * cdef class MRule(TRule): # <<<<<<<<<<<<<< @@ -709,7 +702,7 @@ struct __pyx_obj_5_cdec_MRule { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":100 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":100 * self.cs.AddKBestCandidates(hypergraph.hg[0], k, self.scorer.get()) * * cdef class SegmentEvaluator: # <<<<<<<<<<<<<< @@ -740,7 +733,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_25_genexpr { }; -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":61 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":61 * yield self[i] * * def todot(self): # <<<<<<<<<<<<<< @@ -753,7 +746,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_19_todot { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":12 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":12 * return stats * * cdef class Candidate: # <<<<<<<<<<<<<< @@ -767,7 +760,7 @@ struct __pyx_obj_5_cdec_Candidate { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":173 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":173 * * def __str__(self): * scores = ' '.join('%s=%s' % feat for feat in self.scores) # <<<<<<<<<<<<<< @@ -784,7 +777,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_6_genexpr { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":8 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":8 * return ' '.join(w.encode('utf8') if isinstance(w, unicode) else str(w) for w in phrase) * * cdef class NT: # <<<<<<<<<<<<<< @@ -798,7 +791,7 @@ struct __pyx_obj_5_cdec_NT { }; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":3 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":3 * from libc.stdio cimport FILE * * cdef class FloatList: # <<<<<<<<<<<<<< @@ -815,7 +808,7 @@ struct __pyx_obj_4cdec_2sa_3_sa_FloatList { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":196 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":196 * return vector * * cdef class HypergraphEdge: # <<<<<<<<<<<<<< @@ -831,7 +824,7 @@ struct __pyx_obj_5_cdec_HypergraphEdge { }; -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":72 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":72 * self.vector.set_value(fid, value) * * def __iter__(self): # <<<<<<<<<<<<<< @@ -848,7 +841,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_1___iter__ { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":256 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":256 * * property in_edges: * def __get__(self): # <<<<<<<<<<<<<< @@ -864,7 +857,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_16___get__ { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":131 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":131 * * property a: * def __get__(self): # <<<<<<<<<<<<<< @@ -881,7 +874,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_4___get__ { }; -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":56 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":56 * return unicode(str(self), 'utf8') * * def __iter__(self): # <<<<<<<<<<<<<< @@ -911,7 +904,7 @@ struct __pyx_obj_5_cdec_Decoder { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":246 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":246 * raise NotImplemented('comparison not implemented for HypergraphEdge') * * cdef class HypergraphNode: # <<<<<<<<<<<<<< @@ -926,7 +919,7 @@ struct __pyx_obj_5_cdec_HypergraphNode { }; -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":48 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":48 * return sparse * * cdef class SparseVector: # <<<<<<<<<<<<<< @@ -939,7 +932,7 @@ struct __pyx_obj_5_cdec_SparseVector { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":262 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":262 * * property out_edges: * def __get__(self): # <<<<<<<<<<<<<< @@ -955,7 +948,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_17___get__ { }; -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":32 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":32 * self.vector[0][fid] = value * * def __iter__(self): # <<<<<<<<<<<<<< @@ -971,7 +964,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct____iter__ { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":44 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":44 * return self.stats.size() * * def __iter__(self): # <<<<<<<<<<<<<< @@ -988,7 +981,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_21___iter__ { }; -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":3 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":3 * from cython.operator cimport preincrement as pinc * * cdef class DenseVector: # <<<<<<<<<<<<<< @@ -1002,7 +995,7 @@ struct __pyx_obj_5_cdec_DenseVector { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":199 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":199 * del self.grammar * * def __iter__(self): # <<<<<<<<<<<<<< @@ -1021,7 +1014,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_7___iter__ { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":176 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":176 * out.fields[i] = ss[i] * * cdef class Metric: # <<<<<<<<<<<<<< @@ -1034,7 +1027,7 @@ struct __pyx_obj_5_cdec_Metric { }; -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":26 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":26 * return fmap * * cdef class SufficientStats: # <<<<<<<<<<<<<< @@ -1048,7 +1041,7 @@ struct __pyx_obj_5_cdec_SufficientStats { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":49 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":49 * return unicode(hypergraph.JoshuaVisualizationString(self.hg[0]).c_str(), 'utf8') * * def kbest(self, size): # <<<<<<<<<<<<<< @@ -1067,7 +1060,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_8_kbest { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":81 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":81 * del e_derivations * * def kbest_features(self, size): # <<<<<<<<<<<<<< @@ -1087,7 +1080,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_10_kbest_features { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":216 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":216 * * property tail_nodes: * def __get__(self): # <<<<<<<<<<<<<< @@ -1103,7 +1096,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_15___get__ { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":172 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":172 * self.rule.get().lhs_ = -TDConvert((<NT> lhs).cat) * * def __str__(self): # <<<<<<<<<<<<<< @@ -1116,7 +1109,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_5___str__ { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":6 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":6 * * def _phrase(phrase): * return ' '.join(w.encode('utf8') if isinstance(w, unicode) else str(w) for w in phrase) # <<<<<<<<<<<<<< @@ -1133,7 +1126,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_3_genexpr { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":62 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":62 * del derivations * * def kbest_trees(self, size): # <<<<<<<<<<<<<< @@ -1156,7 +1149,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_9_kbest_trees { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":4 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":4 * cimport kbest * * cdef class Hypergraph: # <<<<<<<<<<<<<< @@ -1171,7 +1164,7 @@ struct __pyx_obj_5_cdec_Hypergraph { }; -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":3 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":3 * cimport lattice * * cdef class Lattice: # <<<<<<<<<<<<<< @@ -1184,7 +1177,7 @@ struct __pyx_obj_5_cdec_Lattice { }; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":97 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":97 * del derivations * * def sample(self, unsigned n): # <<<<<<<<<<<<<< @@ -1225,7 +1218,7 @@ struct __pyx_obj_5_cdec___pyx_scope_struct_23__make_config { }; -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":217 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":217 * self.grammar.get().SetGrammarName(name) * * cdef class TextGrammar(Grammar): # <<<<<<<<<<<<<< @@ -1238,7 +1231,7 @@ struct __pyx_obj_5_cdec_TextGrammar { -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":4 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":4 * cimport kbest * * cdef class Hypergraph: # <<<<<<<<<<<<<< @@ -1252,7 +1245,7 @@ struct __pyx_vtabstruct_5_cdec_Hypergraph { static struct __pyx_vtabstruct_5_cdec_Hypergraph *__pyx_vtabptr_5_cdec_Hypergraph; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":196 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":196 * return vector * * cdef class HypergraphEdge: # <<<<<<<<<<<<<< @@ -1266,7 +1259,7 @@ struct __pyx_vtabstruct_5_cdec_HypergraphEdge { static struct __pyx_vtabstruct_5_cdec_HypergraphEdge *__pyx_vtabptr_5_cdec_HypergraphEdge; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":12 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":12 * cdef void read_handle(self, FILE* f) * * cdef class IntList: # <<<<<<<<<<<<<< @@ -1286,7 +1279,7 @@ struct __pyx_vtabstruct_4cdec_2sa_3_sa_IntList { static struct __pyx_vtabstruct_4cdec_2sa_3_sa_IntList *__pyx_vtabptr_4cdec_2sa_3_sa_IntList; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":3 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":3 * from libc.stdio cimport FILE * * cdef class FloatList: # <<<<<<<<<<<<<< @@ -1302,7 +1295,7 @@ struct __pyx_vtabstruct_4cdec_2sa_3_sa_FloatList { static struct __pyx_vtabstruct_4cdec_2sa_3_sa_FloatList *__pyx_vtabptr_4cdec_2sa_3_sa_FloatList; -/* "/home/pauldb/workspace/cdec/python/src/cdec.sa._sa.pxd":29 +/* "/usr0/home/cdyer/cdec-merge/python/src/cdec.sa._sa.pxd":29 * cdef FloatList values * * cdef class Phrase: # <<<<<<<<<<<<<< @@ -1317,7 +1310,7 @@ struct __pyx_vtabstruct_4cdec_2sa_3_sa_Phrase { static struct __pyx_vtabstruct_4cdec_2sa_3_sa_Phrase *__pyx_vtabptr_4cdec_2sa_3_sa_Phrase; -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":246 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":246 * raise NotImplemented('comparison not implemented for HypergraphEdge') * * cdef class HypergraphNode: # <<<<<<<<<<<<<< @@ -1530,6 +1523,35 @@ static CYTHON_INLINE int __Pyx_PyDict_Contains(PyObject* item, PyObject* dict, i #define __Pyx_PyIter_Next(obj) __Pyx_PyIter_Next2(obj, NULL) static CYTHON_INLINE PyObject *__Pyx_PyIter_Next2(PyObject *, PyObject *); /*proto*/ +static PyObject* __Pyx_PyDict_GetItemDefault(PyObject* d, PyObject* key, PyObject* default_value) { + PyObject* value; +#if PY_MAJOR_VERSION >= 3 + value = PyDict_GetItemWithError(d, key); + if (unlikely(!value)) { + if (unlikely(PyErr_Occurred())) + return NULL; + value = default_value; + } + Py_INCREF(value); +#else + if (PyString_CheckExact(key) || PyUnicode_CheckExact(key) || PyInt_CheckExact(key)) { + value = PyDict_GetItem(d, key); + if (unlikely(!value)) { + value = default_value; + } + Py_INCREF(value); + } else { + PyObject *m; + m = __Pyx_GetAttrString(d, "get"); + if (!m) return NULL; + value = PyObject_CallFunctionObjArgs(m, key, + (default_value == Py_None) ? NULL : default_value, NULL); + Py_DECREF(m); + } +#endif + return value; +} + static double __Pyx__PyObject_AsDouble(PyObject* obj); /* proto */ #if CYTHON_COMPILING_IN_PYPY #define __Pyx_PyObject_AsDouble(obj) \ @@ -1547,12 +1569,12 @@ static int __Pyx_GetException(PyObject **type, PyObject **value, PyObject **tb); static CYTHON_INLINE void __Pyx_ExceptionSave(PyObject **type, PyObject **value, PyObject **tb); /*proto*/ static void __Pyx_ExceptionReset(PyObject *type, PyObject *value, PyObject *tb); /*proto*/ -static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, long level); /*proto*/ +static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, int level); /*proto*/ static PyObject *__Pyx_FindPy2Metaclass(PyObject *bases); /*proto*/ static PyObject *__Pyx_CreateClass(PyObject *bases, PyObject *dict, PyObject *name, - PyObject *modname); /*proto*/ + PyObject *qualname, PyObject *modname); /*proto*/ #ifndef __Pyx_CppExn2PyErr #include <new> @@ -1616,6 +1638,7 @@ typedef struct { PyObject *func_dict; PyObject *func_weakreflist; PyObject *func_name; + PyObject *func_qualname; PyObject *func_doc; PyObject *func_code; PyObject *func_closure; @@ -1626,10 +1649,10 @@ typedef struct { PyObject *(*defaults_getter)(PyObject *); } __pyx_CyFunctionObject; static PyTypeObject *__pyx_CyFunctionType = 0; -#define __Pyx_CyFunction_NewEx(ml, flags, self, module, code) \ - __Pyx_CyFunction_New(__pyx_CyFunctionType, ml, flags, self, module, code) -static PyObject *__Pyx_CyFunction_New(PyTypeObject *, - PyMethodDef *ml, int flags, +#define __Pyx_CyFunction_NewEx(ml, flags, qualname, self, module, code) \ + __Pyx_CyFunction_New(__pyx_CyFunctionType, ml, flags, qualname, self, module, code) +static PyObject *__Pyx_CyFunction_New(PyTypeObject *, PyMethodDef *ml, + int flags, PyObject* qualname, PyObject *self, PyObject *module, PyObject* code); static CYTHON_INLINE void *__Pyx_CyFunction_InitDefaults(PyObject *m, @@ -2026,20 +2049,21 @@ static char __pyx_k_26[] = "digraph lattice {"; static char __pyx_k_31[] = "\\\""; static char __pyx_k_33[] = "%d [shape=doublecircle]"; static char __pyx_k_34[] = "}"; -static char __pyx_k_37[] = "/home/pauldb/workspace/cdec/python/src/lattice.pxi"; -static char __pyx_k_38[] = "\n"; -static char __pyx_k_40[] = "sufficient stats vector index out of range"; -static char __pyx_k_42[] = "candidate set index out of range"; -static char __pyx_k_44[] = "%s %s"; -static char __pyx_k_45[] = "%s = %s"; +static char __pyx_k_37[] = "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi"; +static char __pyx_k_38[] = "Lattice.todot.<locals>.lines"; +static char __pyx_k_39[] = "\n"; +static char __pyx_k_41[] = "sufficient stats vector index out of range"; +static char __pyx_k_43[] = "candidate set index out of range"; +static char __pyx_k_45[] = "%s %s"; +static char __pyx_k_46[] = "%s = %s"; static char __pyx_k_47[] = "formalism \"%s\" unknown"; static char __pyx_k_48[] = "cannot initialize weights with %s"; static char __pyx_k_49[] = "#"; static char __pyx_k_52[] = "Cannot translate input type %s"; static char __pyx_k_53[] = "cdec.sa._sa"; static char __pyx_k_54[] = "*"; -static char __pyx_k_57[] = "/home/pauldb/workspace/cdec/python/src/grammar.pxi"; -static char __pyx_k_63[] = "/home/pauldb/workspace/cdec/python/src/_cdec.pyx"; +static char __pyx_k_57[] = "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi"; +static char __pyx_k_65[] = "/usr0/home/cdyer/cdec-merge/python/src/_cdec.pyx"; static char __pyx_k__a[] = "a"; static char __pyx_k__e[] = "e"; static char __pyx_k__f[] = "f"; @@ -2048,12 +2072,12 @@ static char __pyx_k__k[] = "k"; static char __pyx_k__pb[] = "pb"; static char __pyx_k__yn[] = "yn"; static char __pyx_k__CER[] = "CER"; +static char __pyx_k__SSK[] = "SSK"; static char __pyx_k__TER[] = "TER"; static char __pyx_k___sa[] = "_sa"; static char __pyx_k__cat[] = "cat"; static char __pyx_k__dot[] = "dot"; static char __pyx_k__fst[] = "fst"; -static char __pyx_k__get[] = "get"; static char __pyx_k__hyp[] = "hyp"; static char __pyx_k__inp[] = "inp"; static char __pyx_k__key[] = "key"; @@ -2062,6 +2086,7 @@ static char __pyx_k__plf[] = "plf"; static char __pyx_k__ref[] = "ref"; static char __pyx_k__rhs[] = "rhs"; static char __pyx_k__BLEU[] = "BLEU"; +static char __pyx_k__QCRI[] = "QCRI"; static char __pyx_k__eval[] = "eval"; static char __pyx_k__info[] = "info"; static char __pyx_k__join[] = "join"; @@ -2111,6 +2136,7 @@ static char __pyx_k__lexalign[] = "lexalign"; static char __pyx_k__lextrans[] = "lextrans"; static char __pyx_k__sentence[] = "sentence"; static char __pyx_k__Exception[] = "Exception"; +static char __pyx_k__QCRI_BLEU[] = "QCRI_BLEU"; static char __pyx_k__TypeError[] = "TypeError"; static char __pyx_k____class__[] = "__class__"; static char __pyx_k____enter__[] = "__enter__"; @@ -2151,12 +2177,13 @@ static PyObject *__pyx_kp_s_31; static PyObject *__pyx_kp_s_33; static PyObject *__pyx_kp_s_34; static PyObject *__pyx_kp_s_37; -static PyObject *__pyx_kp_s_38; +static PyObject *__pyx_n_s_38; +static PyObject *__pyx_kp_s_39; static PyObject *__pyx_kp_s_4; -static PyObject *__pyx_kp_s_40; -static PyObject *__pyx_kp_s_42; -static PyObject *__pyx_kp_s_44; +static PyObject *__pyx_kp_s_41; +static PyObject *__pyx_kp_s_43; static PyObject *__pyx_kp_s_45; +static PyObject *__pyx_kp_s_46; static PyObject *__pyx_kp_s_47; static PyObject *__pyx_kp_s_48; static PyObject *__pyx_kp_s_49; @@ -2164,7 +2191,7 @@ static PyObject *__pyx_kp_s_52; static PyObject *__pyx_n_s_53; static PyObject *__pyx_n_s_54; static PyObject *__pyx_kp_s_57; -static PyObject *__pyx_kp_s_63; +static PyObject *__pyx_kp_s_65; static PyObject *__pyx_kp_s_7; static PyObject *__pyx_kp_s_8; static PyObject *__pyx_kp_s_9; @@ -2177,6 +2204,9 @@ static PyObject *__pyx_n_s__InvalidConfig; static PyObject *__pyx_n_s__KeyError; static PyObject *__pyx_n_s__NotImplemented; static PyObject *__pyx_n_s__ParseFailed; +static PyObject *__pyx_n_s__QCRI; +static PyObject *__pyx_n_s__QCRI_BLEU; +static PyObject *__pyx_n_s__SSK; static PyObject *__pyx_n_s__TER; static PyObject *__pyx_n_s__TypeError; static PyObject *__pyx_n_s__ValueError; @@ -2214,7 +2244,6 @@ static PyObject *__pyx_n_s__formalism; static PyObject *__pyx_n_s__format; static PyObject *__pyx_n_s__fst; static PyObject *__pyx_n_s__genexpr; -static PyObject *__pyx_n_s__get; static PyObject *__pyx_n_s__grammar; static PyObject *__pyx_n_s__hyp; static PyObject *__pyx_n_s__hypergraph; @@ -2270,10 +2299,9 @@ static PyObject *__pyx_k_tuple_24; static PyObject *__pyx_k_tuple_25; static PyObject *__pyx_k_tuple_32; static PyObject *__pyx_k_tuple_35; -static PyObject *__pyx_k_tuple_39; -static PyObject *__pyx_k_tuple_41; -static PyObject *__pyx_k_tuple_43; -static PyObject *__pyx_k_tuple_46; +static PyObject *__pyx_k_tuple_40; +static PyObject *__pyx_k_tuple_42; +static PyObject *__pyx_k_tuple_44; static PyObject *__pyx_k_tuple_50; static PyObject *__pyx_k_tuple_51; static PyObject *__pyx_k_tuple_55; @@ -2281,11 +2309,13 @@ static PyObject *__pyx_k_tuple_58; static PyObject *__pyx_k_tuple_59; static PyObject *__pyx_k_tuple_60; static PyObject *__pyx_k_tuple_61; -static PyObject *__pyx_k_tuple_64; +static PyObject *__pyx_k_tuple_62; +static PyObject *__pyx_k_tuple_63; +static PyObject *__pyx_k_tuple_66; static PyObject *__pyx_k_codeobj_36; static PyObject *__pyx_k_codeobj_56; -static PyObject *__pyx_k_codeobj_62; -static PyObject *__pyx_k_codeobj_65; +static PyObject *__pyx_k_codeobj_64; +static PyObject *__pyx_k_codeobj_67; /* "_cdec.pyx":6 * cimport decoder @@ -2300,8 +2330,8 @@ static PyObject *__pyx_f_5_cdec_as_str(PyObject *__pyx_v_data, struct __pyx_opt_ PyObject *__pyx_v_ret = 0; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; PyObject *__pyx_t_4 = NULL; int __pyx_lineno = 0; @@ -2321,11 +2351,8 @@ static PyObject *__pyx_f_5_cdec_as_str(PyObject *__pyx_v_data, struct __pyx_opt_ * ret = data.encode('utf8') * elif isinstance(data, str): */ - __pyx_t_1 = ((PyObject *)((PyObject*)(&PyUnicode_Type))); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_data, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyUnicode_Check(__pyx_v_data); + if (__pyx_t_1) { /* "_cdec.pyx":9 * cdef bytes ret @@ -2334,11 +2361,11 @@ static PyObject *__pyx_f_5_cdec_as_str(PyObject *__pyx_v_data, struct __pyx_opt_ * elif isinstance(data, str): * ret = data */ - __pyx_t_1 = PyObject_GetAttr(__pyx_v_data, __pyx_n_s__encode); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_k_tuple_2), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = PyObject_GetAttr(__pyx_v_data, __pyx_n_s__encode); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_k_tuple_2), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; if (!(likely(PyBytes_CheckExact(__pyx_t_3))||((__pyx_t_3) == Py_None)||(PyErr_Format(PyExc_TypeError, "Expected bytes, got %.200s", Py_TYPE(__pyx_t_3)->tp_name), 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_ret = ((PyObject*)__pyx_t_3); __pyx_t_3 = 0; @@ -2352,11 +2379,8 @@ static PyObject *__pyx_f_5_cdec_as_str(PyObject *__pyx_v_data, struct __pyx_opt_ * ret = data * else: */ - __pyx_t_3 = ((PyObject *)((PyObject*)(&PyString_Type))); - __Pyx_INCREF(__pyx_t_3); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_data, __pyx_t_3); - __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyString_Check(__pyx_v_data); + if (__pyx_t_1) { /* "_cdec.pyx":11 * ret = data.encode('utf8') @@ -2381,17 +2405,17 @@ static PyObject *__pyx_f_5_cdec_as_str(PyObject *__pyx_v_data, struct __pyx_opt_ */ __pyx_t_3 = PyBytes_FromString(__pyx_v_error_msg); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_3)); - __pyx_t_1 = PyObject_GetAttr(((PyObject *)__pyx_t_3), __pyx_n_s__format); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + __pyx_t_2 = PyObject_GetAttr(((PyObject *)__pyx_t_3), __pyx_n_s__format); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_INCREF(((PyObject *)Py_TYPE(__pyx_v_data))); PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)Py_TYPE(__pyx_v_data))); __Pyx_GIVEREF(((PyObject *)Py_TYPE(__pyx_v_data))); - __pyx_t_4 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_4 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); @@ -2422,7 +2446,7 @@ static PyObject *__pyx_f_5_cdec_as_str(PyObject *__pyx_v_data, struct __pyx_opt_ __pyx_r = ((PyObject*)Py_None); __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_4); __Pyx_AddTraceback("_cdec.as_str", __pyx_clineno, __pyx_lineno, __pyx_filename); @@ -2452,7 +2476,7 @@ static int __pyx_pw_5_cdec_11DenseVector_1__init__(PyObject *__pyx_v_self, PyObj return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":7 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":7 * cdef bint owned # if True, do not manage memory * * def __init__(self): # <<<<<<<<<<<<<< @@ -2469,17 +2493,22 @@ static int __pyx_pf_5_cdec_11DenseVector___init__(struct __pyx_obj_5_cdec_DenseV int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":9 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":9 * def __init__(self): * """DenseVector() -> Dense weight/feature vector.""" * self.vector = new vector[weight_t]() # <<<<<<<<<<<<<< * self.owned = False * */ - try {__pyx_t_1 = new std::vector<weight_t>();} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[1]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_1 = new std::vector<weight_t>(); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[1]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_v_self->vector = __pyx_t_1; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":10 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":10 * """DenseVector() -> Dense weight/feature vector.""" * self.vector = new vector[weight_t]() * self.owned = False # <<<<<<<<<<<<<< @@ -2507,7 +2536,7 @@ static void __pyx_pw_5_cdec_11DenseVector_3__dealloc__(PyObject *__pyx_v_self) { __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":12 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":12 * self.owned = False * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -2520,7 +2549,7 @@ static void __pyx_pf_5_cdec_11DenseVector_2__dealloc__(struct __pyx_obj_5_cdec_D int __pyx_t_1; __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":13 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":13 * * def __dealloc__(self): * if not self.owned: # <<<<<<<<<<<<<< @@ -2530,7 +2559,7 @@ static void __pyx_pf_5_cdec_11DenseVector_2__dealloc__(struct __pyx_obj_5_cdec_D __pyx_t_1 = (!__pyx_v_self->owned); if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":14 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":14 * def __dealloc__(self): * if not self.owned: * del self.vector # <<<<<<<<<<<<<< @@ -2556,7 +2585,7 @@ static Py_ssize_t __pyx_pw_5_cdec_11DenseVector_5__len__(PyObject *__pyx_v_self) return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":16 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":16 * del self.vector * * def __len__(self): # <<<<<<<<<<<<<< @@ -2569,7 +2598,7 @@ static Py_ssize_t __pyx_pf_5_cdec_11DenseVector_4__len__(struct __pyx_obj_5_cdec __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__len__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":17 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":17 * * def __len__(self): * return self.vector.size() # <<<<<<<<<<<<<< @@ -2606,7 +2635,7 @@ static PyObject *__pyx_pw_5_cdec_11DenseVector_7__getitem__(PyObject *__pyx_v_se return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":19 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":19 * return self.vector.size() * * def __getitem__(self, char* fname): # <<<<<<<<<<<<<< @@ -2626,7 +2655,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_6__getitem__(struct __pyx_obj_5_c int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__getitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":20 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":20 * * def __getitem__(self, char* fname): * cdef int fid = FDConvert(fname) # <<<<<<<<<<<<<< @@ -2635,7 +2664,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_6__getitem__(struct __pyx_obj_5_c */ __pyx_v_fid = FD::Convert(__pyx_v_fname); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":21 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":21 * def __getitem__(self, char* fname): * cdef int fid = FDConvert(fname) * if 0 <= fid < self.vector.size(): # <<<<<<<<<<<<<< @@ -2648,7 +2677,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_6__getitem__(struct __pyx_obj_5_c } if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":22 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":22 * cdef int fid = FDConvert(fname) * if 0 <= fid < self.vector.size(): * return self.vector[0][fid] # <<<<<<<<<<<<<< @@ -2665,7 +2694,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_6__getitem__(struct __pyx_obj_5_c } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":23 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":23 * if 0 <= fid < self.vector.size(): * return self.vector[0][fid] * raise KeyError(fname) # <<<<<<<<<<<<<< @@ -2724,7 +2753,7 @@ static int __pyx_pw_5_cdec_11DenseVector_9__setitem__(PyObject *__pyx_v_self, Py return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":25 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":25 * raise KeyError(fname) * * def __setitem__(self, char* fname, float value): # <<<<<<<<<<<<<< @@ -2744,7 +2773,7 @@ static int __pyx_pf_5_cdec_11DenseVector_8__setitem__(struct __pyx_obj_5_cdec_De int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__setitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":26 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":26 * * def __setitem__(self, char* fname, float value): * cdef int fid = FDConvert(fname) # <<<<<<<<<<<<<< @@ -2753,7 +2782,7 @@ static int __pyx_pf_5_cdec_11DenseVector_8__setitem__(struct __pyx_obj_5_cdec_De */ __pyx_v_fid = FD::Convert(__pyx_v_fname); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":27 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":27 * def __setitem__(self, char* fname, float value): * cdef int fid = FDConvert(fname) * if fid < 0: raise KeyError(fname) # <<<<<<<<<<<<<< @@ -2779,7 +2808,7 @@ static int __pyx_pf_5_cdec_11DenseVector_8__setitem__(struct __pyx_obj_5_cdec_De } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":28 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":28 * cdef int fid = FDConvert(fname) * if fid < 0: raise KeyError(fname) * if self.vector.size() <= fid: # <<<<<<<<<<<<<< @@ -2789,7 +2818,7 @@ static int __pyx_pf_5_cdec_11DenseVector_8__setitem__(struct __pyx_obj_5_cdec_De __pyx_t_1 = (__pyx_v_self->vector->size() <= __pyx_v_fid); if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":29 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":29 * if fid < 0: raise KeyError(fname) * if self.vector.size() <= fid: * self.vector.resize(fid + 1) # <<<<<<<<<<<<<< @@ -2801,7 +2830,7 @@ static int __pyx_pf_5_cdec_11DenseVector_8__setitem__(struct __pyx_obj_5_cdec_De } __pyx_L4:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":30 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":30 * if self.vector.size() <= fid: * self.vector.resize(fid + 1) * self.vector[0][fid] = value # <<<<<<<<<<<<<< @@ -2834,7 +2863,7 @@ static PyObject *__pyx_pw_5_cdec_11DenseVector_11__iter__(PyObject *__pyx_v_self return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":32 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":32 * self.vector[0][fid] = value * * def __iter__(self): # <<<<<<<<<<<<<< @@ -2899,7 +2928,7 @@ static PyObject *__pyx_gb_5_cdec_11DenseVector_12generator(__pyx_GeneratorObject __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 32; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":34 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":34 * def __iter__(self): * cdef unsigned fid * for fid in range(1, self.vector.size()): # <<<<<<<<<<<<<< @@ -2910,7 +2939,7 @@ static PyObject *__pyx_gb_5_cdec_11DenseVector_12generator(__pyx_GeneratorObject for (__pyx_t_2 = 1; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_fid = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":35 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":35 * cdef unsigned fid * for fid in range(1, self.vector.size()): * yield str(FDConvert(fid).c_str()), self.vector[0][fid] # <<<<<<<<<<<<<< @@ -2983,7 +3012,7 @@ static PyObject *__pyx_pw_5_cdec_11DenseVector_14dot(PyObject *__pyx_v_self, PyO return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":37 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":37 * yield str(FDConvert(fid).c_str()), self.vector[0][fid] * * def dot(self, SparseVector other): # <<<<<<<<<<<<<< @@ -3002,7 +3031,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_13dot(struct __pyx_obj_5_cdec_Den int __pyx_clineno = 0; __Pyx_RefNannySetupContext("dot", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":39 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":39 * def dot(self, SparseVector other): * """vector.dot(SparseVector other) -> Dot product of the two vectors.""" * return other.dot(self) # <<<<<<<<<<<<<< @@ -3051,7 +3080,7 @@ static PyObject *__pyx_pw_5_cdec_11DenseVector_16tosparse(PyObject *__pyx_v_self return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":41 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":41 * return other.dot(self) * * def tosparse(self): # <<<<<<<<<<<<<< @@ -3069,7 +3098,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_15tosparse(struct __pyx_obj_5_cde int __pyx_clineno = 0; __Pyx_RefNannySetupContext("tosparse", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":43 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":43 * def tosparse(self): * """vector.tosparse() -> Equivalent SparseVector.""" * cdef SparseVector sparse = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -3082,7 +3111,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_15tosparse(struct __pyx_obj_5_cde __pyx_v_sparse = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":44 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":44 * """vector.tosparse() -> Equivalent SparseVector.""" * cdef SparseVector sparse = SparseVector.__new__(SparseVector) * sparse.vector = new FastSparseVector[weight_t]() # <<<<<<<<<<<<<< @@ -3091,7 +3120,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_15tosparse(struct __pyx_obj_5_cde */ __pyx_v_sparse->vector = new FastSparseVector<weight_t>(); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":45 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":45 * cdef SparseVector sparse = SparseVector.__new__(SparseVector) * sparse.vector = new FastSparseVector[weight_t]() * InitSparseVector(self.vector[0], sparse.vector) # <<<<<<<<<<<<<< @@ -3100,7 +3129,7 @@ static PyObject *__pyx_pf_5_cdec_11DenseVector_15tosparse(struct __pyx_obj_5_cde */ Weights::InitSparseVector((__pyx_v_self->vector[0]), __pyx_v_sparse->vector); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":46 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":46 * sparse.vector = new FastSparseVector[weight_t]() * InitSparseVector(self.vector[0], sparse.vector) * return sparse # <<<<<<<<<<<<<< @@ -3143,7 +3172,7 @@ static int __pyx_pw_5_cdec_12SparseVector_1__init__(PyObject *__pyx_v_self, PyOb return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":51 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":51 * cdef FastSparseVector[weight_t]* vector * * def __init__(self): # <<<<<<<<<<<<<< @@ -3156,7 +3185,7 @@ static int __pyx_pf_5_cdec_12SparseVector___init__(struct __pyx_obj_5_cdec_Spars __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":53 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":53 * def __init__(self): * """SparseVector() -> Sparse feature/weight vector.""" * self.vector = new FastSparseVector[weight_t]() # <<<<<<<<<<<<<< @@ -3179,7 +3208,7 @@ static void __pyx_pw_5_cdec_12SparseVector_3__dealloc__(PyObject *__pyx_v_self) __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":55 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":55 * self.vector = new FastSparseVector[weight_t]() * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -3191,7 +3220,7 @@ static void __pyx_pf_5_cdec_12SparseVector_2__dealloc__(CYTHON_UNUSED struct __p __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":56 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":56 * * def __dealloc__(self): * del self.vector # <<<<<<<<<<<<<< @@ -3215,7 +3244,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_5copy(PyObject *__pyx_v_self, CY return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":58 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":58 * del self.vector * * def copy(self): # <<<<<<<<<<<<<< @@ -3232,7 +3261,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_4copy(struct __pyx_obj_5_cdec_Sp int __pyx_clineno = 0; __Pyx_RefNannySetupContext("copy", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":60 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":60 * def copy(self): * """vector.copy() -> SparseVector copy.""" * return self * 1 # <<<<<<<<<<<<<< @@ -3279,7 +3308,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_7__getitem__(PyObject *__pyx_v_s return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":62 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":62 * return self * 1 * * def __getitem__(self, char* fname): # <<<<<<<<<<<<<< @@ -3299,7 +3328,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_6__getitem__(struct __pyx_obj_5_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__getitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":63 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":63 * * def __getitem__(self, char* fname): * cdef int fid = FDConvert(fname) # <<<<<<<<<<<<<< @@ -3308,7 +3337,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_6__getitem__(struct __pyx_obj_5_ */ __pyx_v_fid = FD::Convert(__pyx_v_fname); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":64 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":64 * def __getitem__(self, char* fname): * cdef int fid = FDConvert(fname) * if fid < 0: raise KeyError(fname) # <<<<<<<<<<<<<< @@ -3334,7 +3363,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_6__getitem__(struct __pyx_obj_5_ } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":65 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":65 * cdef int fid = FDConvert(fname) * if fid < 0: raise KeyError(fname) * return self.vector.value(fid) # <<<<<<<<<<<<<< @@ -3386,7 +3415,7 @@ static int __pyx_pw_5_cdec_12SparseVector_9__setitem__(PyObject *__pyx_v_self, P return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":67 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":67 * return self.vector.value(fid) * * def __setitem__(self, char* fname, float value): # <<<<<<<<<<<<<< @@ -3406,7 +3435,7 @@ static int __pyx_pf_5_cdec_12SparseVector_8__setitem__(struct __pyx_obj_5_cdec_S int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__setitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":68 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":68 * * def __setitem__(self, char* fname, float value): * cdef int fid = FDConvert(fname) # <<<<<<<<<<<<<< @@ -3415,7 +3444,7 @@ static int __pyx_pf_5_cdec_12SparseVector_8__setitem__(struct __pyx_obj_5_cdec_S */ __pyx_v_fid = FD::Convert(__pyx_v_fname); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":69 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":69 * def __setitem__(self, char* fname, float value): * cdef int fid = FDConvert(fname) * if fid < 0: raise KeyError(fname) # <<<<<<<<<<<<<< @@ -3441,7 +3470,7 @@ static int __pyx_pf_5_cdec_12SparseVector_8__setitem__(struct __pyx_obj_5_cdec_S } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":70 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":70 * cdef int fid = FDConvert(fname) * if fid < 0: raise KeyError(fname) * self.vector.set_value(fid, value) # <<<<<<<<<<<<<< @@ -3474,7 +3503,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_11__iter__(PyObject *__pyx_v_sel return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":72 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":72 * self.vector.set_value(fid, value) * * def __iter__(self): # <<<<<<<<<<<<<< @@ -3539,7 +3568,7 @@ static PyObject *__pyx_gb_5_cdec_12SparseVector_12generator1(__pyx_GeneratorObje __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":73 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":73 * * def __iter__(self): * cdef FastSparseVector[weight_t].const_iterator* it = new FastSparseVector[weight_t].const_iterator(self.vector[0], False) # <<<<<<<<<<<<<< @@ -3548,7 +3577,7 @@ static PyObject *__pyx_gb_5_cdec_12SparseVector_12generator1(__pyx_GeneratorObje */ __pyx_cur_scope->__pyx_v_it = new FastSparseVector<weight_t>::const_iterator((__pyx_cur_scope->__pyx_v_self->vector[0]), 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":75 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":75 * cdef FastSparseVector[weight_t].const_iterator* it = new FastSparseVector[weight_t].const_iterator(self.vector[0], False) * cdef unsigned i * try: # <<<<<<<<<<<<<< @@ -3557,7 +3586,7 @@ static PyObject *__pyx_gb_5_cdec_12SparseVector_12generator1(__pyx_GeneratorObje */ /*try:*/ { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":76 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":76 * cdef unsigned i * try: * for i in range(self.vector.size()): # <<<<<<<<<<<<<< @@ -3568,7 +3597,7 @@ static PyObject *__pyx_gb_5_cdec_12SparseVector_12generator1(__pyx_GeneratorObje for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":77 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":77 * try: * for i in range(self.vector.size()): * yield (str(FDConvert(it[0].ptr().first).c_str()), it[0].ptr().second) # <<<<<<<<<<<<<< @@ -3609,7 +3638,7 @@ static PyObject *__pyx_gb_5_cdec_12SparseVector_12generator1(__pyx_GeneratorObje __pyx_t_2 = __pyx_cur_scope->__pyx_t_1; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 77; __pyx_clineno = __LINE__; goto __pyx_L5;} - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":78 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":78 * for i in range(self.vector.size()): * yield (str(FDConvert(it[0].ptr().first).c_str()), it[0].ptr().second) * pinc(it[0]) # ++it # <<<<<<<<<<<<<< @@ -3620,7 +3649,7 @@ static PyObject *__pyx_gb_5_cdec_12SparseVector_12generator1(__pyx_GeneratorObje } } - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":80 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":80 * pinc(it[0]) # ++it * finally: * del it # <<<<<<<<<<<<<< @@ -3682,7 +3711,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_14dot(PyObject *__pyx_v_self, Py return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":82 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":82 * del it * * def dot(self, other): # <<<<<<<<<<<<<< @@ -3693,28 +3722,25 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_14dot(PyObject *__pyx_v_self, Py static PyObject *__pyx_pf_5_cdec_12SparseVector_13dot(struct __pyx_obj_5_cdec_SparseVector *__pyx_v_self, PyObject *__pyx_v_other) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("dot", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":84 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":84 * def dot(self, other): * """vector.dot(SparseVector/DenseVector other) -> Dot product of the two vectors.""" * if isinstance(other, DenseVector): # <<<<<<<<<<<<<< * return self.vector.dot((<DenseVector> other).vector[0]) * elif isinstance(other, SparseVector): */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_DenseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_other, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_other, ((PyObject*)__pyx_ptype_5_cdec_DenseVector)); + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":85 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":85 * """vector.dot(SparseVector/DenseVector other) -> Dot product of the two vectors.""" * if isinstance(other, DenseVector): * return self.vector.dot((<DenseVector> other).vector[0]) # <<<<<<<<<<<<<< @@ -3722,28 +3748,25 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_13dot(struct __pyx_obj_5_cdec_Sp * return self.vector.dot((<SparseVector> other).vector[0]) */ __Pyx_XDECREF(__pyx_r); - __pyx_t_1 = PyFloat_FromDouble(__pyx_v_self->vector->dot((((struct __pyx_obj_5_cdec_DenseVector *)__pyx_v_other)->vector[0]))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_r = __pyx_t_1; - __pyx_t_1 = 0; + __pyx_t_2 = PyFloat_FromDouble(__pyx_v_self->vector->dot((((struct __pyx_obj_5_cdec_DenseVector *)__pyx_v_other)->vector[0]))); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 85; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_r = __pyx_t_2; + __pyx_t_2 = 0; goto __pyx_L0; goto __pyx_L3; } - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":86 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":86 * if isinstance(other, DenseVector): * return self.vector.dot((<DenseVector> other).vector[0]) * elif isinstance(other, SparseVector): # <<<<<<<<<<<<<< * return self.vector.dot((<SparseVector> other).vector[0]) * raise TypeError('cannot take the dot product of %s and SparseVector' % type(other)) */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SparseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_other, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_other, ((PyObject*)__pyx_ptype_5_cdec_SparseVector)); + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":87 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":87 * return self.vector.dot((<DenseVector> other).vector[0]) * elif isinstance(other, SparseVector): * return self.vector.dot((<SparseVector> other).vector[0]) # <<<<<<<<<<<<<< @@ -3751,40 +3774,40 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_13dot(struct __pyx_obj_5_cdec_Sp * */ __Pyx_XDECREF(__pyx_r); - __pyx_t_1 = PyFloat_FromDouble(__pyx_v_self->vector->dot((((struct __pyx_obj_5_cdec_SparseVector *)__pyx_v_other)->vector[0]))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 87; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_r = __pyx_t_1; - __pyx_t_1 = 0; + __pyx_t_2 = PyFloat_FromDouble(__pyx_v_self->vector->dot((((struct __pyx_obj_5_cdec_SparseVector *)__pyx_v_other)->vector[0]))); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 87; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_r = __pyx_t_2; + __pyx_t_2 = 0; goto __pyx_L0; goto __pyx_L3; } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":88 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":88 * elif isinstance(other, SparseVector): * return self.vector.dot((<SparseVector> other).vector[0]) * raise TypeError('cannot take the dot product of %s and SparseVector' % type(other)) # <<<<<<<<<<<<<< * * def __richcmp__(SparseVector x, SparseVector y, int op): */ - __pyx_t_1 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_3), ((PyObject *)Py_TYPE(__pyx_v_other))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 88; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(((PyObject *)__pyx_t_1)); + __pyx_t_2 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_3), ((PyObject *)Py_TYPE(__pyx_v_other))); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 88; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(((PyObject *)__pyx_t_2)); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 88; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_t_1)); - __Pyx_GIVEREF(((PyObject *)__pyx_t_1)); - __pyx_t_1 = 0; - __pyx_t_1 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 88; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_t_2)); + __Pyx_GIVEREF(((PyObject *)__pyx_t_2)); + __pyx_t_2 = 0; + __pyx_t_2 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 88; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; - __Pyx_Raise(__pyx_t_1, 0, 0, 0); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_Raise(__pyx_t_2, 0, 0, 0); + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; {__pyx_filename = __pyx_f[1]; __pyx_lineno = 88; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_AddTraceback("_cdec.SparseVector.dot", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; @@ -3811,7 +3834,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_16__richcmp__(PyObject *__pyx_v_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":90 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":90 * raise TypeError('cannot take the dot product of %s and SparseVector' % type(other)) * * def __richcmp__(SparseVector x, SparseVector y, int op): # <<<<<<<<<<<<<< @@ -3829,7 +3852,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_15__richcmp__(struct __pyx_obj_5 int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__richcmp__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":93 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":93 * if op == 2: # == * return x.vector[0] == y.vector[0] * elif op == 3: # != # <<<<<<<<<<<<<< @@ -3838,7 +3861,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_15__richcmp__(struct __pyx_obj_5 */ switch (__pyx_v_op) { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":91 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":91 * * def __richcmp__(SparseVector x, SparseVector y, int op): * if op == 2: # == # <<<<<<<<<<<<<< @@ -3847,7 +3870,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_15__richcmp__(struct __pyx_obj_5 */ case 2: - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":92 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":92 * def __richcmp__(SparseVector x, SparseVector y, int op): * if op == 2: # == * return x.vector[0] == y.vector[0] # <<<<<<<<<<<<<< @@ -3862,7 +3885,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_15__richcmp__(struct __pyx_obj_5 goto __pyx_L0; break; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":93 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":93 * if op == 2: # == * return x.vector[0] == y.vector[0] * elif op == 3: # != # <<<<<<<<<<<<<< @@ -3871,7 +3894,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_15__richcmp__(struct __pyx_obj_5 */ case 3: - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":94 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":94 * return x.vector[0] == y.vector[0] * elif op == 3: # != * return not (x == y) # <<<<<<<<<<<<<< @@ -3890,7 +3913,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_15__richcmp__(struct __pyx_obj_5 break; } - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":95 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":95 * elif op == 3: # != * return not (x == y) * raise NotImplemented('comparison not implemented for SparseVector') # <<<<<<<<<<<<<< @@ -3926,7 +3949,7 @@ static Py_ssize_t __pyx_pw_5_cdec_12SparseVector_18__len__(PyObject *__pyx_v_sel return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":97 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":97 * raise NotImplemented('comparison not implemented for SparseVector') * * def __len__(self): # <<<<<<<<<<<<<< @@ -3939,7 +3962,7 @@ static Py_ssize_t __pyx_pf_5_cdec_12SparseVector_17__len__(struct __pyx_obj_5_cd __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__len__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":98 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":98 * * def __len__(self): * return self.vector.size() # <<<<<<<<<<<<<< @@ -3976,7 +3999,7 @@ static int __pyx_pw_5_cdec_12SparseVector_20__contains__(PyObject *__pyx_v_self, return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":100 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":100 * return self.vector.size() * * def __contains__(self, char* fname): # <<<<<<<<<<<<<< @@ -3989,7 +4012,7 @@ static int __pyx_pf_5_cdec_12SparseVector_19__contains__(struct __pyx_obj_5_cdec __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__contains__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":101 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":101 * * def __contains__(self, char* fname): * return self.vector.nonzero(FDConvert(fname)) # <<<<<<<<<<<<<< @@ -4016,7 +4039,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_22__neg__(PyObject *__pyx_v_self return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":103 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":103 * return self.vector.nonzero(FDConvert(fname)) * * def __neg__(self): # <<<<<<<<<<<<<< @@ -4034,7 +4057,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_21__neg__(struct __pyx_obj_5_cde int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__neg__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":104 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":104 * * def __neg__(self): * cdef SparseVector result = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -4047,7 +4070,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_21__neg__(struct __pyx_obj_5_cde __pyx_v_result = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":105 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":105 * def __neg__(self): * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](self.vector[0]) # <<<<<<<<<<<<<< @@ -4056,7 +4079,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_21__neg__(struct __pyx_obj_5_cde */ __pyx_v_result->vector = new FastSparseVector<weight_t>((__pyx_v_self->vector[0])); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":106 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":106 * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](self.vector[0]) * result.vector[0] *= -1.0 # <<<<<<<<<<<<<< @@ -4065,7 +4088,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_21__neg__(struct __pyx_obj_5_cde */ (__pyx_v_result->vector[0]) *= -1.0; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":107 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":107 * result.vector = new FastSparseVector[weight_t](self.vector[0]) * result.vector[0] *= -1.0 * return result # <<<<<<<<<<<<<< @@ -4106,7 +4129,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_24__iadd__(PyObject *__pyx_v_sel return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":109 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":109 * return result * * def __iadd__(SparseVector self, SparseVector other): # <<<<<<<<<<<<<< @@ -4119,7 +4142,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_23__iadd__(struct __pyx_obj_5_cd __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__iadd__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":110 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":110 * * def __iadd__(SparseVector self, SparseVector other): * self.vector[0] += other.vector[0] # <<<<<<<<<<<<<< @@ -4128,7 +4151,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_23__iadd__(struct __pyx_obj_5_cd */ (__pyx_v_self->vector[0]) += (__pyx_v_other->vector[0]); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":111 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":111 * def __iadd__(SparseVector self, SparseVector other): * self.vector[0] += other.vector[0] * return self # <<<<<<<<<<<<<< @@ -4163,7 +4186,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_26__isub__(PyObject *__pyx_v_sel return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":113 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":113 * return self * * def __isub__(SparseVector self, SparseVector other): # <<<<<<<<<<<<<< @@ -4176,7 +4199,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_25__isub__(struct __pyx_obj_5_cd __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__isub__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":114 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":114 * * def __isub__(SparseVector self, SparseVector other): * self.vector[0] -= other.vector[0] # <<<<<<<<<<<<<< @@ -4185,7 +4208,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_25__isub__(struct __pyx_obj_5_cd */ (__pyx_v_self->vector[0]) -= (__pyx_v_other->vector[0]); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":115 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":115 * def __isub__(SparseVector self, SparseVector other): * self.vector[0] -= other.vector[0] * return self # <<<<<<<<<<<<<< @@ -4225,7 +4248,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_28__imul__(PyObject *__pyx_v_sel return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":117 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":117 * return self * * def __imul__(SparseVector self, float scalar): # <<<<<<<<<<<<<< @@ -4238,7 +4261,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_27__imul__(struct __pyx_obj_5_cd __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__imul__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":118 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":118 * * def __imul__(SparseVector self, float scalar): * self.vector[0] *= scalar # <<<<<<<<<<<<<< @@ -4247,7 +4270,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_27__imul__(struct __pyx_obj_5_cd */ (__pyx_v_self->vector[0]) *= __pyx_v_scalar; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":119 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":119 * def __imul__(SparseVector self, float scalar): * self.vector[0] *= scalar * return self # <<<<<<<<<<<<<< @@ -4289,7 +4312,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_30__idiv__(PyObject *__pyx_v_sel } #endif /*!(#if PY_MAJOR_VERSION < 3)*/ -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":121 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":121 * return self * * def __idiv__(SparseVector self, float scalar): # <<<<<<<<<<<<<< @@ -4303,7 +4326,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_29__idiv__(struct __pyx_obj_5_cd __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__idiv__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":122 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":122 * * def __idiv__(SparseVector self, float scalar): * self.vector[0] /= scalar # <<<<<<<<<<<<<< @@ -4312,7 +4335,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_29__idiv__(struct __pyx_obj_5_cd */ (__pyx_v_self->vector[0]) /= __pyx_v_scalar; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":123 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":123 * def __idiv__(SparseVector self, float scalar): * self.vector[0] /= scalar * return self # <<<<<<<<<<<<<< @@ -4349,7 +4372,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_32__add__(PyObject *__pyx_v_x, P return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":125 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":125 * return self * * def __add__(SparseVector x, SparseVector y): # <<<<<<<<<<<<<< @@ -4367,7 +4390,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_31__add__(struct __pyx_obj_5_cde int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__add__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":126 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":126 * * def __add__(SparseVector x, SparseVector y): * cdef SparseVector result = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -4380,7 +4403,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_31__add__(struct __pyx_obj_5_cde __pyx_v_result = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":127 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":127 * def __add__(SparseVector x, SparseVector y): * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](x.vector[0] + y.vector[0]) # <<<<<<<<<<<<<< @@ -4389,7 +4412,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_31__add__(struct __pyx_obj_5_cde */ __pyx_v_result->vector = new FastSparseVector<weight_t>(((__pyx_v_x->vector[0]) + (__pyx_v_y->vector[0]))); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":128 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":128 * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](x.vector[0] + y.vector[0]) * return result # <<<<<<<<<<<<<< @@ -4431,7 +4454,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_34__sub__(PyObject *__pyx_v_x, P return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":130 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":130 * return result * * def __sub__(SparseVector x, SparseVector y): # <<<<<<<<<<<<<< @@ -4449,7 +4472,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_33__sub__(struct __pyx_obj_5_cde int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__sub__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":131 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":131 * * def __sub__(SparseVector x, SparseVector y): * cdef SparseVector result = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -4462,7 +4485,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_33__sub__(struct __pyx_obj_5_cde __pyx_v_result = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":132 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":132 * def __sub__(SparseVector x, SparseVector y): * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](x.vector[0] - y.vector[0]) # <<<<<<<<<<<<<< @@ -4471,7 +4494,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_33__sub__(struct __pyx_obj_5_cde */ __pyx_v_result->vector = new FastSparseVector<weight_t>(((__pyx_v_x->vector[0]) - (__pyx_v_y->vector[0]))); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":133 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":133 * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](x.vector[0] - y.vector[0]) * return result # <<<<<<<<<<<<<< @@ -4507,7 +4530,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_36__mul__(PyObject *__pyx_v_x, P return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":135 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":135 * return result * * def __mul__(x, y): # <<<<<<<<<<<<<< @@ -4521,38 +4544,35 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_35__mul__(PyObject *__pyx_v_x, P struct __pyx_obj_5_cdec_SparseVector *__pyx_v_result = 0; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; float __pyx_t_3; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__mul__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":138 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":138 * cdef SparseVector vector * cdef float scalar * if isinstance(x, SparseVector): vector, scalar = x, y # <<<<<<<<<<<<<< * else: vector, scalar = y, x * cdef SparseVector result = SparseVector.__new__(SparseVector) */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SparseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_x, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_x, ((PyObject*)__pyx_ptype_5_cdec_SparseVector)); + if (__pyx_t_1) { if (!(likely(((__pyx_v_x) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_x, __pyx_ptype_5_cdec_SparseVector))))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_1 = __pyx_v_x; - __Pyx_INCREF(__pyx_t_1); + __pyx_t_2 = __pyx_v_x; + __Pyx_INCREF(__pyx_t_2); __pyx_t_3 = __pyx_PyFloat_AsFloat(__pyx_v_y); if (unlikely((__pyx_t_3 == (float)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_2); + __pyx_t_2 = 0; __pyx_v_scalar = __pyx_t_3; goto __pyx_L3; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":139 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":139 * cdef float scalar * if isinstance(x, SparseVector): vector, scalar = x, y * else: vector, scalar = y, x # <<<<<<<<<<<<<< @@ -4560,29 +4580,29 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_35__mul__(PyObject *__pyx_v_x, P * result.vector = new FastSparseVector[weight_t](vector.vector[0] * scalar) */ if (!(likely(((__pyx_v_y) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_y, __pyx_ptype_5_cdec_SparseVector))))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_1 = __pyx_v_y; - __Pyx_INCREF(__pyx_t_1); + __pyx_t_2 = __pyx_v_y; + __Pyx_INCREF(__pyx_t_2); __pyx_t_3 = __pyx_PyFloat_AsFloat(__pyx_v_x); if (unlikely((__pyx_t_3 == (float)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_2); + __pyx_t_2 = 0; __pyx_v_scalar = __pyx_t_3; } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":140 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":140 * if isinstance(x, SparseVector): vector, scalar = x, y * else: vector, scalar = y, x * cdef SparseVector result = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< * result.vector = new FastSparseVector[weight_t](vector.vector[0] * scalar) * return result */ - __pyx_t_1 = __Pyx_tp_new(((PyObject*)__pyx_ptype_5_cdec_SparseVector)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - if (!(likely(__Pyx_TypeTest(__pyx_t_1, __pyx_ptype_5_cdec_SparseVector)))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_v_result = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_t_2 = __Pyx_tp_new(((PyObject*)__pyx_ptype_5_cdec_SparseVector)); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + if (!(likely(__Pyx_TypeTest(__pyx_t_2, __pyx_ptype_5_cdec_SparseVector)))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 140; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_v_result = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_2); + __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":141 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":141 * else: vector, scalar = y, x * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](vector.vector[0] * scalar) # <<<<<<<<<<<<<< @@ -4591,7 +4611,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_35__mul__(PyObject *__pyx_v_x, P */ __pyx_v_result->vector = new FastSparseVector<weight_t>(((__pyx_v_vector->vector[0]) * __pyx_v_scalar)); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":142 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":142 * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](vector.vector[0] * scalar) * return result # <<<<<<<<<<<<<< @@ -4606,7 +4626,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_35__mul__(PyObject *__pyx_v_x, P __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_AddTraceback("_cdec.SparseVector.__mul__", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; __pyx_L0:; @@ -4630,7 +4650,7 @@ static PyObject *__pyx_pw_5_cdec_12SparseVector_38__div__(PyObject *__pyx_v_x, P } #endif /*!(#if PY_MAJOR_VERSION < 3)*/ -/* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":144 +/* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":144 * return result * * def __div__(x, y): # <<<<<<<<<<<<<< @@ -4645,38 +4665,35 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_37__div__(PyObject *__pyx_v_x, P struct __pyx_obj_5_cdec_SparseVector *__pyx_v_result = 0; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; float __pyx_t_3; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__div__", 0); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":147 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":147 * cdef SparseVector vector * cdef float scalar * if isinstance(x, SparseVector): vector, scalar = x, y # <<<<<<<<<<<<<< * else: vector, scalar = y, x * cdef SparseVector result = SparseVector.__new__(SparseVector) */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SparseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_x, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_x, ((PyObject*)__pyx_ptype_5_cdec_SparseVector)); + if (__pyx_t_1) { if (!(likely(((__pyx_v_x) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_x, __pyx_ptype_5_cdec_SparseVector))))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 147; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_1 = __pyx_v_x; - __Pyx_INCREF(__pyx_t_1); + __pyx_t_2 = __pyx_v_x; + __Pyx_INCREF(__pyx_t_2); __pyx_t_3 = __pyx_PyFloat_AsFloat(__pyx_v_y); if (unlikely((__pyx_t_3 == (float)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 147; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_2); + __pyx_t_2 = 0; __pyx_v_scalar = __pyx_t_3; goto __pyx_L3; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":148 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":148 * cdef float scalar * if isinstance(x, SparseVector): vector, scalar = x, y * else: vector, scalar = y, x # <<<<<<<<<<<<<< @@ -4684,29 +4701,29 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_37__div__(PyObject *__pyx_v_x, P * result.vector = new FastSparseVector[weight_t](vector.vector[0] / scalar) */ if (!(likely(((__pyx_v_y) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_y, __pyx_ptype_5_cdec_SparseVector))))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 148; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_1 = __pyx_v_y; - __Pyx_INCREF(__pyx_t_1); + __pyx_t_2 = __pyx_v_y; + __Pyx_INCREF(__pyx_t_2); __pyx_t_3 = __pyx_PyFloat_AsFloat(__pyx_v_x); if (unlikely((__pyx_t_3 == (float)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 148; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_2); + __pyx_t_2 = 0; __pyx_v_scalar = __pyx_t_3; } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":149 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":149 * if isinstance(x, SparseVector): vector, scalar = x, y * else: vector, scalar = y, x * cdef SparseVector result = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< * result.vector = new FastSparseVector[weight_t](vector.vector[0] / scalar) * return result */ - __pyx_t_1 = __Pyx_tp_new(((PyObject*)__pyx_ptype_5_cdec_SparseVector)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 149; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - if (!(likely(__Pyx_TypeTest(__pyx_t_1, __pyx_ptype_5_cdec_SparseVector)))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 149; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_v_result = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_t_2 = __Pyx_tp_new(((PyObject*)__pyx_ptype_5_cdec_SparseVector)); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 149; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + if (!(likely(__Pyx_TypeTest(__pyx_t_2, __pyx_ptype_5_cdec_SparseVector)))) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 149; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_v_result = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_2); + __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":150 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":150 * else: vector, scalar = y, x * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](vector.vector[0] / scalar) # <<<<<<<<<<<<<< @@ -4714,7 +4731,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_37__div__(PyObject *__pyx_v_x, P */ __pyx_v_result->vector = new FastSparseVector<weight_t>(((__pyx_v_vector->vector[0]) / __pyx_v_scalar)); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":151 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":151 * cdef SparseVector result = SparseVector.__new__(SparseVector) * result.vector = new FastSparseVector[weight_t](vector.vector[0] / scalar) * return result # <<<<<<<<<<<<<< @@ -4727,7 +4744,7 @@ static PyObject *__pyx_pf_5_cdec_12SparseVector_37__div__(PyObject *__pyx_v_x, P __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_AddTraceback("_cdec.SparseVector.__div__", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; __pyx_L0:; @@ -4752,7 +4769,7 @@ static PyObject *__pyx_pw_5_cdec_1_phrase(PyObject *__pyx_self, PyObject *__pyx_ } static PyObject *__pyx_gb_5_cdec_7_phrase_2generator18(__pyx_GeneratorObject *__pyx_generator, PyObject *__pyx_sent_value); /* proto */ -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":6 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":6 * * def _phrase(phrase): * return ' '.join(w.encode('utf8') if isinstance(w, unicode) else str(w) for w in phrase) # <<<<<<<<<<<<<< @@ -4804,8 +4821,8 @@ static PyObject *__pyx_gb_5_cdec_7_phrase_2generator18(__pyx_GeneratorObject *__ Py_ssize_t __pyx_t_2; PyObject *(*__pyx_t_3)(PyObject *); PyObject *__pyx_t_4 = NULL; - PyObject *__pyx_t_5 = NULL; - int __pyx_t_6; + int __pyx_t_5; + PyObject *__pyx_t_6 = NULL; PyObject *__pyx_t_7 = NULL; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("None", 0); @@ -4858,16 +4875,13 @@ static PyObject *__pyx_gb_5_cdec_7_phrase_2generator18(__pyx_GeneratorObject *__ __Pyx_GIVEREF(__pyx_t_4); __pyx_cur_scope->__pyx_v_w = __pyx_t_4; __pyx_t_4 = 0; - __pyx_t_5 = ((PyObject *)((PyObject*)(&PyUnicode_Type))); - __Pyx_INCREF(__pyx_t_5); - __pyx_t_6 = __Pyx_TypeCheck(__pyx_cur_scope->__pyx_v_w, __pyx_t_5); - __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; - if (__pyx_t_6) { - __pyx_t_5 = PyObject_GetAttr(__pyx_cur_scope->__pyx_v_w, __pyx_n_s__encode); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_5); - __pyx_t_7 = PyObject_Call(__pyx_t_5, ((PyObject *)__pyx_k_tuple_6), NULL); if (unlikely(!__pyx_t_7)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_5 = PyUnicode_Check(__pyx_cur_scope->__pyx_v_w); + if (__pyx_t_5) { + __pyx_t_6 = PyObject_GetAttr(__pyx_cur_scope->__pyx_v_w, __pyx_n_s__encode); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_6); + __pyx_t_7 = PyObject_Call(__pyx_t_6, ((PyObject *)__pyx_k_tuple_6), NULL); if (unlikely(!__pyx_t_7)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_7); - __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; + __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; __pyx_t_4 = __pyx_t_7; __pyx_t_7 = 0; } else { @@ -4876,11 +4890,11 @@ static PyObject *__pyx_gb_5_cdec_7_phrase_2generator18(__pyx_GeneratorObject *__ __Pyx_INCREF(__pyx_cur_scope->__pyx_v_w); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_cur_scope->__pyx_v_w); __Pyx_GIVEREF(__pyx_cur_scope->__pyx_v_w); - __pyx_t_5 = PyObject_Call(((PyObject *)((PyObject*)(&PyString_Type))), ((PyObject *)__pyx_t_7), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_5); + __pyx_t_6 = PyObject_Call(((PyObject *)((PyObject*)(&PyString_Type))), ((PyObject *)__pyx_t_7), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_6); __Pyx_DECREF(((PyObject *)__pyx_t_7)); __pyx_t_7 = 0; - __pyx_t_4 = __pyx_t_5; - __pyx_t_5 = 0; + __pyx_t_4 = __pyx_t_6; + __pyx_t_6 = 0; } __pyx_r = __pyx_t_4; __pyx_t_4 = 0; @@ -4907,7 +4921,7 @@ static PyObject *__pyx_gb_5_cdec_7_phrase_2generator18(__pyx_GeneratorObject *__ __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_4); - __Pyx_XDECREF(__pyx_t_5); + __Pyx_XDECREF(__pyx_t_6); __Pyx_XDECREF(__pyx_t_7); __Pyx_AddTraceback("genexpr", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_L0:; @@ -4918,7 +4932,7 @@ static PyObject *__pyx_gb_5_cdec_7_phrase_2generator18(__pyx_GeneratorObject *__ return NULL; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":5 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":5 * import cdec.sa._sa as _sa * * def _phrase(phrase): # <<<<<<<<<<<<<< @@ -4947,7 +4961,7 @@ static PyObject *__pyx_pf_5_cdec__phrase(CYTHON_UNUSED PyObject *__pyx_self, PyO __Pyx_INCREF(__pyx_cur_scope->__pyx_v_phrase); __Pyx_GIVEREF(__pyx_cur_scope->__pyx_v_phrase); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":6 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":6 * * def _phrase(phrase): * return ' '.join(w.encode('utf8') if isinstance(w, unicode) else str(w) for w in phrase) # <<<<<<<<<<<<<< @@ -5058,7 +5072,7 @@ static int __pyx_pw_5_cdec_2NT_1__init__(PyObject *__pyx_v_self, PyObject *__pyx return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":11 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":11 * cdef public bytes cat * cdef public unsigned ref * def __init__(self, bytes cat, unsigned ref=0): # <<<<<<<<<<<<<< @@ -5071,7 +5085,7 @@ static int __pyx_pf_5_cdec_2NT___init__(struct __pyx_obj_5_cdec_NT *__pyx_v_self __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":13 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":13 * def __init__(self, bytes cat, unsigned ref=0): * """NT(bytes cat, int ref=0) -> Non-terminal from category `cat`.""" * self.cat = cat # <<<<<<<<<<<<<< @@ -5084,7 +5098,7 @@ static int __pyx_pf_5_cdec_2NT___init__(struct __pyx_obj_5_cdec_NT *__pyx_v_self __Pyx_DECREF(((PyObject *)__pyx_v_self->cat)); __pyx_v_self->cat = __pyx_v_cat; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":14 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":14 * """NT(bytes cat, int ref=0) -> Non-terminal from category `cat`.""" * self.cat = cat * self.ref = ref # <<<<<<<<<<<<<< @@ -5109,7 +5123,7 @@ static PyObject *__pyx_pw_5_cdec_2NT_3__str__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":16 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":16 * self.ref = ref * * def __str__(self): # <<<<<<<<<<<<<< @@ -5128,7 +5142,7 @@ static PyObject *__pyx_pf_5_cdec_2NT_2__str__(struct __pyx_obj_5_cdec_NT *__pyx_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__str__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":17 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":17 * * def __str__(self): * if self.ref > 0: # <<<<<<<<<<<<<< @@ -5138,7 +5152,7 @@ static PyObject *__pyx_pf_5_cdec_2NT_2__str__(struct __pyx_obj_5_cdec_NT *__pyx_ __pyx_t_1 = (__pyx_v_self->ref > 0); if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":18 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":18 * def __str__(self): * if self.ref > 0: * return '[%s,%d]' % (self.cat, self.ref) # <<<<<<<<<<<<<< @@ -5166,7 +5180,7 @@ static PyObject *__pyx_pf_5_cdec_2NT_2__str__(struct __pyx_obj_5_cdec_NT *__pyx_ } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":19 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":19 * if self.ref > 0: * return '[%s,%d]' % (self.cat, self.ref) * return '[%s]' % self.cat # <<<<<<<<<<<<<< @@ -5204,7 +5218,7 @@ static PyObject *__pyx_pw_5_cdec_2NT_3cat_1__get__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":9 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":9 * * cdef class NT: * cdef public bytes cat # <<<<<<<<<<<<<< @@ -5300,7 +5314,7 @@ static PyObject *__pyx_pw_5_cdec_2NT_3ref_1__get__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":10 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":10 * cdef class NT: * cdef public bytes cat * cdef public unsigned ref # <<<<<<<<<<<<<< @@ -5418,7 +5432,7 @@ static int __pyx_pw_5_cdec_5NTRef_1__init__(PyObject *__pyx_v_self, PyObject *__ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":23 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":23 * cdef class NTRef: * cdef public unsigned ref * def __init__(self, unsigned ref): # <<<<<<<<<<<<<< @@ -5431,7 +5445,7 @@ static int __pyx_pf_5_cdec_5NTRef___init__(struct __pyx_obj_5_cdec_NTRef *__pyx_ __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":25 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":25 * def __init__(self, unsigned ref): * """NTRef(int ref) -> Non-terminal reference.""" * self.ref = ref # <<<<<<<<<<<<<< @@ -5456,7 +5470,7 @@ static PyObject *__pyx_pw_5_cdec_5NTRef_3__str__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":27 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":27 * self.ref = ref * * def __str__(self): # <<<<<<<<<<<<<< @@ -5474,7 +5488,7 @@ static PyObject *__pyx_pf_5_cdec_5NTRef_2__str__(struct __pyx_obj_5_cdec_NTRef * int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__str__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":28 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":28 * * def __str__(self): * return '[%d]' % self.ref # <<<<<<<<<<<<<< @@ -5515,7 +5529,7 @@ static PyObject *__pyx_pw_5_cdec_5NTRef_3ref_1__get__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":22 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":22 * * cdef class NTRef: * cdef public unsigned ref # <<<<<<<<<<<<<< @@ -5582,7 +5596,7 @@ static int __pyx_pf_5_cdec_5NTRef_3ref_2__set__(struct __pyx_obj_5_cdec_NTRef *_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":30 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":30 * return '[%d]' % self.ref * * cdef TRule convert_rule(_sa.Rule rule): # <<<<<<<<<<<<<< @@ -5603,16 +5617,17 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; PyObject *__pyx_t_2 = NULL; - int __pyx_t_3; + int *__pyx_t_3; int __pyx_t_4; int __pyx_t_5; int __pyx_t_6; + int __pyx_t_7; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("convert_rule", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":31 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":31 * * cdef TRule convert_rule(_sa.Rule rule): * lhs = _sa.sym_tocat(rule.lhs) # <<<<<<<<<<<<<< @@ -5621,7 +5636,7 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o */ __pyx_v_lhs = __pyx_f_4cdec_2sa_3_sa_sym_tocat(__pyx_v_rule->lhs); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":32 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":32 * cdef TRule convert_rule(_sa.Rule rule): * lhs = _sa.sym_tocat(rule.lhs) * scores = dict(rule.scores) # <<<<<<<<<<<<<< @@ -5639,7 +5654,7 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o __pyx_v_scores = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":33 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":33 * lhs = _sa.sym_tocat(rule.lhs) * scores = dict(rule.scores) * f, e = [], [] # <<<<<<<<<<<<<< @@ -5650,42 +5665,43 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o __Pyx_GOTREF(__pyx_t_2); __pyx_t_1 = PyList_New(0); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 33; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); - __pyx_v_f = __pyx_t_2; + __pyx_v_f = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0; - __pyx_v_e = __pyx_t_1; + __pyx_v_e = ((PyObject*)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":34 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":34 * scores = dict(rule.scores) * f, e = [], [] * cdef int* fsyms = rule.f.syms # <<<<<<<<<<<<<< * for i in range(rule.f.n): * if _sa.sym_isvar(fsyms[i]): */ - __pyx_v_fsyms = __pyx_v_rule->f->syms; + __pyx_t_3 = __pyx_v_rule->f->syms; + __pyx_v_fsyms = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":35 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":35 * f, e = [], [] * cdef int* fsyms = rule.f.syms * for i in range(rule.f.n): # <<<<<<<<<<<<<< * if _sa.sym_isvar(fsyms[i]): * f.append(NT(_sa.sym_tocat(fsyms[i]))) */ - __pyx_t_3 = __pyx_v_rule->f->n; - for (__pyx_t_4 = 0; __pyx_t_4 < __pyx_t_3; __pyx_t_4+=1) { - __pyx_v_i = __pyx_t_4; + __pyx_t_4 = __pyx_v_rule->f->n; + for (__pyx_t_5 = 0; __pyx_t_5 < __pyx_t_4; __pyx_t_5+=1) { + __pyx_v_i = __pyx_t_5; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":36 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":36 * cdef int* fsyms = rule.f.syms * for i in range(rule.f.n): * if _sa.sym_isvar(fsyms[i]): # <<<<<<<<<<<<<< * f.append(NT(_sa.sym_tocat(fsyms[i]))) * else: */ - __pyx_t_5 = __pyx_f_4cdec_2sa_3_sa_sym_isvar((__pyx_v_fsyms[__pyx_v_i])); - if (__pyx_t_5) { + __pyx_t_6 = __pyx_f_4cdec_2sa_3_sa_sym_isvar((__pyx_v_fsyms[__pyx_v_i])); + if (__pyx_t_6) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":37 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":37 * for i in range(rule.f.n): * if _sa.sym_isvar(fsyms[i]): * f.append(NT(_sa.sym_tocat(fsyms[i]))) # <<<<<<<<<<<<<< @@ -5702,13 +5718,13 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o __pyx_t_1 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NT)), ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 37; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; - __pyx_t_6 = PyList_Append(__pyx_v_f, __pyx_t_1); if (unlikely(__pyx_t_6 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 37; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_7 = PyList_Append(__pyx_v_f, __pyx_t_1); if (unlikely(__pyx_t_7 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 37; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; goto __pyx_L5; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":39 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":39 * f.append(NT(_sa.sym_tocat(fsyms[i]))) * else: * f.append(_sa.sym_tostring(fsyms[i])) # <<<<<<<<<<<<<< @@ -5717,43 +5733,44 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o */ __pyx_t_1 = PyBytes_FromString(__pyx_f_4cdec_2sa_3_sa_sym_tostring((__pyx_v_fsyms[__pyx_v_i]))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 39; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_1)); - __pyx_t_6 = PyList_Append(__pyx_v_f, ((PyObject *)__pyx_t_1)); if (unlikely(__pyx_t_6 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 39; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_7 = PyList_Append(__pyx_v_f, ((PyObject *)__pyx_t_1)); if (unlikely(__pyx_t_7 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 39; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; } __pyx_L5:; } - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":40 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":40 * else: * f.append(_sa.sym_tostring(fsyms[i])) * cdef int* esyms = rule.e.syms # <<<<<<<<<<<<<< * for i in range(rule.e.n): * if _sa.sym_isvar(esyms[i]): */ - __pyx_v_esyms = __pyx_v_rule->e->syms; + __pyx_t_3 = __pyx_v_rule->e->syms; + __pyx_v_esyms = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":41 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":41 * f.append(_sa.sym_tostring(fsyms[i])) * cdef int* esyms = rule.e.syms * for i in range(rule.e.n): # <<<<<<<<<<<<<< * if _sa.sym_isvar(esyms[i]): * e.append(NTRef(_sa.sym_getindex(esyms[i]))) */ - __pyx_t_3 = __pyx_v_rule->e->n; - for (__pyx_t_4 = 0; __pyx_t_4 < __pyx_t_3; __pyx_t_4+=1) { - __pyx_v_i = __pyx_t_4; + __pyx_t_4 = __pyx_v_rule->e->n; + for (__pyx_t_5 = 0; __pyx_t_5 < __pyx_t_4; __pyx_t_5+=1) { + __pyx_v_i = __pyx_t_5; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":42 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":42 * cdef int* esyms = rule.e.syms * for i in range(rule.e.n): * if _sa.sym_isvar(esyms[i]): # <<<<<<<<<<<<<< * e.append(NTRef(_sa.sym_getindex(esyms[i]))) * else: */ - __pyx_t_5 = __pyx_f_4cdec_2sa_3_sa_sym_isvar((__pyx_v_esyms[__pyx_v_i])); - if (__pyx_t_5) { + __pyx_t_6 = __pyx_f_4cdec_2sa_3_sa_sym_isvar((__pyx_v_esyms[__pyx_v_i])); + if (__pyx_t_6) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":43 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":43 * for i in range(rule.e.n): * if _sa.sym_isvar(esyms[i]): * e.append(NTRef(_sa.sym_getindex(esyms[i]))) # <<<<<<<<<<<<<< @@ -5770,13 +5787,13 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o __pyx_t_1 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NTRef)), ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 43; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; - __pyx_t_6 = PyList_Append(__pyx_v_e, __pyx_t_1); if (unlikely(__pyx_t_6 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 43; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_7 = PyList_Append(__pyx_v_e, __pyx_t_1); if (unlikely(__pyx_t_7 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 43; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; goto __pyx_L8; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":45 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":45 * e.append(NTRef(_sa.sym_getindex(esyms[i]))) * else: * e.append(_sa.sym_tostring(esyms[i])) # <<<<<<<<<<<<<< @@ -5785,13 +5802,13 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o */ __pyx_t_1 = PyBytes_FromString(__pyx_f_4cdec_2sa_3_sa_sym_tostring((__pyx_v_esyms[__pyx_v_i]))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 45; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_1)); - __pyx_t_6 = PyList_Append(__pyx_v_e, ((PyObject *)__pyx_t_1)); if (unlikely(__pyx_t_6 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 45; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_7 = PyList_Append(__pyx_v_e, ((PyObject *)__pyx_t_1)); if (unlikely(__pyx_t_7 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 45; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; } __pyx_L8:; } - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":46 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":46 * else: * e.append(_sa.sym_tostring(esyms[i])) * a = list(rule.alignments()) # <<<<<<<<<<<<<< @@ -5814,7 +5831,7 @@ static struct __pyx_obj_5_cdec_TRule *__pyx_f_5_cdec_convert_rule(struct __pyx_o __pyx_v_a = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":47 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":47 * e.append(_sa.sym_tostring(esyms[i])) * a = list(rule.alignments()) * return TRule(lhs, f, e, scores, a) # <<<<<<<<<<<<<< @@ -5884,7 +5901,7 @@ static int __pyx_pw_5_cdec_5TRule_1__init__(PyObject *__pyx_v_self, PyObject *__ static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__lhs,&__pyx_n_s__f,&__pyx_n_s__e,&__pyx_n_s__scores,&__pyx_n_s__a,0}; PyObject* values[5] = {0,0,0,0,0}; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":52 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":52 * cdef shared_ptr[grammar.TRule]* rule * * def __init__(self, lhs, f, e, scores, a=None): # <<<<<<<<<<<<<< @@ -5973,17 +5990,22 @@ static int __pyx_pf_5_cdec_5TRule___init__(struct __pyx_obj_5_cdec_TRule *__pyx_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":59 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":59 * scores: dictionary of feature scores * a: optional list of alignment points""" * self.rule = new shared_ptr[grammar.TRule](new grammar.TRule()) # <<<<<<<<<<<<<< * self.lhs = lhs * self.e = e */ - try {__pyx_t_1 = new TRule();} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[2]; __pyx_lineno = 59; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_1 = new TRule(); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[2]; __pyx_lineno = 59; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_v_self->rule = new boost::shared_ptr<TRule>(__pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":60 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":60 * a: optional list of alignment points""" * self.rule = new shared_ptr[grammar.TRule](new grammar.TRule()) * self.lhs = lhs # <<<<<<<<<<<<<< @@ -5992,7 +6014,7 @@ static int __pyx_pf_5_cdec_5TRule___init__(struct __pyx_obj_5_cdec_TRule *__pyx_ */ if (PyObject_SetAttr(((PyObject *)__pyx_v_self), __pyx_n_s__lhs, __pyx_v_lhs) < 0) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 60; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":61 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":61 * self.rule = new shared_ptr[grammar.TRule](new grammar.TRule()) * self.lhs = lhs * self.e = e # <<<<<<<<<<<<<< @@ -6001,7 +6023,7 @@ static int __pyx_pf_5_cdec_5TRule___init__(struct __pyx_obj_5_cdec_TRule *__pyx_ */ if (PyObject_SetAttr(((PyObject *)__pyx_v_self), __pyx_n_s__e, __pyx_v_e) < 0) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 61; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":62 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":62 * self.lhs = lhs * self.e = e * self.f = f # <<<<<<<<<<<<<< @@ -6010,7 +6032,7 @@ static int __pyx_pf_5_cdec_5TRule___init__(struct __pyx_obj_5_cdec_TRule *__pyx_ */ if (PyObject_SetAttr(((PyObject *)__pyx_v_self), __pyx_n_s__f, __pyx_v_f) < 0) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":63 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":63 * self.e = e * self.f = f * self.scores = scores # <<<<<<<<<<<<<< @@ -6019,7 +6041,7 @@ static int __pyx_pf_5_cdec_5TRule___init__(struct __pyx_obj_5_cdec_TRule *__pyx_ */ if (PyObject_SetAttr(((PyObject *)__pyx_v_self), __pyx_n_s__scores, __pyx_v_scores) < 0) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":64 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":64 * self.f = f * self.scores = scores * if a: # <<<<<<<<<<<<<< @@ -6029,7 +6051,7 @@ static int __pyx_pf_5_cdec_5TRule___init__(struct __pyx_obj_5_cdec_TRule *__pyx_ __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_v_a); if (unlikely(__pyx_t_2 < 0)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 64; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (__pyx_t_2) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":65 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":65 * self.scores = scores * if a: * self.a = a # <<<<<<<<<<<<<< @@ -6041,7 +6063,7 @@ static int __pyx_pf_5_cdec_5TRule___init__(struct __pyx_obj_5_cdec_TRule *__pyx_ } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":66 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":66 * if a: * self.a = a * self.rule.get().ComputeArity() # <<<<<<<<<<<<<< @@ -6069,7 +6091,7 @@ static void __pyx_pw_5_cdec_5TRule_3__dealloc__(PyObject *__pyx_v_self) { __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":68 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":68 * self.rule.get().ComputeArity() * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -6081,7 +6103,7 @@ static void __pyx_pf_5_cdec_5TRule_2__dealloc__(CYTHON_UNUSED struct __pyx_obj_5 __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":69 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":69 * * def __dealloc__(self): * del self.rule # <<<<<<<<<<<<<< @@ -6104,7 +6126,7 @@ static PyObject *__pyx_pw_5_cdec_5TRule_5arity_1__get__(PyObject *__pyx_v_self) return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":72 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":72 * * property arity: * def __get__(self): # <<<<<<<<<<<<<< @@ -6121,7 +6143,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_5arity___get__(struct __pyx_obj_5_cdec_T int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":73 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":73 * property arity: * def __get__(self): * return self.rule.get().arity_ # <<<<<<<<<<<<<< @@ -6158,7 +6180,7 @@ static PyObject *__pyx_pw_5_cdec_5TRule_1f_1__get__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":76 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":76 * * property f: * def __get__(self): # <<<<<<<<<<<<<< @@ -6185,7 +6207,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":77 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":77 * property f: * def __get__(self): * cdef vector[WordID]* f_ = &self.rule.get().f_ # <<<<<<<<<<<<<< @@ -6194,7 +6216,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_f_ = (&__pyx_v_self->rule->get()->f_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":79 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":79 * cdef vector[WordID]* f_ = &self.rule.get().f_ * cdef WordID w * cdef f = [] # <<<<<<<<<<<<<< @@ -6206,7 +6228,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule __pyx_v_f = ((PyObject *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":81 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":81 * cdef f = [] * cdef unsigned i * cdef int idx = 0 # <<<<<<<<<<<<<< @@ -6215,7 +6237,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_idx = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":82 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":82 * cdef unsigned i * cdef int idx = 0 * for i in range(f_.size()): # <<<<<<<<<<<<<< @@ -6226,7 +6248,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_2; __pyx_t_3+=1) { __pyx_v_i = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":83 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":83 * cdef int idx = 0 * for i in range(f_.size()): * w = f_[0][i] # <<<<<<<<<<<<<< @@ -6235,7 +6257,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_w = ((__pyx_v_f_[0])[__pyx_v_i]); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":84 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":84 * for i in range(f_.size()): * w = f_[0][i] * if w < 0: # <<<<<<<<<<<<<< @@ -6245,7 +6267,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule __pyx_t_4 = (__pyx_v_w < 0); if (__pyx_t_4) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":85 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":85 * w = f_[0][i] * if w < 0: * idx += 1 # <<<<<<<<<<<<<< @@ -6254,7 +6276,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_idx = (__pyx_v_idx + 1); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":86 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":86 * if w < 0: * idx += 1 * f.append(NT(TDConvert(-w).c_str(), idx)) # <<<<<<<<<<<<<< @@ -6284,7 +6306,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":88 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":88 * f.append(NT(TDConvert(-w).c_str(), idx)) * else: * f.append(unicode(TDConvert(w).c_str(), encoding='utf8')) # <<<<<<<<<<<<<< @@ -6313,7 +6335,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1f___get__(struct __pyx_obj_5_cdec_TRule __pyx_L5:; } - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":89 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":89 * else: * f.append(unicode(TDConvert(w).c_str(), encoding='utf8')) * return f # <<<<<<<<<<<<<< @@ -6351,7 +6373,7 @@ static int __pyx_pw_5_cdec_5TRule_1f_3__set__(PyObject *__pyx_v_self, PyObject * return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":91 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":91 * return f * * def __set__(self, f): # <<<<<<<<<<<<<< @@ -6369,15 +6391,15 @@ static int __pyx_pf_5_cdec_5TRule_1f_2__set__(struct __pyx_obj_5_cdec_TRule *__p Py_ssize_t __pyx_t_1; unsigned int __pyx_t_2; PyObject *__pyx_t_3 = NULL; - PyObject *__pyx_t_4 = NULL; - int __pyx_t_5; - char *__pyx_t_6; + int __pyx_t_4; + char *__pyx_t_5; + PyObject *__pyx_t_6 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__set__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":92 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":92 * * def __set__(self, f): * cdef vector[WordID]* f_ = &self.rule.get().f_ # <<<<<<<<<<<<<< @@ -6386,7 +6408,7 @@ static int __pyx_pf_5_cdec_5TRule_1f_2__set__(struct __pyx_obj_5_cdec_TRule *__p */ __pyx_v_f_ = (&__pyx_v_self->rule->get()->f_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":93 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":93 * def __set__(self, f): * cdef vector[WordID]* f_ = &self.rule.get().f_ * f_.resize(len(f)) # <<<<<<<<<<<<<< @@ -6396,7 +6418,7 @@ static int __pyx_pf_5_cdec_5TRule_1f_2__set__(struct __pyx_obj_5_cdec_TRule *__p __pyx_t_1 = PyObject_Length(__pyx_v_f); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_f_->resize(__pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":95 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":95 * f_.resize(len(f)) * cdef unsigned i * cdef int idx = 0 # <<<<<<<<<<<<<< @@ -6405,7 +6427,7 @@ static int __pyx_pf_5_cdec_5TRule_1f_2__set__(struct __pyx_obj_5_cdec_TRule *__p */ __pyx_v_idx = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":96 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":96 * cdef unsigned i * cdef int idx = 0 * for i in range(len(f)): # <<<<<<<<<<<<<< @@ -6416,7 +6438,7 @@ static int __pyx_pf_5_cdec_5TRule_1f_2__set__(struct __pyx_obj_5_cdec_TRule *__p for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":97 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":97 * cdef int idx = 0 * for i in range(len(f)): * if isinstance(f[i], NT): # <<<<<<<<<<<<<< @@ -6425,54 +6447,51 @@ static int __pyx_pf_5_cdec_5TRule_1f_2__set__(struct __pyx_obj_5_cdec_TRule *__p */ __pyx_t_3 = __Pyx_GetItemInt(__pyx_v_f, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_3) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 97; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __pyx_t_4 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NT)); - __Pyx_INCREF(__pyx_t_4); - __pyx_t_5 = __Pyx_TypeCheck(__pyx_t_3, __pyx_t_4); + __pyx_t_4 = __Pyx_TypeCheck(__pyx_t_3, ((PyObject*)__pyx_ptype_5_cdec_NT)); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; - if (__pyx_t_5) { + if (__pyx_t_4) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":98 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":98 * for i in range(len(f)): * if isinstance(f[i], NT): * f_[0][i] = -TDConvert((<NT> f[i]).cat) # <<<<<<<<<<<<<< * else: * fi = as_str(f[i]) */ - __pyx_t_4 = __Pyx_GetItemInt(__pyx_v_f, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_4) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_4); - __pyx_t_6 = PyBytes_AsString(((PyObject *)((struct __pyx_obj_5_cdec_NT *)__pyx_t_4)->cat)); if (unlikely((!__pyx_t_6) && PyErr_Occurred())) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; - ((__pyx_v_f_[0])[__pyx_v_i]) = (-TD::Convert(__pyx_t_6)); + __pyx_t_3 = __Pyx_GetItemInt(__pyx_v_f, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_3) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); + __pyx_t_5 = PyBytes_AsString(((PyObject *)((struct __pyx_obj_5_cdec_NT *)__pyx_t_3)->cat)); if (unlikely((!__pyx_t_5) && PyErr_Occurred())) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 98; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; + ((__pyx_v_f_[0])[__pyx_v_i]) = (-TD::Convert(__pyx_t_5)); goto __pyx_L5; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":100 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":100 * f_[0][i] = -TDConvert((<NT> f[i]).cat) * else: * fi = as_str(f[i]) # <<<<<<<<<<<<<< * f_[0][i] = TDConvert(fi) * */ - __pyx_t_4 = __Pyx_GetItemInt(__pyx_v_f, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_4) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_4); - __pyx_t_3 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_4, NULL)); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = __Pyx_GetItemInt(__pyx_v_f, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_3) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; + __pyx_t_6 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_3, NULL)); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 100; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_6); + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_XDECREF(((PyObject *)__pyx_v_fi)); - __pyx_v_fi = ((PyObject*)__pyx_t_3); - __pyx_t_3 = 0; + __pyx_v_fi = ((PyObject*)__pyx_t_6); + __pyx_t_6 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":101 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":101 * else: * fi = as_str(f[i]) * f_[0][i] = TDConvert(fi) # <<<<<<<<<<<<<< * * property e: */ - __pyx_t_6 = PyBytes_AsString(((PyObject *)__pyx_v_fi)); if (unlikely((!__pyx_t_6) && PyErr_Occurred())) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - ((__pyx_v_f_[0])[__pyx_v_i]) = TD::Convert(__pyx_t_6); + __pyx_t_5 = PyBytes_AsString(((PyObject *)__pyx_v_fi)); if (unlikely((!__pyx_t_5) && PyErr_Occurred())) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + ((__pyx_v_f_[0])[__pyx_v_i]) = TD::Convert(__pyx_t_5); } __pyx_L5:; } @@ -6481,7 +6500,7 @@ static int __pyx_pf_5_cdec_5TRule_1f_2__set__(struct __pyx_obj_5_cdec_TRule *__p goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_3); - __Pyx_XDECREF(__pyx_t_4); + __Pyx_XDECREF(__pyx_t_6); __Pyx_AddTraceback("_cdec.TRule.f.__set__", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = -1; __pyx_L0:; @@ -6501,7 +6520,7 @@ static PyObject *__pyx_pw_5_cdec_5TRule_1e_1__get__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":104 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":104 * * property e: * def __get__(self): # <<<<<<<<<<<<<< @@ -6528,7 +6547,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":105 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":105 * property e: * def __get__(self): * cdef vector[WordID]* e_ = &self.rule.get().e_ # <<<<<<<<<<<<<< @@ -6537,7 +6556,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_e_ = (&__pyx_v_self->rule->get()->e_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":107 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":107 * cdef vector[WordID]* e_ = &self.rule.get().e_ * cdef WordID w * cdef e = [] # <<<<<<<<<<<<<< @@ -6549,7 +6568,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule __pyx_v_e = ((PyObject *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":109 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":109 * cdef e = [] * cdef unsigned i * cdef int idx = 0 # <<<<<<<<<<<<<< @@ -6558,7 +6577,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_idx = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":110 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":110 * cdef unsigned i * cdef int idx = 0 * for i in range(e_.size()): # <<<<<<<<<<<<<< @@ -6569,7 +6588,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_2; __pyx_t_3+=1) { __pyx_v_i = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":111 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":111 * cdef int idx = 0 * for i in range(e_.size()): * w = e_[0][i] # <<<<<<<<<<<<<< @@ -6578,7 +6597,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_w = ((__pyx_v_e_[0])[__pyx_v_i]); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":112 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":112 * for i in range(e_.size()): * w = e_[0][i] * if w < 1: # <<<<<<<<<<<<<< @@ -6588,7 +6607,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule __pyx_t_4 = (__pyx_v_w < 1); if (__pyx_t_4) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":113 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":113 * w = e_[0][i] * if w < 1: * idx += 1 # <<<<<<<<<<<<<< @@ -6597,7 +6616,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_idx = (__pyx_v_idx + 1); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":114 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":114 * if w < 1: * idx += 1 * e.append(NTRef(1-w)) # <<<<<<<<<<<<<< @@ -6622,7 +6641,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":116 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":116 * e.append(NTRef(1-w)) * else: * e.append(unicode(TDConvert(w).c_str(), encoding='utf8')) # <<<<<<<<<<<<<< @@ -6651,7 +6670,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_1e___get__(struct __pyx_obj_5_cdec_TRule __pyx_L5:; } - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":117 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":117 * else: * e.append(unicode(TDConvert(w).c_str(), encoding='utf8')) * return e # <<<<<<<<<<<<<< @@ -6689,7 +6708,7 @@ static int __pyx_pw_5_cdec_5TRule_1e_3__set__(PyObject *__pyx_v_self, PyObject * return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":119 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":119 * return e * * def __set__(self, e): # <<<<<<<<<<<<<< @@ -6706,8 +6725,8 @@ static int __pyx_pf_5_cdec_5TRule_1e_2__set__(struct __pyx_obj_5_cdec_TRule *__p Py_ssize_t __pyx_t_1; unsigned int __pyx_t_2; PyObject *__pyx_t_3 = NULL; - PyObject *__pyx_t_4 = NULL; - int __pyx_t_5; + int __pyx_t_4; + PyObject *__pyx_t_5 = NULL; WordID __pyx_t_6; char *__pyx_t_7; int __pyx_lineno = 0; @@ -6715,7 +6734,7 @@ static int __pyx_pf_5_cdec_5TRule_1e_2__set__(struct __pyx_obj_5_cdec_TRule *__p int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__set__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":120 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":120 * * def __set__(self, e): * cdef vector[WordID]* e_ = &self.rule.get().e_ # <<<<<<<<<<<<<< @@ -6724,7 +6743,7 @@ static int __pyx_pf_5_cdec_5TRule_1e_2__set__(struct __pyx_obj_5_cdec_TRule *__p */ __pyx_v_e_ = (&__pyx_v_self->rule->get()->e_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":121 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":121 * def __set__(self, e): * cdef vector[WordID]* e_ = &self.rule.get().e_ * e_.resize(len(e)) # <<<<<<<<<<<<<< @@ -6734,7 +6753,7 @@ static int __pyx_pf_5_cdec_5TRule_1e_2__set__(struct __pyx_obj_5_cdec_TRule *__p __pyx_t_1 = PyObject_Length(__pyx_v_e); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 121; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_e_->resize(__pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":123 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":123 * e_.resize(len(e)) * cdef unsigned i * for i in range(len(e)): # <<<<<<<<<<<<<< @@ -6745,7 +6764,7 @@ static int __pyx_pf_5_cdec_5TRule_1e_2__set__(struct __pyx_obj_5_cdec_TRule *__p for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":124 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":124 * cdef unsigned i * for i in range(len(e)): * if isinstance(e[i], NTRef): # <<<<<<<<<<<<<< @@ -6754,52 +6773,49 @@ static int __pyx_pf_5_cdec_5TRule_1e_2__set__(struct __pyx_obj_5_cdec_TRule *__p */ __pyx_t_3 = __Pyx_GetItemInt(__pyx_v_e, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_3) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 124; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __pyx_t_4 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NTRef)); - __Pyx_INCREF(__pyx_t_4); - __pyx_t_5 = __Pyx_TypeCheck(__pyx_t_3, __pyx_t_4); + __pyx_t_4 = __Pyx_TypeCheck(__pyx_t_3, ((PyObject*)__pyx_ptype_5_cdec_NTRef)); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; - if (__pyx_t_5) { + if (__pyx_t_4) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":125 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":125 * for i in range(len(e)): * if isinstance(e[i], NTRef): * e_[0][i] = 1-e[i].ref # <<<<<<<<<<<<<< * else: * ei = as_str(e[i]) */ - __pyx_t_4 = __Pyx_GetItemInt(__pyx_v_e, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_4) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_4); - __pyx_t_3 = PyObject_GetAttr(__pyx_t_4, __pyx_n_s__ref); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = __Pyx_GetItemInt(__pyx_v_e, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_3) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; - __pyx_t_4 = PyNumber_Subtract(__pyx_int_1, __pyx_t_3); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_4); + __pyx_t_5 = PyObject_GetAttr(__pyx_t_3, __pyx_n_s__ref); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_5); + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; + __pyx_t_3 = PyNumber_Subtract(__pyx_int_1, __pyx_t_5); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); + __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; + __pyx_t_6 = __Pyx_PyInt_from_py_WordID(__pyx_t_3); if (unlikely((__pyx_t_6 == (WordID)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - __pyx_t_6 = __Pyx_PyInt_from_py_WordID(__pyx_t_4); if (unlikely((__pyx_t_6 == (WordID)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; ((__pyx_v_e_[0])[__pyx_v_i]) = __pyx_t_6; goto __pyx_L5; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":127 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":127 * e_[0][i] = 1-e[i].ref * else: * ei = as_str(e[i]) # <<<<<<<<<<<<<< * e_[0][i] = TDConvert(ei) * */ - __pyx_t_4 = __Pyx_GetItemInt(__pyx_v_e, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_4) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_4); - __pyx_t_3 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_4, NULL)); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = __Pyx_GetItemInt(__pyx_v_e, __pyx_v_i, sizeof(unsigned int)+1, PyLong_FromUnsignedLong); if (!__pyx_t_3) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; + __pyx_t_5 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_3, NULL)); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_5); + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_XDECREF(((PyObject *)__pyx_v_ei)); - __pyx_v_ei = ((PyObject*)__pyx_t_3); - __pyx_t_3 = 0; + __pyx_v_ei = ((PyObject*)__pyx_t_5); + __pyx_t_5 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":128 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":128 * else: * ei = as_str(e[i]) * e_[0][i] = TDConvert(ei) # <<<<<<<<<<<<<< @@ -6816,7 +6832,7 @@ static int __pyx_pf_5_cdec_5TRule_1e_2__set__(struct __pyx_obj_5_cdec_TRule *__p goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_3); - __Pyx_XDECREF(__pyx_t_4); + __Pyx_XDECREF(__pyx_t_5); __Pyx_AddTraceback("_cdec.TRule.e.__set__", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = -1; __pyx_L0:; @@ -6837,7 +6853,7 @@ static PyObject *__pyx_pw_5_cdec_5TRule_1a_1__get__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":131 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":131 * * property a: * def __get__(self): # <<<<<<<<<<<<<< @@ -6902,7 +6918,7 @@ static PyObject *__pyx_gb_5_cdec_5TRule_1a_2generator2(__pyx_GeneratorObject *__ __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 131; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":133 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":133 * def __get__(self): * cdef unsigned i * cdef vector[grammar.AlignmentPoint]* a = &self.rule.get().a_ # <<<<<<<<<<<<<< @@ -6911,7 +6927,7 @@ static PyObject *__pyx_gb_5_cdec_5TRule_1a_2generator2(__pyx_GeneratorObject *__ */ __pyx_cur_scope->__pyx_v_a = (&__pyx_cur_scope->__pyx_v_self->rule->get()->a_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":134 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":134 * cdef unsigned i * cdef vector[grammar.AlignmentPoint]* a = &self.rule.get().a_ * for i in range(a.size()): # <<<<<<<<<<<<<< @@ -6922,7 +6938,7 @@ static PyObject *__pyx_gb_5_cdec_5TRule_1a_2generator2(__pyx_GeneratorObject *__ for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":135 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":135 * cdef vector[grammar.AlignmentPoint]* a = &self.rule.get().a_ * for i in range(a.size()): * yield (a[0][i].s_, a[0][i].t_) # <<<<<<<<<<<<<< @@ -6981,7 +6997,7 @@ static int __pyx_pw_5_cdec_5TRule_1a_4__set__(PyObject *__pyx_v_self, PyObject * return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":137 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":137 * yield (a[0][i].s_, a[0][i].t_) * * def __set__(self, a): # <<<<<<<<<<<<<< @@ -7010,7 +7026,7 @@ static int __pyx_pf_5_cdec_5TRule_1a_3__set__(struct __pyx_obj_5_cdec_TRule *__p int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__set__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":138 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":138 * * def __set__(self, a): * cdef vector[grammar.AlignmentPoint]* a_ = &self.rule.get().a_ # <<<<<<<<<<<<<< @@ -7019,7 +7035,7 @@ static int __pyx_pf_5_cdec_5TRule_1a_3__set__(struct __pyx_obj_5_cdec_TRule *__p */ __pyx_v_a_ = (&__pyx_v_self->rule->get()->a_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":139 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":139 * def __set__(self, a): * cdef vector[grammar.AlignmentPoint]* a_ = &self.rule.get().a_ * a_.resize(len(a)) # <<<<<<<<<<<<<< @@ -7029,7 +7045,7 @@ static int __pyx_pf_5_cdec_5TRule_1a_3__set__(struct __pyx_obj_5_cdec_TRule *__p __pyx_t_1 = PyObject_Length(__pyx_v_a); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_a_->resize(__pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":142 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":142 * cdef unsigned i * cdef int s, t * for i in range(len(a)): # <<<<<<<<<<<<<< @@ -7040,7 +7056,7 @@ static int __pyx_pf_5_cdec_5TRule_1a_3__set__(struct __pyx_obj_5_cdec_TRule *__p for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":143 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":143 * cdef int s, t * for i in range(len(a)): * s, t = a[i] # <<<<<<<<<<<<<< @@ -7073,7 +7089,9 @@ static int __pyx_pf_5_cdec_5TRule_1a_3__set__(struct __pyx_obj_5_cdec_TRule *__p __Pyx_INCREF(__pyx_t_5); #else __pyx_t_4 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 143; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_4); __pyx_t_5 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 143; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_5); #endif __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; } else @@ -7105,7 +7123,7 @@ static int __pyx_pf_5_cdec_5TRule_1a_3__set__(struct __pyx_obj_5_cdec_TRule *__p __pyx_v_s = __pyx_t_8; __pyx_v_t = __pyx_t_9; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":144 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":144 * for i in range(len(a)): * s, t = a[i] * a_[0][i] = grammar.AlignmentPoint(s, t) # <<<<<<<<<<<<<< @@ -7140,7 +7158,7 @@ static PyObject *__pyx_pw_5_cdec_5TRule_6scores_1__get__(PyObject *__pyx_v_self) return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":147 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":147 * * property scores: * def __get__(self): # <<<<<<<<<<<<<< @@ -7158,7 +7176,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_6scores___get__(struct __pyx_obj_5_cdec_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":148 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":148 * property scores: * def __get__(self): * cdef SparseVector scores = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -7171,7 +7189,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_6scores___get__(struct __pyx_obj_5_cdec_ __pyx_v_scores = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":149 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":149 * def __get__(self): * cdef SparseVector scores = SparseVector.__new__(SparseVector) * scores.vector = new FastSparseVector[double](self.rule.get().scores_) # <<<<<<<<<<<<<< @@ -7180,7 +7198,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_6scores___get__(struct __pyx_obj_5_cdec_ */ __pyx_v_scores->vector = new FastSparseVector<double>(__pyx_v_self->rule->get()->scores_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":150 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":150 * cdef SparseVector scores = SparseVector.__new__(SparseVector) * scores.vector = new FastSparseVector[double](self.rule.get().scores_) * return scores # <<<<<<<<<<<<<< @@ -7216,7 +7234,7 @@ static int __pyx_pw_5_cdec_5TRule_6scores_3__set__(PyObject *__pyx_v_self, PyObj return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":152 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":152 * return scores * * def __set__(self, scores): # <<<<<<<<<<<<<< @@ -7248,7 +7266,7 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__set__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":153 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":153 * * def __set__(self, scores): * cdef FastSparseVector[double]* scores_ = &self.rule.get().scores_ # <<<<<<<<<<<<<< @@ -7257,7 +7275,7 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_scores_ = (&__pyx_v_self->rule->get()->scores_); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":154 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":154 * def __set__(self, scores): * cdef FastSparseVector[double]* scores_ = &self.rule.get().scores_ * scores_.clear() # <<<<<<<<<<<<<< @@ -7266,7 +7284,7 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule */ __pyx_v_scores_->clear(); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":157 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":157 * cdef int fid * cdef float fval * for fname, fval in scores.items(): # <<<<<<<<<<<<<< @@ -7337,7 +7355,9 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule __Pyx_INCREF(__pyx_t_6); #else __pyx_t_5 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 157; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_5); __pyx_t_6 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 157; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_6); #endif __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; } else @@ -7369,7 +7389,7 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule __pyx_t_5 = 0; __pyx_v_fval = __pyx_t_9; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":158 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":158 * cdef float fval * for fname, fval in scores.items(): * fn = as_str(fname) # <<<<<<<<<<<<<< @@ -7382,7 +7402,7 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule __pyx_v_fn = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":159 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":159 * for fname, fval in scores.items(): * fn = as_str(fname) * fid = FDConvert(fn) # <<<<<<<<<<<<<< @@ -7392,7 +7412,7 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule __pyx_t_10 = PyBytes_AsString(((PyObject *)__pyx_v_fn)); if (unlikely((!__pyx_t_10) && PyErr_Occurred())) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 159; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_fid = FD::Convert(__pyx_t_10); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":160 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":160 * fn = as_str(fname) * fid = FDConvert(fn) * if fid < 0: raise KeyError(fname) # <<<<<<<<<<<<<< @@ -7416,7 +7436,7 @@ static int __pyx_pf_5_cdec_5TRule_6scores_2__set__(struct __pyx_obj_5_cdec_TRule } __pyx_L7:; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":161 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":161 * fid = FDConvert(fn) * if fid < 0: raise KeyError(fname) * scores_.set_value(fid, fval) # <<<<<<<<<<<<<< @@ -7455,7 +7475,7 @@ static PyObject *__pyx_pw_5_cdec_5TRule_3lhs_1__get__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":164 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":164 * * property lhs: * def __get__(self): # <<<<<<<<<<<<<< @@ -7473,7 +7493,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_3lhs___get__(struct __pyx_obj_5_cdec_TRu int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":165 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":165 * property lhs: * def __get__(self): * return NT(TDConvert(-self.rule.get().lhs_).c_str()) # <<<<<<<<<<<<<< @@ -7519,7 +7539,7 @@ static int __pyx_pw_5_cdec_5TRule_3lhs_3__set__(PyObject *__pyx_v_self, PyObject return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":167 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":167 * return NT(TDConvert(-self.rule.get().lhs_).c_str()) * * def __set__(self, lhs): # <<<<<<<<<<<<<< @@ -7530,9 +7550,9 @@ static int __pyx_pw_5_cdec_5TRule_3lhs_3__set__(PyObject *__pyx_v_self, PyObject static int __pyx_pf_5_cdec_5TRule_3lhs_2__set__(struct __pyx_obj_5_cdec_TRule *__pyx_v_self, PyObject *__pyx_v_lhs) { int __pyx_r; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; + int __pyx_t_1; int __pyx_t_2; - int __pyx_t_3; + PyObject *__pyx_t_3 = NULL; PyObject *__pyx_t_4 = NULL; char *__pyx_t_5; int __pyx_lineno = 0; @@ -7541,35 +7561,32 @@ static int __pyx_pf_5_cdec_5TRule_3lhs_2__set__(struct __pyx_obj_5_cdec_TRule *_ __Pyx_RefNannySetupContext("__set__", 0); __Pyx_INCREF(__pyx_v_lhs); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":168 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":168 * * def __set__(self, lhs): * if not isinstance(lhs, NT): # <<<<<<<<<<<<<< * lhs = NT(lhs) * self.rule.get().lhs_ = -TDConvert((<NT> lhs).cat) */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NT)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_lhs, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - __pyx_t_3 = (!__pyx_t_2); - if (__pyx_t_3) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_lhs, ((PyObject*)__pyx_ptype_5_cdec_NT)); + __pyx_t_2 = (!__pyx_t_1); + if (__pyx_t_2) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":169 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":169 * def __set__(self, lhs): * if not isinstance(lhs, NT): * lhs = NT(lhs) # <<<<<<<<<<<<<< * self.rule.get().lhs_ = -TDConvert((<NT> lhs).cat) * */ - __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 169; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 169; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); __Pyx_INCREF(__pyx_v_lhs); - PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_v_lhs); + PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_v_lhs); __Pyx_GIVEREF(__pyx_v_lhs); - __pyx_t_4 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NT)), ((PyObject *)__pyx_t_1), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 169; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_4 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NT)), ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 169; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); - __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; + __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; __Pyx_DECREF(__pyx_v_lhs); __pyx_v_lhs = __pyx_t_4; __pyx_t_4 = 0; @@ -7577,7 +7594,7 @@ static int __pyx_pf_5_cdec_5TRule_3lhs_2__set__(struct __pyx_obj_5_cdec_TRule *_ } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":170 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":170 * if not isinstance(lhs, NT): * lhs = NT(lhs) * self.rule.get().lhs_ = -TDConvert((<NT> lhs).cat) # <<<<<<<<<<<<<< @@ -7590,7 +7607,7 @@ static int __pyx_pf_5_cdec_5TRule_3lhs_2__set__(struct __pyx_obj_5_cdec_TRule *_ __pyx_r = 0; goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_4); __Pyx_AddTraceback("_cdec.TRule.lhs.__set__", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = -1; @@ -7612,7 +7629,7 @@ static PyObject *__pyx_pw_5_cdec_5TRule_5__str__(PyObject *__pyx_v_self) { } static PyObject *__pyx_gb_5_cdec_5TRule_7__str___2generator19(__pyx_GeneratorObject *__pyx_generator, PyObject *__pyx_sent_value); /* proto */ -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":173 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":173 * * def __str__(self): * scores = ' '.join('%s=%s' % feat for feat in self.scores) # <<<<<<<<<<<<<< @@ -7754,7 +7771,7 @@ static PyObject *__pyx_gb_5_cdec_5TRule_7__str___2generator19(__pyx_GeneratorObj return NULL; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":172 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":172 * self.rule.get().lhs_ = -TDConvert((<NT> lhs).cat) * * def __str__(self): # <<<<<<<<<<<<<< @@ -7786,7 +7803,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_4__str__(struct __pyx_obj_5_cdec_TRule * __Pyx_INCREF((PyObject *)__pyx_cur_scope->__pyx_v_self); __Pyx_GIVEREF((PyObject *)__pyx_cur_scope->__pyx_v_self); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":173 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":173 * * def __str__(self): * scores = ' '.join('%s=%s' % feat for feat in self.scores) # <<<<<<<<<<<<<< @@ -7809,7 +7826,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_4__str__(struct __pyx_obj_5_cdec_TRule * __pyx_v_scores = __pyx_t_2; __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":174 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":174 * def __str__(self): * scores = ' '.join('%s=%s' % feat for feat in self.scores) * return '%s ||| %s ||| %s ||| %s' % (self.lhs, # <<<<<<<<<<<<<< @@ -7820,7 +7837,7 @@ static PyObject *__pyx_pf_5_cdec_5TRule_4__str__(struct __pyx_obj_5_cdec_TRule * __pyx_t_2 = PyObject_GetAttr(((PyObject *)__pyx_cur_scope->__pyx_v_self), __pyx_n_s__lhs); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 174; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":175 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":175 * scores = ' '.join('%s=%s' % feat for feat in self.scores) * return '%s ||| %s ||| %s ||| %s' % (self.lhs, * _phrase(self.f), _phrase(self.e), scores) # <<<<<<<<<<<<<< @@ -7961,7 +7978,7 @@ static int __pyx_pw_5_cdec_5MRule_1__init__(PyObject *__pyx_v_self, PyObject *__ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":178 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":178 * * cdef class MRule(TRule): * def __init__(self, lhs, rhs, scores): # <<<<<<<<<<<<<< @@ -7987,7 +8004,7 @@ static int __pyx_pf_5_cdec_5MRule___init__(struct __pyx_obj_5_cdec_MRule *__pyx_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":183 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":183 * rhs: right hand side phrase (list of words/NT) * scores: dictionary of feature scores""" * cdef unsigned i = 1 # <<<<<<<<<<<<<< @@ -7996,7 +8013,7 @@ static int __pyx_pf_5_cdec_5MRule___init__(struct __pyx_obj_5_cdec_MRule *__pyx_ */ __pyx_v_i = 1; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":184 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":184 * scores: dictionary of feature scores""" * cdef unsigned i = 1 * e = [] # <<<<<<<<<<<<<< @@ -8005,10 +8022,10 @@ static int __pyx_pf_5_cdec_5MRule___init__(struct __pyx_obj_5_cdec_MRule *__pyx_ */ __pyx_t_1 = PyList_New(0); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 184; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); - __pyx_v_e = __pyx_t_1; + __pyx_v_e = ((PyObject*)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":185 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":185 * cdef unsigned i = 1 * e = [] * for s in rhs: # <<<<<<<<<<<<<< @@ -8053,20 +8070,17 @@ static int __pyx_pf_5_cdec_5MRule___init__(struct __pyx_obj_5_cdec_MRule *__pyx_ __pyx_v_s = __pyx_t_4; __pyx_t_4 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":186 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":186 * e = [] * for s in rhs: * if isinstance(s, NT): # <<<<<<<<<<<<<< * e.append(NTRef(i)) * i += 1 */ - __pyx_t_4 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_NT)); - __Pyx_INCREF(__pyx_t_4); - __pyx_t_5 = __Pyx_TypeCheck(__pyx_v_s, __pyx_t_4); - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; + __pyx_t_5 = __Pyx_TypeCheck(__pyx_v_s, ((PyObject*)__pyx_ptype_5_cdec_NT)); if (__pyx_t_5) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":187 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":187 * for s in rhs: * if isinstance(s, NT): * e.append(NTRef(i)) # <<<<<<<<<<<<<< @@ -8086,7 +8100,7 @@ static int __pyx_pf_5_cdec_5MRule___init__(struct __pyx_obj_5_cdec_MRule *__pyx_ __pyx_t_7 = PyList_Append(__pyx_v_e, __pyx_t_4); if (unlikely(__pyx_t_7 == -1)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 187; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":188 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":188 * if isinstance(s, NT): * e.append(NTRef(i)) * i += 1 # <<<<<<<<<<<<<< @@ -8098,7 +8112,7 @@ static int __pyx_pf_5_cdec_5MRule___init__(struct __pyx_obj_5_cdec_MRule *__pyx_ } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":190 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":190 * i += 1 * else: * e.append(s) # <<<<<<<<<<<<<< @@ -8111,7 +8125,7 @@ static int __pyx_pf_5_cdec_5MRule___init__(struct __pyx_obj_5_cdec_MRule *__pyx_ } __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":191 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":191 * else: * e.append(s) * super(MRule, self).__init__(lhs, rhs, e, scores, None) # <<<<<<<<<<<<<< @@ -8179,7 +8193,7 @@ static void __pyx_pw_5_cdec_7Grammar_1__dealloc__(PyObject *__pyx_v_self) { __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":196 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":196 * cdef shared_ptr[grammar.Grammar]* grammar * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -8191,7 +8205,7 @@ static void __pyx_pf_5_cdec_7Grammar___dealloc__(CYTHON_UNUSED struct __pyx_obj_ __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":197 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":197 * * def __dealloc__(self): * del self.grammar # <<<<<<<<<<<<<< @@ -8215,7 +8229,7 @@ static PyObject *__pyx_pw_5_cdec_7Grammar_3__iter__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":199 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":199 * del self.grammar * * def __iter__(self): # <<<<<<<<<<<<<< @@ -8278,7 +8292,7 @@ static PyObject *__pyx_gb_5_cdec_7Grammar_4generator3(__pyx_GeneratorObject *__p __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 199; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":200 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":200 * * def __iter__(self): * cdef grammar.const_GrammarIter* root = self.grammar.get().GetRoot() # <<<<<<<<<<<<<< @@ -8287,7 +8301,7 @@ static PyObject *__pyx_gb_5_cdec_7Grammar_4generator3(__pyx_GeneratorObject *__p */ __pyx_cur_scope->__pyx_v_root = __pyx_cur_scope->__pyx_v_self->grammar->get()->GetRoot(); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":201 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":201 * def __iter__(self): * cdef grammar.const_GrammarIter* root = self.grammar.get().GetRoot() * cdef grammar.const_RuleBin* rbin = root.GetRules() # <<<<<<<<<<<<<< @@ -8296,7 +8310,7 @@ static PyObject *__pyx_gb_5_cdec_7Grammar_4generator3(__pyx_GeneratorObject *__p */ __pyx_cur_scope->__pyx_v_rbin = __pyx_cur_scope->__pyx_v_root->GetRules(); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":204 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":204 * cdef TRule trule * cdef unsigned i * for i in range(rbin.GetNumRules()): # <<<<<<<<<<<<<< @@ -8307,7 +8321,7 @@ static PyObject *__pyx_gb_5_cdec_7Grammar_4generator3(__pyx_GeneratorObject *__p for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":205 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":205 * cdef unsigned i * for i in range(rbin.GetNumRules()): * trule = TRule.__new__(TRule) # <<<<<<<<<<<<<< @@ -8323,7 +8337,7 @@ static PyObject *__pyx_gb_5_cdec_7Grammar_4generator3(__pyx_GeneratorObject *__p __pyx_cur_scope->__pyx_v_trule = ((struct __pyx_obj_5_cdec_TRule *)__pyx_t_3); __pyx_t_3 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":206 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":206 * for i in range(rbin.GetNumRules()): * trule = TRule.__new__(TRule) * trule.rule = new shared_ptr[grammar.TRule](rbin.GetIthRule(i)) # <<<<<<<<<<<<<< @@ -8332,7 +8346,7 @@ static PyObject *__pyx_gb_5_cdec_7Grammar_4generator3(__pyx_GeneratorObject *__p */ __pyx_cur_scope->__pyx_v_trule->rule = new boost::shared_ptr<TRule>(__pyx_cur_scope->__pyx_v_rbin->GetIthRule(__pyx_cur_scope->__pyx_v_i)); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":207 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":207 * trule = TRule.__new__(TRule) * trule.rule = new shared_ptr[grammar.TRule](rbin.GetIthRule(i)) * yield trule # <<<<<<<<<<<<<< @@ -8377,7 +8391,7 @@ static PyObject *__pyx_pw_5_cdec_7Grammar_4name_1__get__(PyObject *__pyx_v_self) return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":210 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":210 * * property name: * def __get__(self): # <<<<<<<<<<<<<< @@ -8395,7 +8409,7 @@ static PyObject *__pyx_pf_5_cdec_7Grammar_4name___get__(struct __pyx_obj_5_cdec_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":211 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":211 * property name: * def __get__(self): * str(self.grammar.get().GetGrammarName().c_str()) # <<<<<<<<<<<<<< @@ -8438,7 +8452,7 @@ static int __pyx_pw_5_cdec_7Grammar_4name_3__set__(PyObject *__pyx_v_self, PyObj return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":213 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":213 * str(self.grammar.get().GetGrammarName().c_str()) * * def __set__(self, name): # <<<<<<<<<<<<<< @@ -8457,7 +8471,7 @@ static int __pyx_pf_5_cdec_7Grammar_4name_2__set__(struct __pyx_obj_5_cdec_Gramm __Pyx_RefNannySetupContext("__set__", 0); __Pyx_INCREF(__pyx_v_name); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":214 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":214 * * def __set__(self, name): * name = as_str(name) # <<<<<<<<<<<<<< @@ -8470,7 +8484,7 @@ static int __pyx_pf_5_cdec_7Grammar_4name_2__set__(struct __pyx_obj_5_cdec_Gramm __pyx_v_name = __pyx_t_1; __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":215 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":215 * def __set__(self, name): * name = as_str(name) * self.grammar.get().SetGrammarName(name) # <<<<<<<<<<<<<< @@ -8543,7 +8557,7 @@ static int __pyx_pw_5_cdec_11TextGrammar_1__init__(PyObject *__pyx_v_self, PyObj return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":218 +/* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":218 * * cdef class TextGrammar(Grammar): * def __init__(self, rules): # <<<<<<<<<<<<<< @@ -8561,14 +8575,13 @@ static int __pyx_pf_5_cdec_11TextGrammar___init__(struct __pyx_obj_5_cdec_TextGr PyObject *(*__pyx_t_3)(PyObject *); PyObject *__pyx_t_4 = NULL; int __pyx_t_5; - PyObject *__pyx_t_6 = NULL; - int __pyx_t_7; + int __pyx_t_6; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":220 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":220 * def __init__(self, rules): * """TextGrammar(rules) -> SCFG Grammar containing the rules.""" * self.grammar = new shared_ptr[grammar.Grammar](new grammar.TextGrammar()) # <<<<<<<<<<<<<< @@ -8577,7 +8590,7 @@ static int __pyx_pf_5_cdec_11TextGrammar___init__(struct __pyx_obj_5_cdec_TextGr */ __pyx_v_self->__pyx_base.grammar = new boost::shared_ptr<Grammar>(new TextGrammar()); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":221 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":221 * """TextGrammar(rules) -> SCFG Grammar containing the rules.""" * self.grammar = new shared_ptr[grammar.Grammar](new grammar.TextGrammar()) * cdef grammar.TextGrammar* _g = <grammar.TextGrammar*> self.grammar.get() # <<<<<<<<<<<<<< @@ -8586,7 +8599,7 @@ static int __pyx_pf_5_cdec_11TextGrammar___init__(struct __pyx_obj_5_cdec_TextGr */ __pyx_v__g = ((TextGrammar *)__pyx_v_self->__pyx_base.grammar->get()); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":222 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":222 * self.grammar = new shared_ptr[grammar.Grammar](new grammar.TextGrammar()) * cdef grammar.TextGrammar* _g = <grammar.TextGrammar*> self.grammar.get() * for trule in rules: # <<<<<<<<<<<<<< @@ -8631,20 +8644,17 @@ static int __pyx_pf_5_cdec_11TextGrammar___init__(struct __pyx_obj_5_cdec_TextGr __pyx_v_trule = __pyx_t_4; __pyx_t_4 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":223 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":223 * cdef grammar.TextGrammar* _g = <grammar.TextGrammar*> self.grammar.get() * for trule in rules: * if isinstance(trule, _sa.Rule): # <<<<<<<<<<<<<< * trule = convert_rule(trule) * elif not isinstance(trule, TRule): */ - __pyx_t_4 = ((PyObject *)((PyObject*)__pyx_ptype_4cdec_2sa_3_sa_Rule)); - __Pyx_INCREF(__pyx_t_4); - __pyx_t_5 = __Pyx_TypeCheck(__pyx_v_trule, __pyx_t_4); - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; + __pyx_t_5 = __Pyx_TypeCheck(__pyx_v_trule, ((PyObject*)__pyx_ptype_4cdec_2sa_3_sa_Rule)); if (__pyx_t_5) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":224 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":224 * for trule in rules: * if isinstance(trule, _sa.Rule): * trule = convert_rule(trule) # <<<<<<<<<<<<<< @@ -8652,47 +8662,41 @@ static int __pyx_pf_5_cdec_11TextGrammar___init__(struct __pyx_obj_5_cdec_TextGr * raise ValueError('the grammar should contain TRule objects') */ if (!(likely(((__pyx_v_trule) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_trule, __pyx_ptype_4cdec_2sa_3_sa_Rule))))) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 224; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_4 = __pyx_v_trule; - __Pyx_INCREF(__pyx_t_4); - __pyx_t_6 = ((PyObject *)__pyx_f_5_cdec_convert_rule(((struct __pyx_obj_4cdec_2sa_3_sa_Rule *)__pyx_t_4))); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 224; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_6); - __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; + __pyx_t_4 = ((PyObject *)__pyx_f_5_cdec_convert_rule(((struct __pyx_obj_4cdec_2sa_3_sa_Rule *)__pyx_v_trule))); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 224; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(__pyx_v_trule); - __pyx_v_trule = __pyx_t_6; - __pyx_t_6 = 0; + __pyx_v_trule = __pyx_t_4; + __pyx_t_4 = 0; goto __pyx_L5; } - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":225 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":225 * if isinstance(trule, _sa.Rule): * trule = convert_rule(trule) * elif not isinstance(trule, TRule): # <<<<<<<<<<<<<< * raise ValueError('the grammar should contain TRule objects') * _g.AddRule((<TRule> trule).rule[0]) */ - __pyx_t_6 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_TRule)); - __Pyx_INCREF(__pyx_t_6); - __pyx_t_5 = __Pyx_TypeCheck(__pyx_v_trule, __pyx_t_6); - __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; - __pyx_t_7 = (!__pyx_t_5); - if (__pyx_t_7) { + __pyx_t_5 = __Pyx_TypeCheck(__pyx_v_trule, ((PyObject*)__pyx_ptype_5_cdec_TRule)); + __pyx_t_6 = (!__pyx_t_5); + if (__pyx_t_6) { - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":226 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":226 * trule = convert_rule(trule) * elif not isinstance(trule, TRule): * raise ValueError('the grammar should contain TRule objects') # <<<<<<<<<<<<<< * _g.AddRule((<TRule> trule).rule[0]) */ - __pyx_t_6 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_14), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 226; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_6); - __Pyx_Raise(__pyx_t_6, 0, 0, 0); - __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; + __pyx_t_4 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_14), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 226; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_4); + __Pyx_Raise(__pyx_t_4, 0, 0, 0); + __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; {__pyx_filename = __pyx_f[2]; __pyx_lineno = 226; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L5; } __pyx_L5:; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":227 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":227 * elif not isinstance(trule, TRule): * raise ValueError('the grammar should contain TRule objects') * _g.AddRule((<TRule> trule).rule[0]) # <<<<<<<<<<<<<< @@ -8706,7 +8710,6 @@ static int __pyx_pf_5_cdec_11TextGrammar___init__(struct __pyx_obj_5_cdec_TextGr __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); __Pyx_XDECREF(__pyx_t_4); - __Pyx_XDECREF(__pyx_t_6); __Pyx_AddTraceback("_cdec.TextGrammar.__init__", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = -1; __pyx_L0:; @@ -8724,7 +8727,7 @@ static void __pyx_pw_5_cdec_10Hypergraph_1__dealloc__(PyObject *__pyx_v_self) { __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":8 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":8 * cdef MT19937* rng * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -8737,7 +8740,7 @@ static void __pyx_pf_5_cdec_10Hypergraph___dealloc__(struct __pyx_obj_5_cdec_Hyp int __pyx_t_1; __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":9 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":9 * * def __dealloc__(self): * del self.hg # <<<<<<<<<<<<<< @@ -8746,7 +8749,7 @@ static void __pyx_pf_5_cdec_10Hypergraph___dealloc__(struct __pyx_obj_5_cdec_Hyp */ delete __pyx_v_self->hg; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":10 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":10 * def __dealloc__(self): * del self.hg * if self.rng != NULL: # <<<<<<<<<<<<<< @@ -8756,7 +8759,7 @@ static void __pyx_pf_5_cdec_10Hypergraph___dealloc__(struct __pyx_obj_5_cdec_Hyp __pyx_t_1 = (__pyx_v_self->rng != NULL); if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":11 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":11 * del self.hg * if self.rng != NULL: * del self.rng # <<<<<<<<<<<<<< @@ -8771,7 +8774,7 @@ static void __pyx_pf_5_cdec_10Hypergraph___dealloc__(struct __pyx_obj_5_cdec_Hyp __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":13 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":13 * del self.rng * * cdef MT19937* _rng(self): # <<<<<<<<<<<<<< @@ -8789,7 +8792,7 @@ static MT19937 *__pyx_f_5_cdec_10Hypergraph__rng(struct __pyx_obj_5_cdec_Hypergr int __pyx_clineno = 0; __Pyx_RefNannySetupContext("_rng", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":14 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":14 * * cdef MT19937* _rng(self): * if self.rng == NULL: # <<<<<<<<<<<<<< @@ -8799,20 +8802,25 @@ static MT19937 *__pyx_f_5_cdec_10Hypergraph__rng(struct __pyx_obj_5_cdec_Hypergr __pyx_t_1 = (__pyx_v_self->rng == NULL); if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":15 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":15 * cdef MT19937* _rng(self): * if self.rng == NULL: * self.rng = new MT19937() # <<<<<<<<<<<<<< * return self.rng * */ - try {__pyx_t_2 = new MT19937();} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[3]; __pyx_lineno = 15; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_2 = new MT19937(); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[3]; __pyx_lineno = 15; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_v_self->rng = __pyx_t_2; goto __pyx_L3; } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":16 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":16 * if self.rng == NULL: * self.rng = new MT19937() * return self.rng # <<<<<<<<<<<<<< @@ -8844,7 +8852,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_3viterbi(PyObject *__pyx_v_self, C return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":18 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":18 * return self.rng * * def viterbi(self): # <<<<<<<<<<<<<< @@ -8863,7 +8871,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_2viterbi(struct __pyx_obj_5_cdec_H int __pyx_clineno = 0; __Pyx_RefNannySetupContext("viterbi", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":21 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":21 * """hg.viterbi() -> String for the best hypothesis in the hypergraph.""" * cdef vector[WordID] trans * hypergraph.ViterbiESentence(self.hg[0], &trans) # <<<<<<<<<<<<<< @@ -8872,7 +8880,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_2viterbi(struct __pyx_obj_5_cdec_H */ ViterbiESentence((__pyx_v_self->hg[0]), (&__pyx_v_trans)); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":22 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":22 * cdef vector[WordID] trans * hypergraph.ViterbiESentence(self.hg[0], &trans) * return unicode(GetString(trans).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -8922,7 +8930,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_5viterbi_trees(PyObject *__pyx_v_s return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":24 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":24 * return unicode(GetString(trans).c_str(), 'utf8') * * def viterbi_trees(self): # <<<<<<<<<<<<<< @@ -8942,7 +8950,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_4viterbi_trees(struct __pyx_obj_5_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("viterbi_trees", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":29 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":29 * e_tree: Target tree for the best hypothesis in the hypergraph. * """ * f_tree = unicode(hypergraph.ViterbiFTree(self.hg[0]).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -8965,7 +8973,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_4viterbi_trees(struct __pyx_obj_5_ __pyx_v_f_tree = ((PyObject*)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":30 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":30 * """ * f_tree = unicode(hypergraph.ViterbiFTree(self.hg[0]).c_str(), 'utf8') * e_tree = unicode(hypergraph.ViterbiETree(self.hg[0]).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -8988,7 +8996,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_4viterbi_trees(struct __pyx_obj_5_ __pyx_v_e_tree = ((PyObject*)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":31 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":31 * f_tree = unicode(hypergraph.ViterbiFTree(self.hg[0]).c_str(), 'utf8') * e_tree = unicode(hypergraph.ViterbiETree(self.hg[0]).c_str(), 'utf8') * return (f_tree, e_tree) # <<<<<<<<<<<<<< @@ -9035,7 +9043,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_7viterbi_features(PyObject *__pyx_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":33 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":33 * return (f_tree, e_tree) * * def viterbi_features(self): # <<<<<<<<<<<<<< @@ -9053,7 +9061,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_6viterbi_features(struct __pyx_obj int __pyx_clineno = 0; __Pyx_RefNannySetupContext("viterbi_features", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":36 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":36 * """hg.viterbi_features() -> SparseVector with the features corresponding * to the best derivation in the hypergraph.""" * cdef SparseVector fmap = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -9066,7 +9074,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_6viterbi_features(struct __pyx_obj __pyx_v_fmap = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":37 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":37 * to the best derivation in the hypergraph.""" * cdef SparseVector fmap = SparseVector.__new__(SparseVector) * fmap.vector = new FastSparseVector[weight_t](hypergraph.ViterbiFeatures(self.hg[0])) # <<<<<<<<<<<<<< @@ -9075,7 +9083,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_6viterbi_features(struct __pyx_obj */ __pyx_v_fmap->vector = new FastSparseVector<weight_t>(ViterbiFeatures((__pyx_v_self->hg[0]))); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":38 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":38 * cdef SparseVector fmap = SparseVector.__new__(SparseVector) * fmap.vector = new FastSparseVector[weight_t](hypergraph.ViterbiFeatures(self.hg[0])) * return fmap # <<<<<<<<<<<<<< @@ -9111,7 +9119,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_9viterbi_forest(PyObject *__pyx_v_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":40 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":40 * return fmap * * def viterbi_forest(self): # <<<<<<<<<<<<<< @@ -9129,7 +9137,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_8viterbi_forest(struct __pyx_obj_5 int __pyx_clineno = 0; __Pyx_RefNannySetupContext("viterbi_forest", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":41 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":41 * * def viterbi_forest(self): * cdef Hypergraph hg = Hypergraph() # <<<<<<<<<<<<<< @@ -9141,7 +9149,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_8viterbi_forest(struct __pyx_obj_5 __pyx_v_hg = ((struct __pyx_obj_5_cdec_Hypergraph *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":42 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":42 * def viterbi_forest(self): * cdef Hypergraph hg = Hypergraph() * hg.hg = new hypergraph.Hypergraph(self.hg[0].CreateViterbiHypergraph(NULL).get()[0]) # <<<<<<<<<<<<<< @@ -9150,7 +9158,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_8viterbi_forest(struct __pyx_obj_5 */ __pyx_v_hg->hg = new Hypergraph(((__pyx_v_self->hg[0]).CreateViterbiHypergraph(NULL).get()[0])); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":43 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":43 * cdef Hypergraph hg = Hypergraph() * hg.hg = new hypergraph.Hypergraph(self.hg[0].CreateViterbiHypergraph(NULL).get()[0]) * return hg # <<<<<<<<<<<<<< @@ -9187,7 +9195,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_11viterbi_joshua(PyObject *__pyx_v return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":45 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":45 * return hg * * def viterbi_joshua(self): # <<<<<<<<<<<<<< @@ -9205,7 +9213,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_10viterbi_joshua(struct __pyx_obj_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("viterbi_joshua", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":47 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":47 * def viterbi_joshua(self): * """hg.viterbi_joshua() -> Joshua representation of the best derivation.""" * return unicode(hypergraph.JoshuaVisualizationString(self.hg[0]).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -9256,7 +9264,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_13kbest(PyObject *__pyx_v_self, Py return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":49 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":49 * return unicode(hypergraph.JoshuaVisualizationString(self.hg[0]).c_str(), 'utf8') * * def kbest(self, size): # <<<<<<<<<<<<<< @@ -9324,7 +9332,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_14generator4(__pyx_GeneratorObject __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 49; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":51 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":51 * def kbest(self, size): * """hg.kbest(size) -> List of k-best hypotheses in the hypergraph.""" * cdef kbest.KBestDerivations[vector[WordID], kbest.ESentenceTraversal]* derivations = new kbest.KBestDerivations[vector[WordID], kbest.ESentenceTraversal](self.hg[0], size) # <<<<<<<<<<<<<< @@ -9334,7 +9342,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_14generator4(__pyx_GeneratorObject __pyx_t_1 = __Pyx_PyInt_AsUnsignedInt(__pyx_cur_scope->__pyx_v_size); if (unlikely((__pyx_t_1 == (unsigned int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 51; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_cur_scope->__pyx_v_derivations = new KBest::KBestDerivations<std::vector<WordID>,ESentenceTraversal>((__pyx_cur_scope->__pyx_v_self->hg[0]), __pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":54 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":54 * cdef kbest.KBestDerivations[vector[WordID], kbest.ESentenceTraversal].Derivation* derivation * cdef unsigned k * try: # <<<<<<<<<<<<<< @@ -9343,7 +9351,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_14generator4(__pyx_GeneratorObject */ /*try:*/ { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":55 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":55 * cdef unsigned k * try: * for k in range(size): # <<<<<<<<<<<<<< @@ -9354,7 +9362,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_14generator4(__pyx_GeneratorObject for (__pyx_t_1 = 0; __pyx_t_1 < __pyx_t_2; __pyx_t_1+=1) { __pyx_cur_scope->__pyx_v_k = __pyx_t_1; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":56 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":56 * try: * for k in range(size): * derivation = derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) # <<<<<<<<<<<<<< @@ -9363,7 +9371,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_14generator4(__pyx_GeneratorObject */ __pyx_cur_scope->__pyx_v_derivation = __pyx_cur_scope->__pyx_v_derivations->LazyKthBest((__pyx_cur_scope->__pyx_v_self->hg->nodes_.size() - 1), __pyx_cur_scope->__pyx_v_k); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":57 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":57 * for k in range(size): * derivation = derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * if not derivation: break # <<<<<<<<<<<<<< @@ -9377,7 +9385,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_14generator4(__pyx_GeneratorObject } __pyx_L9:; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":58 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":58 * derivation = derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * if not derivation: break * yield unicode(GetString(derivation._yield).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -9414,7 +9422,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_14generator4(__pyx_GeneratorObject __pyx_L8_break:; } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":60 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":60 * yield unicode(GetString(derivation._yield).c_str(), 'utf8') * finally: * del derivations # <<<<<<<<<<<<<< @@ -9475,7 +9483,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_16kbest_trees(PyObject *__pyx_v_se return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":62 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":62 * del derivations * * def kbest_trees(self, size): # <<<<<<<<<<<<<< @@ -9545,7 +9553,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 62; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":64 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":64 * def kbest_trees(self, size): * """hg.kbest_trees(size) -> List of k-best trees in the hypergraph.""" * cdef kbest.KBestDerivations[vector[WordID], kbest.FTreeTraversal]* f_derivations = new kbest.KBestDerivations[vector[WordID], kbest.FTreeTraversal](self.hg[0], size) # <<<<<<<<<<<<<< @@ -9555,7 +9563,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject __pyx_t_1 = __Pyx_PyInt_AsUnsignedInt(__pyx_cur_scope->__pyx_v_size); if (unlikely((__pyx_t_1 == (unsigned int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 64; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_cur_scope->__pyx_v_f_derivations = new KBest::KBestDerivations<std::vector<WordID>,FTreeTraversal>((__pyx_cur_scope->__pyx_v_self->hg[0]), __pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":66 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":66 * cdef kbest.KBestDerivations[vector[WordID], kbest.FTreeTraversal]* f_derivations = new kbest.KBestDerivations[vector[WordID], kbest.FTreeTraversal](self.hg[0], size) * cdef kbest.KBestDerivations[vector[WordID], kbest.FTreeTraversal].Derivation* f_derivation * cdef kbest.KBestDerivations[vector[WordID], kbest.ETreeTraversal]* e_derivations = new kbest.KBestDerivations[vector[WordID], kbest.ETreeTraversal](self.hg[0], size) # <<<<<<<<<<<<<< @@ -9565,7 +9573,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject __pyx_t_1 = __Pyx_PyInt_AsUnsignedInt(__pyx_cur_scope->__pyx_v_size); if (unlikely((__pyx_t_1 == (unsigned int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 66; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_cur_scope->__pyx_v_e_derivations = new KBest::KBestDerivations<std::vector<WordID>,ETreeTraversal>((__pyx_cur_scope->__pyx_v_self->hg[0]), __pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":69 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":69 * cdef kbest.KBestDerivations[vector[WordID], kbest.ETreeTraversal].Derivation* e_derivation * cdef unsigned k * try: # <<<<<<<<<<<<<< @@ -9574,7 +9582,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject */ /*try:*/ { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":70 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":70 * cdef unsigned k * try: * for k in range(size): # <<<<<<<<<<<<<< @@ -9585,7 +9593,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject for (__pyx_t_1 = 0; __pyx_t_1 < __pyx_t_2; __pyx_t_1+=1) { __pyx_cur_scope->__pyx_v_k = __pyx_t_1; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":71 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":71 * try: * for k in range(size): * f_derivation = f_derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) # <<<<<<<<<<<<<< @@ -9594,7 +9602,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject */ __pyx_cur_scope->__pyx_v_f_derivation = __pyx_cur_scope->__pyx_v_f_derivations->LazyKthBest((__pyx_cur_scope->__pyx_v_self->hg->nodes_.size() - 1), __pyx_cur_scope->__pyx_v_k); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":72 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":72 * for k in range(size): * f_derivation = f_derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * e_derivation = e_derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) # <<<<<<<<<<<<<< @@ -9603,7 +9611,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject */ __pyx_cur_scope->__pyx_v_e_derivation = __pyx_cur_scope->__pyx_v_e_derivations->LazyKthBest((__pyx_cur_scope->__pyx_v_self->hg->nodes_.size() - 1), __pyx_cur_scope->__pyx_v_k); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":73 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":73 * f_derivation = f_derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * e_derivation = e_derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * if not f_derivation or not e_derivation: break # <<<<<<<<<<<<<< @@ -9623,7 +9631,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject } __pyx_L9:; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":74 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":74 * e_derivation = e_derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * if not f_derivation or not e_derivation: break * f_tree = unicode(GetString(f_derivation._yield).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -9649,7 +9657,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject __pyx_cur_scope->__pyx_v_f_tree = ((PyObject*)__pyx_t_6); __pyx_t_6 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":75 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":75 * if not f_derivation or not e_derivation: break * f_tree = unicode(GetString(f_derivation._yield).c_str(), 'utf8') * e_tree = unicode(GetString(e_derivation._yield).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -9675,7 +9683,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject __pyx_cur_scope->__pyx_v_e_tree = ((PyObject*)__pyx_t_6); __pyx_t_6 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":76 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":76 * f_tree = unicode(GetString(f_derivation._yield).c_str(), 'utf8') * e_tree = unicode(GetString(e_derivation._yield).c_str(), 'utf8') * yield (f_tree, e_tree) # <<<<<<<<<<<<<< @@ -9707,7 +9715,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject __pyx_L8_break:; } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":78 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":78 * yield (f_tree, e_tree) * finally: * del f_derivations # <<<<<<<<<<<<<< @@ -9731,7 +9739,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_17generator5(__pyx_GeneratorObject __pyx_L6:; delete __pyx_cur_scope->__pyx_v_f_derivations; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":79 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":79 * finally: * del f_derivations * del e_derivations # <<<<<<<<<<<<<< @@ -9777,7 +9785,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_19kbest_features(PyObject *__pyx_v return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":81 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":81 * del e_derivations * * def kbest_features(self, size): # <<<<<<<<<<<<<< @@ -9844,7 +9852,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 81; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":83 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":83 * def kbest_features(self, size): * """hg.kbest_trees(size) -> List of k-best feature vectors in the hypergraph.""" * cdef kbest.KBestDerivations[FastSparseVector[weight_t], kbest.FeatureVectorTraversal]* derivations = new kbest.KBestDerivations[FastSparseVector[weight_t], kbest.FeatureVectorTraversal](self.hg[0], size) # <<<<<<<<<<<<<< @@ -9854,7 +9862,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject __pyx_t_1 = __Pyx_PyInt_AsUnsignedInt(__pyx_cur_scope->__pyx_v_size); if (unlikely((__pyx_t_1 == (unsigned int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 83; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_cur_scope->__pyx_v_derivations = new KBest::KBestDerivations<FastSparseVector<weight_t>,FeatureVectorTraversal>((__pyx_cur_scope->__pyx_v_self->hg[0]), __pyx_t_1); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":87 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":87 * cdef SparseVector fmap * cdef unsigned k * try: # <<<<<<<<<<<<<< @@ -9863,7 +9871,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject */ /*try:*/ { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":88 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":88 * cdef unsigned k * try: * for k in range(size): # <<<<<<<<<<<<<< @@ -9874,7 +9882,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject for (__pyx_t_1 = 0; __pyx_t_1 < __pyx_t_2; __pyx_t_1+=1) { __pyx_cur_scope->__pyx_v_k = __pyx_t_1; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":89 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":89 * try: * for k in range(size): * derivation = derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) # <<<<<<<<<<<<<< @@ -9883,7 +9891,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject */ __pyx_cur_scope->__pyx_v_derivation = __pyx_cur_scope->__pyx_v_derivations->LazyKthBest((__pyx_cur_scope->__pyx_v_self->hg->nodes_.size() - 1), __pyx_cur_scope->__pyx_v_k); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":90 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":90 * for k in range(size): * derivation = derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * if not derivation: break # <<<<<<<<<<<<<< @@ -9897,7 +9905,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject } __pyx_L9:; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":91 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":91 * derivation = derivations.LazyKthBest(self.hg.nodes_.size() - 1, k) * if not derivation: break * fmap = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -9913,7 +9921,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject __pyx_cur_scope->__pyx_v_fmap = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_4); __pyx_t_4 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":92 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":92 * if not derivation: break * fmap = SparseVector.__new__(SparseVector) * fmap.vector = new FastSparseVector[weight_t](derivation._yield) # <<<<<<<<<<<<<< @@ -9922,7 +9930,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject */ __pyx_cur_scope->__pyx_v_fmap->vector = new FastSparseVector<weight_t>(__pyx_cur_scope->__pyx_v_derivation->yield); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":93 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":93 * fmap = SparseVector.__new__(SparseVector) * fmap.vector = new FastSparseVector[weight_t](derivation._yield) * yield fmap # <<<<<<<<<<<<<< @@ -9946,7 +9954,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_20generator6(__pyx_GeneratorObject __pyx_L8_break:; } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":95 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":95 * yield fmap * finally: * del derivations # <<<<<<<<<<<<<< @@ -10015,7 +10023,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_22sample(PyObject *__pyx_v_self, P return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":97 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":97 * del derivations * * def sample(self, unsigned n): # <<<<<<<<<<<<<< @@ -10081,17 +10089,22 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_23generator7(__pyx_GeneratorObject __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 97; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":99 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":99 * def sample(self, unsigned n): * """hg.sample(n) -> Sample of n hypotheses from the hypergraph.""" * cdef vector[hypergraph.Hypothesis]* hypos = new vector[hypergraph.Hypothesis]() # <<<<<<<<<<<<<< * hypergraph.sample_hypotheses(self.hg[0], n, self._rng(), hypos) * cdef unsigned k */ - try {__pyx_t_1 = new std::vector<HypergraphSampler::Hypothesis>();} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[3]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_1 = new std::vector<HypergraphSampler::Hypothesis>(); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[3]; __pyx_lineno = 99; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_cur_scope->__pyx_v_hypos = __pyx_t_1; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":100 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":100 * """hg.sample(n) -> Sample of n hypotheses from the hypergraph.""" * cdef vector[hypergraph.Hypothesis]* hypos = new vector[hypergraph.Hypothesis]() * hypergraph.sample_hypotheses(self.hg[0], n, self._rng(), hypos) # <<<<<<<<<<<<<< @@ -10100,7 +10113,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_23generator7(__pyx_GeneratorObject */ HypergraphSampler::sample_hypotheses((__pyx_cur_scope->__pyx_v_self->hg[0]), __pyx_cur_scope->__pyx_v_n, ((struct __pyx_vtabstruct_5_cdec_Hypergraph *)__pyx_cur_scope->__pyx_v_self->__pyx_vtab)->_rng(__pyx_cur_scope->__pyx_v_self), __pyx_cur_scope->__pyx_v_hypos); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":102 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":102 * hypergraph.sample_hypotheses(self.hg[0], n, self._rng(), hypos) * cdef unsigned k * try: # <<<<<<<<<<<<<< @@ -10109,7 +10122,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_23generator7(__pyx_GeneratorObject */ /*try:*/ { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":103 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":103 * cdef unsigned k * try: * for k in range(hypos.size()): # <<<<<<<<<<<<<< @@ -10120,7 +10133,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_23generator7(__pyx_GeneratorObject for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_2; __pyx_t_3+=1) { __pyx_cur_scope->__pyx_v_k = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":104 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":104 * try: * for k in range(hypos.size()): * yield unicode(GetString(hypos[0][k].words).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -10156,7 +10169,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_23generator7(__pyx_GeneratorObject } } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":106 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":106 * yield unicode(GetString(hypos[0][k].words).c_str(), 'utf8') * finally: * del hypos # <<<<<<<<<<<<<< @@ -10227,7 +10240,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_25sample_trees(PyObject *__pyx_v_s return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":108 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":108 * del hypos * * def sample_trees(self, unsigned n): # <<<<<<<<<<<<<< @@ -10293,17 +10306,22 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_26generator8(__pyx_GeneratorObject __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":110 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":110 * def sample_trees(self, unsigned n): * """hg.sample_trees(n) -> Sample of n trees from the hypergraph.""" * cdef vector[string]* trees = new vector[string]() # <<<<<<<<<<<<<< * hypergraph.sample_trees(self.hg[0], n, self._rng(), trees) * cdef unsigned k */ - try {__pyx_t_1 = new std::vector<std::string>();} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[3]; __pyx_lineno = 110; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_1 = new std::vector<std::string>(); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[3]; __pyx_lineno = 110; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_cur_scope->__pyx_v_trees = __pyx_t_1; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":111 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":111 * """hg.sample_trees(n) -> Sample of n trees from the hypergraph.""" * cdef vector[string]* trees = new vector[string]() * hypergraph.sample_trees(self.hg[0], n, self._rng(), trees) # <<<<<<<<<<<<<< @@ -10312,7 +10330,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_26generator8(__pyx_GeneratorObject */ HypergraphSampler::sample_trees((__pyx_cur_scope->__pyx_v_self->hg[0]), __pyx_cur_scope->__pyx_v_n, ((struct __pyx_vtabstruct_5_cdec_Hypergraph *)__pyx_cur_scope->__pyx_v_self->__pyx_vtab)->_rng(__pyx_cur_scope->__pyx_v_self), __pyx_cur_scope->__pyx_v_trees); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":113 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":113 * hypergraph.sample_trees(self.hg[0], n, self._rng(), trees) * cdef unsigned k * try: # <<<<<<<<<<<<<< @@ -10321,7 +10339,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_26generator8(__pyx_GeneratorObject */ /*try:*/ { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":114 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":114 * cdef unsigned k * try: * for k in range(trees.size()): # <<<<<<<<<<<<<< @@ -10332,7 +10350,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_26generator8(__pyx_GeneratorObject for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_2; __pyx_t_3+=1) { __pyx_cur_scope->__pyx_v_k = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":115 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":115 * try: * for k in range(trees.size()): * yield unicode(trees[0][k].c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -10368,7 +10386,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_26generator8(__pyx_GeneratorObject } } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":117 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":117 * yield unicode(trees[0][k].c_str(), 'utf8') * finally: * del trees # <<<<<<<<<<<<<< @@ -10428,7 +10446,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_28intersect(PyObject *__pyx_v_self return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":119 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":119 * del trees * * def intersect(self, inp): # <<<<<<<<<<<<<< @@ -10440,28 +10458,25 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_27intersect(struct __pyx_obj_5_cde struct __pyx_obj_5_cdec_Lattice *__pyx_v_lat = 0; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("intersect", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":122 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":122 * """hg.intersect(Lattice/string): Intersect the hypergraph with the provided reference.""" * cdef Lattice lat * if isinstance(inp, Lattice): # <<<<<<<<<<<<<< * lat = <Lattice> inp * elif isinstance(inp, basestring): */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Lattice)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_inp, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_inp, ((PyObject*)__pyx_ptype_5_cdec_Lattice)); + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":123 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":123 * cdef Lattice lat * if isinstance(inp, Lattice): * lat = <Lattice> inp # <<<<<<<<<<<<<< @@ -10473,41 +10488,38 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_27intersect(struct __pyx_obj_5_cde goto __pyx_L3; } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":124 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":124 * if isinstance(inp, Lattice): * lat = <Lattice> inp * elif isinstance(inp, basestring): # <<<<<<<<<<<<<< * lat = Lattice(inp) * else: */ - __pyx_t_1 = __pyx_builtin_basestring; - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = PyObject_IsInstance(__pyx_v_inp, __pyx_t_1); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 124; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyObject_IsInstance(__pyx_v_inp, __pyx_builtin_basestring); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 124; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":125 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":125 * lat = <Lattice> inp * elif isinstance(inp, basestring): * lat = Lattice(inp) # <<<<<<<<<<<<<< * else: * raise TypeError('cannot intersect hypergraph with %s' % type(inp)) */ - __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_INCREF(__pyx_v_inp); - PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_v_inp); + PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_v_inp); __Pyx_GIVEREF(__pyx_v_inp); - __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Lattice)), ((PyObject *)__pyx_t_1), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Lattice)), ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 125; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; + __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; __pyx_v_lat = ((struct __pyx_obj_5_cdec_Lattice *)__pyx_t_3); __pyx_t_3 = 0; goto __pyx_L3; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":127 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":127 * lat = Lattice(inp) * else: * raise TypeError('cannot intersect hypergraph with %s' % type(inp)) # <<<<<<<<<<<<<< @@ -10516,21 +10528,21 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_27intersect(struct __pyx_obj_5_cde */ __pyx_t_3 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_15), ((PyObject *)Py_TYPE(__pyx_v_inp))); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_3)); - __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - PyTuple_SET_ITEM(__pyx_t_1, 0, ((PyObject *)__pyx_t_3)); + __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + PyTuple_SET_ITEM(__pyx_t_2, 0, ((PyObject *)__pyx_t_3)); __Pyx_GIVEREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; - __pyx_t_3 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_1), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; + __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; __Pyx_Raise(__pyx_t_3, 0, 0, 0); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; {__pyx_filename = __pyx_f[3]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":128 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":128 * else: * raise TypeError('cannot intersect hypergraph with %s' % type(inp)) * return hypergraph.Intersect(lat.lattice[0], self.hg) # <<<<<<<<<<<<<< @@ -10547,7 +10559,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_27intersect(struct __pyx_obj_5_cde __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_AddTraceback("_cdec.Hypergraph.intersect", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; @@ -10626,7 +10638,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_30prune(PyObject *__pyx_v_self, Py return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":130 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":130 * return hypergraph.Intersect(lat.lattice[0], self.hg) * * def prune(self, beam_alpha=0, density=0, **kwargs): # <<<<<<<<<<<<<< @@ -10646,7 +10658,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_29prune(struct __pyx_obj_5_cdec_Hy int __pyx_clineno = 0; __Pyx_RefNannySetupContext("prune", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":134 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":134 * beam_alpha: use beam pruning * density: use density pruning""" * cdef hypergraph.EdgeMask* preserve_mask = NULL # <<<<<<<<<<<<<< @@ -10655,7 +10667,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_29prune(struct __pyx_obj_5_cdec_Hy */ __pyx_v_preserve_mask = NULL; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":135 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":135 * density: use density pruning""" * cdef hypergraph.EdgeMask* preserve_mask = NULL * if 'csplit_preserve_full_word' in kwargs: # <<<<<<<<<<<<<< @@ -10665,7 +10677,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_29prune(struct __pyx_obj_5_cdec_Hy __pyx_t_1 = (__Pyx_PyDict_Contains(((PyObject *)__pyx_n_s_16), ((PyObject *)__pyx_v_kwargs), Py_EQ)); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 135; __pyx_clineno = __LINE__; goto __pyx_L1_error;} if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":136 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":136 * cdef hypergraph.EdgeMask* preserve_mask = NULL * if 'csplit_preserve_full_word' in kwargs: * preserve_mask = new hypergraph.EdgeMask(self.hg.edges_.size()) # <<<<<<<<<<<<<< @@ -10674,7 +10686,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_29prune(struct __pyx_obj_5_cdec_Hy */ __pyx_v_preserve_mask = new std::vector<bool>(__pyx_v_self->hg->edges_.size()); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":137 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":137 * if 'csplit_preserve_full_word' in kwargs: * preserve_mask = new hypergraph.EdgeMask(self.hg.edges_.size()) * preserve_mask[0][hypergraph.GetFullWordEdgeIndex(self.hg[0])] = True # <<<<<<<<<<<<<< @@ -10686,7 +10698,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_29prune(struct __pyx_obj_5_cdec_Hy } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":138 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":138 * preserve_mask = new hypergraph.EdgeMask(self.hg.edges_.size()) * preserve_mask[0][hypergraph.GetFullWordEdgeIndex(self.hg[0])] = True * self.hg.PruneInsideOutside(beam_alpha, density, preserve_mask, False, 1, False) # <<<<<<<<<<<<<< @@ -10697,7 +10709,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_29prune(struct __pyx_obj_5_cdec_Hy __pyx_t_3 = __pyx_PyFloat_AsDouble(__pyx_v_density); if (unlikely((__pyx_t_3 == (double)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_self->hg->PruneInsideOutside(__pyx_t_2, __pyx_t_3, __pyx_v_preserve_mask, 0, 1.0, 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":139 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":139 * preserve_mask[0][hypergraph.GetFullWordEdgeIndex(self.hg[0])] = True * self.hg.PruneInsideOutside(beam_alpha, density, preserve_mask, False, 1, False) * if preserve_mask: # <<<<<<<<<<<<<< @@ -10707,7 +10719,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_29prune(struct __pyx_obj_5_cdec_Hy __pyx_t_1 = (__pyx_v_preserve_mask != 0); if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":140 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":140 * self.hg.PruneInsideOutside(beam_alpha, density, preserve_mask, False, 1, False) * if preserve_mask: * del preserve_mask # <<<<<<<<<<<<<< @@ -10742,7 +10754,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_32lattice(PyObject *__pyx_v_self, return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":142 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":142 * del preserve_mask * * def lattice(self): # TODO direct hg -> lattice conversion in cdec # <<<<<<<<<<<<<< @@ -10762,7 +10774,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_31lattice(struct __pyx_obj_5_cdec_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("lattice", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":144 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":144 * def lattice(self): # TODO direct hg -> lattice conversion in cdec * """hg.lattice() -> Lattice corresponding to the hypergraph.""" * cdef bytes plf = hypergraph.AsPLF(self.hg[0], True).c_str() # <<<<<<<<<<<<<< @@ -10771,10 +10783,10 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_31lattice(struct __pyx_obj_5_cdec_ */ __pyx_t_1 = PyBytes_FromString(HypergraphIO::AsPLF((__pyx_v_self->hg[0]), 1).c_str()); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 144; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_1)); - __pyx_v_plf = __pyx_t_1; + __pyx_v_plf = ((PyObject*)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":145 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":145 * """hg.lattice() -> Lattice corresponding to the hypergraph.""" * cdef bytes plf = hypergraph.AsPLF(self.hg[0], True).c_str() * return Lattice(eval(plf)) # <<<<<<<<<<<<<< @@ -10845,7 +10857,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_34plf(PyObject *__pyx_v_self, CYTH return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":147 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":147 * return Lattice(eval(plf)) * * def plf(self): # <<<<<<<<<<<<<< @@ -10863,7 +10875,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_33plf(struct __pyx_obj_5_cdec_Hype int __pyx_clineno = 0; __Pyx_RefNannySetupContext("plf", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":149 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":149 * def plf(self): * """hg.plf() -> Lattice PLF representation corresponding to the hypergraph.""" * return bytes(hypergraph.AsPLF(self.hg[0], True).c_str()) # <<<<<<<<<<<<<< @@ -10910,7 +10922,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_36reweight(PyObject *__pyx_v_self, return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":151 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":151 * return bytes(hypergraph.AsPLF(self.hg[0], True).c_str()) * * def reweight(self, weights): # <<<<<<<<<<<<<< @@ -10921,28 +10933,25 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_36reweight(PyObject *__pyx_v_self, static PyObject *__pyx_pf_5_cdec_10Hypergraph_35reweight(struct __pyx_obj_5_cdec_Hypergraph *__pyx_v_self, PyObject *__pyx_v_weights) { PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("reweight", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":153 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":153 * def reweight(self, weights): * """hg.reweight(SparseVector/DenseVector): Reweight the hypergraph with a new vector.""" * if isinstance(weights, SparseVector): # <<<<<<<<<<<<<< * self.hg.Reweight((<SparseVector> weights).vector[0]) * elif isinstance(weights, DenseVector): */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SparseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_weights, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_weights, ((PyObject*)__pyx_ptype_5_cdec_SparseVector)); + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":154 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":154 * """hg.reweight(SparseVector/DenseVector): Reweight the hypergraph with a new vector.""" * if isinstance(weights, SparseVector): * self.hg.Reweight((<SparseVector> weights).vector[0]) # <<<<<<<<<<<<<< @@ -10953,20 +10962,17 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_35reweight(struct __pyx_obj_5_cdec goto __pyx_L3; } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":155 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":155 * if isinstance(weights, SparseVector): * self.hg.Reweight((<SparseVector> weights).vector[0]) * elif isinstance(weights, DenseVector): # <<<<<<<<<<<<<< * self.hg.Reweight((<DenseVector> weights).vector[0]) * else: */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_DenseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_weights, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_weights, ((PyObject*)__pyx_ptype_5_cdec_DenseVector)); + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":156 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":156 * self.hg.Reweight((<SparseVector> weights).vector[0]) * elif isinstance(weights, DenseVector): * self.hg.Reweight((<DenseVector> weights).vector[0]) # <<<<<<<<<<<<<< @@ -10978,25 +10984,25 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_35reweight(struct __pyx_obj_5_cdec } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":158 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":158 * self.hg.Reweight((<DenseVector> weights).vector[0]) * else: * raise TypeError('cannot reweight hypergraph with %s' % type(weights)) # <<<<<<<<<<<<<< * * property edges: */ - __pyx_t_1 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_17), ((PyObject *)Py_TYPE(__pyx_v_weights))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 158; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(((PyObject *)__pyx_t_1)); + __pyx_t_2 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_17), ((PyObject *)Py_TYPE(__pyx_v_weights))); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 158; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(((PyObject *)__pyx_t_2)); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 158; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_t_1)); - __Pyx_GIVEREF(((PyObject *)__pyx_t_1)); - __pyx_t_1 = 0; - __pyx_t_1 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 158; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_t_2)); + __Pyx_GIVEREF(((PyObject *)__pyx_t_2)); + __pyx_t_2 = 0; + __pyx_t_2 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 158; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; - __Pyx_Raise(__pyx_t_1, 0, 0, 0); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_Raise(__pyx_t_2, 0, 0, 0); + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; {__pyx_filename = __pyx_f[3]; __pyx_lineno = 158; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_L3:; @@ -11004,7 +11010,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_35reweight(struct __pyx_obj_5_cdec __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_AddTraceback("_cdec.Hypergraph.reweight", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; @@ -11026,7 +11032,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_5edges_1__get__(PyObject *__pyx_v_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":161 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":161 * * property edges: * def __get__(self): # <<<<<<<<<<<<<< @@ -11090,7 +11096,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_5edges_2generator9(__pyx_Generator __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 161; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":163 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":163 * def __get__(self): * cdef unsigned i * for i in range(self.hg.edges_.size()): # <<<<<<<<<<<<<< @@ -11101,7 +11107,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_5edges_2generator9(__pyx_Generator for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":164 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":164 * cdef unsigned i * for i in range(self.hg.edges_.size()): * yield HypergraphEdge().init(self.hg, i) # <<<<<<<<<<<<<< @@ -11153,7 +11159,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_5nodes_1__get__(PyObject *__pyx_v_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":167 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":167 * * property nodes: * def __get__(self): # <<<<<<<<<<<<<< @@ -11217,7 +11223,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_5nodes_2generator10(__pyx_Generato __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 167; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":169 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":169 * def __get__(self): * cdef unsigned i * for i in range(self.hg.nodes_.size()): # <<<<<<<<<<<<<< @@ -11228,7 +11234,7 @@ static PyObject *__pyx_gb_5_cdec_10Hypergraph_5nodes_2generator10(__pyx_Generato for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":170 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":170 * cdef unsigned i * for i in range(self.hg.nodes_.size()): * yield HypergraphNode().init(self.hg, i) # <<<<<<<<<<<<<< @@ -11279,7 +11285,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_4goal_1__get__(PyObject *__pyx_v_s return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":173 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":173 * * property goal: * def __get__(self): # <<<<<<<<<<<<<< @@ -11297,7 +11303,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_4goal___get__(struct __pyx_obj_5_c int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":174 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":174 * property goal: * def __get__(self): * return HypergraphNode().init(self.hg, self.hg.GoalNode()) # <<<<<<<<<<<<<< @@ -11338,7 +11344,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_6npaths_1__get__(PyObject *__pyx_v return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":177 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":177 * * property npaths: * def __get__(self): # <<<<<<<<<<<<<< @@ -11355,7 +11361,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_6npaths___get__(struct __pyx_obj_5 int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":178 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":178 * property npaths: * def __get__(self): * return self.hg.NumberOfPaths() # <<<<<<<<<<<<<< @@ -11393,7 +11399,7 @@ static PyObject *__pyx_pw_5_cdec_10Hypergraph_38inside_outside(PyObject *__pyx_v return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":180 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":180 * return self.hg.NumberOfPaths() * * def inside_outside(self): # <<<<<<<<<<<<<< @@ -11417,7 +11423,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("inside_outside", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":182 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":182 * def inside_outside(self): * """hg.inside_outside() -> SparseVector with inside-outside scores for each feature.""" * cdef FastSparseVector[prob_t]* result = new FastSparseVector[prob_t]() # <<<<<<<<<<<<<< @@ -11426,7 +11432,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ __pyx_v_result = new FastSparseVector<prob_t>(); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":183 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":183 * """hg.inside_outside() -> SparseVector with inside-outside scores for each feature.""" * cdef FastSparseVector[prob_t]* result = new FastSparseVector[prob_t]() * cdef prob_t z = hypergraph.InsideOutside(self.hg[0], result) # <<<<<<<<<<<<<< @@ -11435,7 +11441,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ __pyx_v_z = InsideOutside<prob_t, EdgeProb, SparseVector<prob_t>, EdgeFeaturesAndProbWeightFunction>((__pyx_v_self->hg[0]), __pyx_v_result); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":184 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":184 * cdef FastSparseVector[prob_t]* result = new FastSparseVector[prob_t]() * cdef prob_t z = hypergraph.InsideOutside(self.hg[0], result) * result[0] /= z # <<<<<<<<<<<<<< @@ -11444,7 +11450,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ (__pyx_v_result[0]) /= __pyx_v_z; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":185 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":185 * cdef prob_t z = hypergraph.InsideOutside(self.hg[0], result) * result[0] /= z * cdef SparseVector vector = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -11457,7 +11463,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":186 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":186 * result[0] /= z * cdef SparseVector vector = SparseVector.__new__(SparseVector) * vector.vector = new FastSparseVector[double]() # <<<<<<<<<<<<<< @@ -11466,7 +11472,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ __pyx_v_vector->vector = new FastSparseVector<double>(); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":187 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":187 * cdef SparseVector vector = SparseVector.__new__(SparseVector) * vector.vector = new FastSparseVector[double]() * cdef FastSparseVector[prob_t].const_iterator* it = new FastSparseVector[prob_t].const_iterator(result[0], False) # <<<<<<<<<<<<<< @@ -11475,7 +11481,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ __pyx_v_it = new FastSparseVector<prob_t>::const_iterator((__pyx_v_result[0]), 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":189 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":189 * cdef FastSparseVector[prob_t].const_iterator* it = new FastSparseVector[prob_t].const_iterator(result[0], False) * cdef unsigned i * for i in range(result.size()): # <<<<<<<<<<<<<< @@ -11486,7 +11492,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_2; __pyx_t_3+=1) { __pyx_v_i = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":190 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":190 * cdef unsigned i * for i in range(result.size()): * vector.vector.set_value(it[0].ptr().first, log(it[0].ptr().second)) # <<<<<<<<<<<<<< @@ -11495,7 +11501,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ __pyx_v_vector->vector->set_value((__pyx_v_it[0]).operator->()->first, log((__pyx_v_it[0]).operator->()->second)); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":191 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":191 * for i in range(result.size()): * vector.vector.set_value(it[0].ptr().first, log(it[0].ptr().second)) * pinc(it[0]) # ++it # <<<<<<<<<<<<<< @@ -11505,7 +11511,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ (++(__pyx_v_it[0])); } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":192 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":192 * vector.vector.set_value(it[0].ptr().first, log(it[0].ptr().second)) * pinc(it[0]) # ++it * del it # <<<<<<<<<<<<<< @@ -11514,7 +11520,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ delete __pyx_v_it; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":193 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":193 * pinc(it[0]) # ++it * del it * del result # <<<<<<<<<<<<<< @@ -11523,7 +11529,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ */ delete __pyx_v_result; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":194 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":194 * del it * del result * return vector # <<<<<<<<<<<<<< @@ -11548,7 +11554,7 @@ static PyObject *__pyx_pf_5_cdec_10Hypergraph_37inside_outside(struct __pyx_obj_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":201 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":201 * cdef public TRule trule * * cdef init(self, hypergraph.Hypergraph* hg, unsigned i): # <<<<<<<<<<<<<< @@ -11565,7 +11571,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphEdge_init(struct __pyx_obj_5_cdec_Hy int __pyx_clineno = 0; __Pyx_RefNannySetupContext("init", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":202 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":202 * * cdef init(self, hypergraph.Hypergraph* hg, unsigned i): * self.hg = hg # <<<<<<<<<<<<<< @@ -11574,7 +11580,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphEdge_init(struct __pyx_obj_5_cdec_Hy */ __pyx_v_self->hg = __pyx_v_hg; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":203 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":203 * cdef init(self, hypergraph.Hypergraph* hg, unsigned i): * self.hg = hg * self.edge = &hg.edges_[i] # <<<<<<<<<<<<<< @@ -11583,7 +11589,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphEdge_init(struct __pyx_obj_5_cdec_Hy */ __pyx_v_self->edge = (&(__pyx_v_hg->edges_[__pyx_v_i])); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":204 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":204 * self.hg = hg * self.edge = &hg.edges_[i] * self.trule = TRule.__new__(TRule) # <<<<<<<<<<<<<< @@ -11599,7 +11605,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphEdge_init(struct __pyx_obj_5_cdec_Hy __pyx_v_self->trule = ((struct __pyx_obj_5_cdec_TRule *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":205 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":205 * self.edge = &hg.edges_[i] * self.trule = TRule.__new__(TRule) * self.trule.rule = new shared_ptr[grammar.TRule](self.edge.rule_) # <<<<<<<<<<<<<< @@ -11608,7 +11614,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphEdge_init(struct __pyx_obj_5_cdec_Hy */ __pyx_v_self->trule->rule = new boost::shared_ptr<TRule>(__pyx_v_self->edge->rule_); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":206 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":206 * self.trule = TRule.__new__(TRule) * self.trule.rule = new shared_ptr[grammar.TRule](self.edge.rule_) * return self # <<<<<<<<<<<<<< @@ -11643,7 +11649,7 @@ static Py_ssize_t __pyx_pw_5_cdec_14HypergraphEdge_1__len__(PyObject *__pyx_v_se return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":208 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":208 * return self * * def __len__(self): # <<<<<<<<<<<<<< @@ -11656,7 +11662,7 @@ static Py_ssize_t __pyx_pf_5_cdec_14HypergraphEdge___len__(struct __pyx_obj_5_cd __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__len__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":209 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":209 * * def __len__(self): * return self.edge.tail_nodes_.size() # <<<<<<<<<<<<<< @@ -11683,7 +11689,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_9head_node_1__get__(PyObject * return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":212 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":212 * * property head_node: * def __get__(self): # <<<<<<<<<<<<<< @@ -11701,7 +11707,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_9head_node___get__(struct __py int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":213 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":213 * property head_node: * def __get__(self): * return HypergraphNode().init(self.hg, self.edge.head_node_) # <<<<<<<<<<<<<< @@ -11743,7 +11749,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_10tail_nodes_1__get__(PyObject return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":216 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":216 * * property tail_nodes: * def __get__(self): # <<<<<<<<<<<<<< @@ -11807,7 +11813,7 @@ static PyObject *__pyx_gb_5_cdec_14HypergraphEdge_10tail_nodes_2generator11(__py __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 216; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":218 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":218 * def __get__(self): * cdef unsigned i * for i in range(self.edge.tail_nodes_.size()): # <<<<<<<<<<<<<< @@ -11818,7 +11824,7 @@ static PyObject *__pyx_gb_5_cdec_14HypergraphEdge_10tail_nodes_2generator11(__py for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":219 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":219 * cdef unsigned i * for i in range(self.edge.tail_nodes_.size()): * yield HypergraphNode().init(self.hg, self.edge.tail_nodes_[i]) # <<<<<<<<<<<<<< @@ -11869,7 +11875,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_4span_1__get__(PyObject *__pyx return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":222 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":222 * * property span: * def __get__(self): # <<<<<<<<<<<<<< @@ -11888,7 +11894,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_4span___get__(struct __pyx_obj int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":223 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":223 * property span: * def __get__(self): * return (self.edge.i_, self.edge.j_) # <<<<<<<<<<<<<< @@ -11937,7 +11943,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_8src_span_1__get__(PyObject *_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":226 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":226 * * property src_span: * def __get__(self): # <<<<<<<<<<<<<< @@ -11956,7 +11962,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_8src_span___get__(struct __pyx int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":227 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":227 * property src_span: * def __get__(self): * return (self.edge.prev_i_, self.edge.prev_j_) # <<<<<<<<<<<<<< @@ -12005,7 +12011,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_14feature_values_1__get__(PyOb return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":230 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":230 * * property feature_values: * def __get__(self): # <<<<<<<<<<<<<< @@ -12023,7 +12029,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_14feature_values___get__(struc int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":231 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":231 * property feature_values: * def __get__(self): * cdef SparseVector vector = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -12036,7 +12042,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_14feature_values___get__(struc __pyx_v_vector = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":232 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":232 * def __get__(self): * cdef SparseVector vector = SparseVector.__new__(SparseVector) * vector.vector = new FastSparseVector[double](self.edge.feature_values_) # <<<<<<<<<<<<<< @@ -12045,7 +12051,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_14feature_values___get__(struc */ __pyx_v_vector->vector = new FastSparseVector<double>(__pyx_v_self->edge->feature_values_); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":233 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":233 * cdef SparseVector vector = SparseVector.__new__(SparseVector) * vector.vector = new FastSparseVector[double](self.edge.feature_values_) * return vector # <<<<<<<<<<<<<< @@ -12081,7 +12087,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_4prob_1__get__(PyObject *__pyx return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":236 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":236 * * property prob: * def __get__(self): # <<<<<<<<<<<<<< @@ -12098,7 +12104,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_4prob___get__(struct __pyx_obj int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":237 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":237 * property prob: * def __get__(self): * return self.edge.edge_prob_.as_float() # <<<<<<<<<<<<<< @@ -12141,7 +12147,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_3__richcmp__(PyObject *__pyx_v return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":239 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":239 * return self.edge.edge_prob_.as_float() * * def __richcmp__(HypergraphEdge x, HypergraphEdge y, int op): # <<<<<<<<<<<<<< @@ -12159,7 +12165,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_2__richcmp__(struct __pyx_obj_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__richcmp__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":242 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":242 * if op == 2: # == * return x.edge == y.edge * elif op == 3: # != # <<<<<<<<<<<<<< @@ -12168,7 +12174,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_2__richcmp__(struct __pyx_obj_ */ switch (__pyx_v_op) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":240 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":240 * * def __richcmp__(HypergraphEdge x, HypergraphEdge y, int op): * if op == 2: # == # <<<<<<<<<<<<<< @@ -12177,7 +12183,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_2__richcmp__(struct __pyx_obj_ */ case 2: - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":241 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":241 * def __richcmp__(HypergraphEdge x, HypergraphEdge y, int op): * if op == 2: # == * return x.edge == y.edge # <<<<<<<<<<<<<< @@ -12192,7 +12198,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_2__richcmp__(struct __pyx_obj_ goto __pyx_L0; break; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":242 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":242 * if op == 2: # == * return x.edge == y.edge * elif op == 3: # != # <<<<<<<<<<<<<< @@ -12201,7 +12207,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_2__richcmp__(struct __pyx_obj_ */ case 3: - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":243 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":243 * return x.edge == y.edge * elif op == 3: # != * return not (x == y) # <<<<<<<<<<<<<< @@ -12220,7 +12226,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphEdge_2__richcmp__(struct __pyx_obj_ break; } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":244 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":244 * elif op == 3: # != * return not (x == y) * raise NotImplemented('comparison not implemented for HypergraphEdge') # <<<<<<<<<<<<<< @@ -12256,7 +12262,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphEdge_5trule_1__get__(PyObject *__py return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":199 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":199 * cdef hypergraph.Hypergraph* hg * cdef hypergraph.HypergraphEdge* edge * cdef public TRule trule # <<<<<<<<<<<<<< @@ -12341,7 +12347,7 @@ static int __pyx_pf_5_cdec_14HypergraphEdge_5trule_4__del__(struct __pyx_obj_5_c return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":250 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":250 * cdef hypergraph.HypergraphNode* node * * cdef init(self, hypergraph.Hypergraph* hg, unsigned i): # <<<<<<<<<<<<<< @@ -12354,7 +12360,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphNode_init(struct __pyx_obj_5_cdec_Hy __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("init", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":251 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":251 * * cdef init(self, hypergraph.Hypergraph* hg, unsigned i): * self.hg = hg # <<<<<<<<<<<<<< @@ -12363,7 +12369,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphNode_init(struct __pyx_obj_5_cdec_Hy */ __pyx_v_self->hg = __pyx_v_hg; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":252 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":252 * cdef init(self, hypergraph.Hypergraph* hg, unsigned i): * self.hg = hg * self.node = &hg.nodes_[i] # <<<<<<<<<<<<<< @@ -12372,7 +12378,7 @@ static PyObject *__pyx_f_5_cdec_14HypergraphNode_init(struct __pyx_obj_5_cdec_Hy */ __pyx_v_self->node = (&(__pyx_v_hg->nodes_[__pyx_v_i])); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":253 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":253 * self.hg = hg * self.node = &hg.nodes_[i] * return self # <<<<<<<<<<<<<< @@ -12403,7 +12409,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphNode_8in_edges_1__get__(PyObject *_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":256 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":256 * * property in_edges: * def __get__(self): # <<<<<<<<<<<<<< @@ -12467,7 +12473,7 @@ static PyObject *__pyx_gb_5_cdec_14HypergraphNode_8in_edges_2generator12(__pyx_G __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 256; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":258 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":258 * def __get__(self): * cdef unsigned i * for i in range(self.node.in_edges_.size()): # <<<<<<<<<<<<<< @@ -12478,7 +12484,7 @@ static PyObject *__pyx_gb_5_cdec_14HypergraphNode_8in_edges_2generator12(__pyx_G for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":259 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":259 * cdef unsigned i * for i in range(self.node.in_edges_.size()): * yield HypergraphEdge().init(self.hg, self.node.in_edges_[i]) # <<<<<<<<<<<<<< @@ -12530,7 +12536,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphNode_9out_edges_1__get__(PyObject * return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":262 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":262 * * property out_edges: * def __get__(self): # <<<<<<<<<<<<<< @@ -12594,7 +12600,7 @@ static PyObject *__pyx_gb_5_cdec_14HypergraphNode_9out_edges_2generator13(__pyx_ __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 262; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":264 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":264 * def __get__(self): * cdef unsigned i * for i in range(self.node.out_edges_.size()): # <<<<<<<<<<<<<< @@ -12605,7 +12611,7 @@ static PyObject *__pyx_gb_5_cdec_14HypergraphNode_9out_edges_2generator13(__pyx_ for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":265 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":265 * cdef unsigned i * for i in range(self.node.out_edges_.size()): * yield HypergraphEdge().init(self.hg, self.node.out_edges_[i]) # <<<<<<<<<<<<<< @@ -12656,7 +12662,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphNode_4span_1__get__(PyObject *__pyx return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":268 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":268 * * property span: * def __get__(self): # <<<<<<<<<<<<<< @@ -12674,7 +12680,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode_4span___get__(struct __pyx_obj int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":269 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":269 * property span: * def __get__(self): * return next(self.in_edges).span # <<<<<<<<<<<<<< @@ -12718,7 +12724,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphNode_3cat_1__get__(PyObject *__pyx_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":272 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":272 * * property cat: * def __get__(self): # <<<<<<<<<<<<<< @@ -12736,7 +12742,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode_3cat___get__(struct __pyx_obj_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":273 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":273 * property cat: * def __get__(self): * if self.node.cat_: # <<<<<<<<<<<<<< @@ -12745,7 +12751,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode_3cat___get__(struct __pyx_obj_ */ if (__pyx_v_self->node->cat_) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":274 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":274 * def __get__(self): * if self.node.cat_: * return str(TDConvert(-self.node.cat_).c_str()) # <<<<<<<<<<<<<< @@ -12800,7 +12806,7 @@ static PyObject *__pyx_pw_5_cdec_14HypergraphNode_1__richcmp__(PyObject *__pyx_v return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":276 +/* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":276 * return str(TDConvert(-self.node.cat_).c_str()) * * def __richcmp__(HypergraphNode x, HypergraphNode y, int op): # <<<<<<<<<<<<<< @@ -12818,7 +12824,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode___richcmp__(struct __pyx_obj_5 int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__richcmp__", 0); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":279 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":279 * if op == 2: # == * return x.node == y.node * elif op == 3: # != # <<<<<<<<<<<<<< @@ -12827,7 +12833,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode___richcmp__(struct __pyx_obj_5 */ switch (__pyx_v_op) { - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":277 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":277 * * def __richcmp__(HypergraphNode x, HypergraphNode y, int op): * if op == 2: # == # <<<<<<<<<<<<<< @@ -12836,7 +12842,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode___richcmp__(struct __pyx_obj_5 */ case 2: - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":278 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":278 * def __richcmp__(HypergraphNode x, HypergraphNode y, int op): * if op == 2: # == * return x.node == y.node # <<<<<<<<<<<<<< @@ -12851,7 +12857,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode___richcmp__(struct __pyx_obj_5 goto __pyx_L0; break; - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":279 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":279 * if op == 2: # == * return x.node == y.node * elif op == 3: # != # <<<<<<<<<<<<<< @@ -12860,7 +12866,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode___richcmp__(struct __pyx_obj_5 */ case 3: - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":280 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":280 * return x.node == y.node * elif op == 3: # != * return not (x == y) # <<<<<<<<<<<<<< @@ -12878,7 +12884,7 @@ static PyObject *__pyx_pf_5_cdec_14HypergraphNode___richcmp__(struct __pyx_obj_5 break; } - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":281 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":281 * elif op == 3: # != * return not (x == y) * raise NotImplemented('comparison not implemented for HypergraphNode') # <<<<<<<<<<<<<< @@ -12915,7 +12921,7 @@ static int __pyx_pw_5_cdec_7Lattice_1__cinit__(PyObject *__pyx_v_self, PyObject return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":6 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":6 * cdef lattice.Lattice* lattice * * def __cinit__(self): # <<<<<<<<<<<<<< @@ -12928,7 +12934,7 @@ static int __pyx_pf_5_cdec_7Lattice___cinit__(struct __pyx_obj_5_cdec_Lattice *_ __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__cinit__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":7 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":7 * * def __cinit__(self): * self.lattice = new lattice.Lattice() # <<<<<<<<<<<<<< @@ -12993,7 +12999,7 @@ static int __pyx_pw_5_cdec_7Lattice_3__init__(PyObject *__pyx_v_self, PyObject * return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":9 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":9 * self.lattice = new lattice.Lattice() * * def __init__(self, inp): # <<<<<<<<<<<<<< @@ -13006,9 +13012,9 @@ static int __pyx_pf_5_cdec_7Lattice_2__init__(struct __pyx_obj_5_cdec_Lattice *_ PyObject *__pyx_v_arcs = NULL; int __pyx_r; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; - Py_ssize_t __pyx_t_3; + int __pyx_t_1; + Py_ssize_t __pyx_t_2; + PyObject *__pyx_t_3 = NULL; PyObject *__pyx_t_4 = NULL; PyObject *(*__pyx_t_5)(PyObject *); PyObject *__pyx_t_6 = NULL; @@ -13018,30 +13024,27 @@ static int __pyx_pf_5_cdec_7Lattice_2__init__(struct __pyx_obj_5_cdec_Lattice *_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__init__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":12 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":12 * """Lattice(tuple) -> Lattice from node list. * Lattice(string) -> Lattice from PLF representation.""" * if isinstance(inp, tuple): # <<<<<<<<<<<<<< * self.lattice.resize(len(inp)) * for i, arcs in enumerate(inp): */ - __pyx_t_1 = ((PyObject *)((PyObject*)(&PyTuple_Type))); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_inp, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyTuple_Check(__pyx_v_inp); + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":13 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":13 * Lattice(string) -> Lattice from PLF representation.""" * if isinstance(inp, tuple): * self.lattice.resize(len(inp)) # <<<<<<<<<<<<<< * for i, arcs in enumerate(inp): * self[i] = arcs */ - __pyx_t_3 = PyObject_Length(__pyx_v_inp); if (unlikely(__pyx_t_3 == -1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_v_self->lattice->resize(__pyx_t_3); + __pyx_t_2 = PyObject_Length(__pyx_v_inp); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_v_self->lattice->resize(__pyx_t_2); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":14 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":14 * if isinstance(inp, tuple): * self.lattice.resize(len(inp)) * for i, arcs in enumerate(inp): # <<<<<<<<<<<<<< @@ -13049,29 +13052,29 @@ static int __pyx_pf_5_cdec_7Lattice_2__init__(struct __pyx_obj_5_cdec_Lattice *_ * elif isinstance(inp, basestring): */ __Pyx_INCREF(__pyx_int_0); - __pyx_t_1 = __pyx_int_0; + __pyx_t_3 = __pyx_int_0; if (PyList_CheckExact(__pyx_v_inp) || PyTuple_CheckExact(__pyx_v_inp)) { - __pyx_t_4 = __pyx_v_inp; __Pyx_INCREF(__pyx_t_4); __pyx_t_3 = 0; + __pyx_t_4 = __pyx_v_inp; __Pyx_INCREF(__pyx_t_4); __pyx_t_2 = 0; __pyx_t_5 = NULL; } else { - __pyx_t_3 = -1; __pyx_t_4 = PyObject_GetIter(__pyx_v_inp); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = -1; __pyx_t_4 = PyObject_GetIter(__pyx_v_inp); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __pyx_t_5 = Py_TYPE(__pyx_t_4)->tp_iternext; } for (;;) { if (!__pyx_t_5 && PyList_CheckExact(__pyx_t_4)) { - if (__pyx_t_3 >= PyList_GET_SIZE(__pyx_t_4)) break; + if (__pyx_t_2 >= PyList_GET_SIZE(__pyx_t_4)) break; #if CYTHON_COMPILING_IN_CPYTHON - __pyx_t_6 = PyList_GET_ITEM(__pyx_t_4, __pyx_t_3); __Pyx_INCREF(__pyx_t_6); __pyx_t_3++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PyList_GET_ITEM(__pyx_t_4, __pyx_t_2); __Pyx_INCREF(__pyx_t_6); __pyx_t_2++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #else - __pyx_t_6 = PySequence_ITEM(__pyx_t_4, __pyx_t_3); __pyx_t_3++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PySequence_ITEM(__pyx_t_4, __pyx_t_2); __pyx_t_2++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif } else if (!__pyx_t_5 && PyTuple_CheckExact(__pyx_t_4)) { - if (__pyx_t_3 >= PyTuple_GET_SIZE(__pyx_t_4)) break; + if (__pyx_t_2 >= PyTuple_GET_SIZE(__pyx_t_4)) break; #if CYTHON_COMPILING_IN_CPYTHON - __pyx_t_6 = PyTuple_GET_ITEM(__pyx_t_4, __pyx_t_3); __Pyx_INCREF(__pyx_t_6); __pyx_t_3++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PyTuple_GET_ITEM(__pyx_t_4, __pyx_t_2); __Pyx_INCREF(__pyx_t_6); __pyx_t_2++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #else - __pyx_t_6 = PySequence_ITEM(__pyx_t_4, __pyx_t_3); __pyx_t_3++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PySequence_ITEM(__pyx_t_4, __pyx_t_2); __pyx_t_2++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif } else { __pyx_t_6 = __pyx_t_5(__pyx_t_4); @@ -13087,16 +13090,16 @@ static int __pyx_pf_5_cdec_7Lattice_2__init__(struct __pyx_obj_5_cdec_Lattice *_ __Pyx_XDECREF(__pyx_v_arcs); __pyx_v_arcs = __pyx_t_6; __pyx_t_6 = 0; - __Pyx_INCREF(__pyx_t_1); + __Pyx_INCREF(__pyx_t_3); __Pyx_XDECREF(__pyx_v_i); - __pyx_v_i = __pyx_t_1; - __pyx_t_6 = PyNumber_Add(__pyx_t_1, __pyx_int_1); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_v_i = __pyx_t_3; + __pyx_t_6 = PyNumber_Add(__pyx_t_3, __pyx_int_1); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 14; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_6); - __Pyx_DECREF(__pyx_t_1); - __pyx_t_1 = __pyx_t_6; + __Pyx_DECREF(__pyx_t_3); + __pyx_t_3 = __pyx_t_6; __pyx_t_6 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":15 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":15 * self.lattice.resize(len(inp)) * for i, arcs in enumerate(inp): * self[i] = arcs # <<<<<<<<<<<<<< @@ -13106,58 +13109,55 @@ static int __pyx_pf_5_cdec_7Lattice_2__init__(struct __pyx_obj_5_cdec_Lattice *_ if (PyObject_SetItem(((PyObject *)__pyx_v_self), __pyx_v_i, __pyx_v_arcs) < 0) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 15; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; goto __pyx_L3; } - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":16 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":16 * for i, arcs in enumerate(inp): * self[i] = arcs * elif isinstance(inp, basestring): # <<<<<<<<<<<<<< * lattice.ConvertTextOrPLF(as_str(inp), self.lattice) * else: */ - __pyx_t_1 = __pyx_builtin_basestring; - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = PyObject_IsInstance(__pyx_v_inp, __pyx_t_1); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 16; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyObject_IsInstance(__pyx_v_inp, __pyx_builtin_basestring); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 16; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":17 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":17 * self[i] = arcs * elif isinstance(inp, basestring): * lattice.ConvertTextOrPLF(as_str(inp), self.lattice) # <<<<<<<<<<<<<< * else: * raise TypeError('cannot create lattice from %s' % type(inp)) */ - __pyx_t_1 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_v_inp, NULL)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_t_7 = __pyx_convert_string_from_py_(__pyx_t_1); if (unlikely(PyErr_Occurred())) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __pyx_t_3 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_v_inp, NULL)); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); + __pyx_t_7 = __pyx_convert_string_from_py_(__pyx_t_3); if (unlikely(PyErr_Occurred())) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 17; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; LatticeTools::ConvertTextOrPLF(__pyx_t_7, __pyx_v_self->lattice); goto __pyx_L3; } /*else*/ { - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":19 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":19 * lattice.ConvertTextOrPLF(as_str(inp), self.lattice) * else: * raise TypeError('cannot create lattice from %s' % type(inp)) # <<<<<<<<<<<<<< * * def __dealloc__(self): */ - __pyx_t_1 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_22), ((PyObject *)Py_TYPE(__pyx_v_inp))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 19; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(((PyObject *)__pyx_t_1)); + __pyx_t_3 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_22), ((PyObject *)Py_TYPE(__pyx_v_inp))); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 19; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(((PyObject *)__pyx_t_3)); __pyx_t_4 = PyTuple_New(1); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 19; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); - PyTuple_SET_ITEM(__pyx_t_4, 0, ((PyObject *)__pyx_t_1)); - __Pyx_GIVEREF(((PyObject *)__pyx_t_1)); - __pyx_t_1 = 0; - __pyx_t_1 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 19; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + PyTuple_SET_ITEM(__pyx_t_4, 0, ((PyObject *)__pyx_t_3)); + __Pyx_GIVEREF(((PyObject *)__pyx_t_3)); + __pyx_t_3 = 0; + __pyx_t_3 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 19; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0; - __Pyx_Raise(__pyx_t_1, 0, 0, 0); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_Raise(__pyx_t_3, 0, 0, 0); + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; {__pyx_filename = __pyx_f[4]; __pyx_lineno = 19; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_L3:; @@ -13165,7 +13165,7 @@ static int __pyx_pf_5_cdec_7Lattice_2__init__(struct __pyx_obj_5_cdec_Lattice *_ __pyx_r = 0; goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_4); __Pyx_XDECREF(__pyx_t_6); __Pyx_AddTraceback("_cdec.Lattice.__init__", __pyx_clineno, __pyx_lineno, __pyx_filename); @@ -13186,7 +13186,7 @@ static void __pyx_pw_5_cdec_7Lattice_5__dealloc__(PyObject *__pyx_v_self) { __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":21 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":21 * raise TypeError('cannot create lattice from %s' % type(inp)) * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -13198,7 +13198,7 @@ static void __pyx_pf_5_cdec_7Lattice_4__dealloc__(CYTHON_UNUSED struct __pyx_obj __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":22 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":22 * * def __dealloc__(self): * del self.lattice # <<<<<<<<<<<<<< @@ -13231,7 +13231,7 @@ static PyObject *__pyx_pw_5_cdec_7Lattice_7__getitem__(PyObject *__pyx_v_self, P return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":24 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":24 * del self.lattice * * def __getitem__(self, int index): # <<<<<<<<<<<<<< @@ -13261,7 +13261,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__getitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":25 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":25 * * def __getitem__(self, int index): * if not 0 <= index < len(self): # <<<<<<<<<<<<<< @@ -13276,7 +13276,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L __pyx_t_3 = (!__pyx_t_1); if (__pyx_t_3) { - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":26 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":26 * def __getitem__(self, int index): * if not 0 <= index < len(self): * raise IndexError('lattice index out of range') # <<<<<<<<<<<<<< @@ -13292,7 +13292,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":27 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":27 * if not 0 <= index < len(self): * raise IndexError('lattice index out of range') * arcs = [] # <<<<<<<<<<<<<< @@ -13301,10 +13301,10 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L */ __pyx_t_4 = PyList_New(0); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 27; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); - __pyx_v_arcs = __pyx_t_4; + __pyx_v_arcs = ((PyObject*)__pyx_t_4); __pyx_t_4 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":28 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":28 * raise IndexError('lattice index out of range') * arcs = [] * cdef vector[lattice.LatticeArc] arc_vector = self.lattice[0][index] # <<<<<<<<<<<<<< @@ -13313,7 +13313,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L */ __pyx_v_arc_vector = ((__pyx_v_self->lattice[0])[__pyx_v_index]); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":31 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":31 * cdef lattice.LatticeArc* arc * cdef unsigned i * for i in range(arc_vector.size()): # <<<<<<<<<<<<<< @@ -13324,7 +13324,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L for (__pyx_t_6 = 0; __pyx_t_6 < __pyx_t_5; __pyx_t_6+=1) { __pyx_v_i = __pyx_t_6; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":32 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":32 * cdef unsigned i * for i in range(arc_vector.size()): * arc = &arc_vector[i] # <<<<<<<<<<<<<< @@ -13333,7 +13333,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L */ __pyx_v_arc = (&(__pyx_v_arc_vector[__pyx_v_i])); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":33 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":33 * for i in range(arc_vector.size()): * arc = &arc_vector[i] * label = unicode(TDConvert(arc.label).c_str(), 'utf8') # <<<<<<<<<<<<<< @@ -13357,7 +13357,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L __pyx_v_label = ((PyObject*)__pyx_t_4); __pyx_t_4 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":34 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":34 * arc = &arc_vector[i] * label = unicode(TDConvert(arc.label).c_str(), 'utf8') * arcs.append((label, arc.cost, arc.dist2next)) # <<<<<<<<<<<<<< @@ -13383,7 +13383,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_6__getitem__(struct __pyx_obj_5_cdec_L __Pyx_DECREF(((PyObject *)__pyx_t_8)); __pyx_t_8 = 0; } - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":35 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":35 * label = unicode(TDConvert(arc.label).c_str(), 'utf8') * arcs.append((label, arc.cost, arc.dist2next)) * return tuple(arcs) # <<<<<<<<<<<<<< @@ -13439,7 +13439,7 @@ static int __pyx_pw_5_cdec_7Lattice_9__setitem__(PyObject *__pyx_v_self, PyObjec return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":37 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":37 * return tuple(arcs) * * def __setitem__(self, int index, tuple arcs): # <<<<<<<<<<<<<< @@ -13473,7 +13473,7 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__setitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":38 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":38 * * def __setitem__(self, int index, tuple arcs): * if not 0 <= index < len(self): # <<<<<<<<<<<<<< @@ -13488,7 +13488,7 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice __pyx_t_3 = (!__pyx_t_1); if (__pyx_t_3) { - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":39 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":39 * def __setitem__(self, int index, tuple arcs): * if not 0 <= index < len(self): * raise IndexError('lattice index out of range') # <<<<<<<<<<<<<< @@ -13504,7 +13504,7 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":41 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":41 * raise IndexError('lattice index out of range') * cdef lattice.LatticeArc* arc * for (label, cost, dist2next) in arcs: # <<<<<<<<<<<<<< @@ -13550,8 +13550,11 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice __Pyx_INCREF(__pyx_t_8); #else __pyx_t_6 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_6); __pyx_t_7 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_7)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_7); __pyx_t_8 = PySequence_ITEM(sequence, 2); if (unlikely(!__pyx_t_8)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 41; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_8); #endif __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; } else @@ -13588,7 +13591,7 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice __pyx_v_dist2next = __pyx_t_8; __pyx_t_8 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":42 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":42 * cdef lattice.LatticeArc* arc * for (label, cost, dist2next) in arcs: * label_str = as_str(label) # <<<<<<<<<<<<<< @@ -13601,7 +13604,7 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice __pyx_v_label_str = ((PyObject*)__pyx_t_5); __pyx_t_5 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":43 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":43 * for (label, cost, dist2next) in arcs: * label_str = as_str(label) * arc = new lattice.LatticeArc(TDConvert(label_str), cost, dist2next) # <<<<<<<<<<<<<< @@ -13613,7 +13616,7 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice __pyx_t_13 = __Pyx_PyInt_AsInt(__pyx_v_dist2next); if (unlikely((__pyx_t_13 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 43; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_arc = new LatticeArc(TD::Convert(__pyx_t_11), __pyx_t_12, __pyx_t_13); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":44 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":44 * label_str = as_str(label) * arc = new lattice.LatticeArc(TDConvert(label_str), cost, dist2next) * self.lattice[0][index].push_back(arc[0]) # <<<<<<<<<<<<<< @@ -13622,7 +13625,7 @@ static int __pyx_pf_5_cdec_7Lattice_8__setitem__(struct __pyx_obj_5_cdec_Lattice */ ((__pyx_v_self->lattice[0])[__pyx_v_index]).push_back((__pyx_v_arc[0])); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":45 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":45 * arc = new lattice.LatticeArc(TDConvert(label_str), cost, dist2next) * self.lattice[0][index].push_back(arc[0]) * del arc # <<<<<<<<<<<<<< @@ -13664,7 +13667,7 @@ static Py_ssize_t __pyx_pw_5_cdec_7Lattice_11__len__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":47 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":47 * del arc * * def __len__(self): # <<<<<<<<<<<<<< @@ -13677,7 +13680,7 @@ static Py_ssize_t __pyx_pf_5_cdec_7Lattice_10__len__(struct __pyx_obj_5_cdec_Lat __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__len__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":48 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":48 * * def __len__(self): * return self.lattice.size() # <<<<<<<<<<<<<< @@ -13704,7 +13707,7 @@ static PyObject *__pyx_pw_5_cdec_7Lattice_13__str__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":50 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":50 * return self.lattice.size() * * def __str__(self): # <<<<<<<<<<<<<< @@ -13722,7 +13725,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_12__str__(struct __pyx_obj_5_cdec_Latt int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__str__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":51 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":51 * * def __str__(self): * return str(hypergraph.AsPLF(self.lattice[0], True).c_str()) # <<<<<<<<<<<<<< @@ -13768,7 +13771,7 @@ static PyObject *__pyx_pw_5_cdec_7Lattice_15__unicode__(PyObject *__pyx_v_self, return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":53 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":53 * return str(hypergraph.AsPLF(self.lattice[0], True).c_str()) * * def __unicode__(self): # <<<<<<<<<<<<<< @@ -13786,7 +13789,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_14__unicode__(struct __pyx_obj_5_cdec_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__unicode__", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":54 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":54 * * def __unicode__(self): * return unicode(str(self), 'utf8') # <<<<<<<<<<<<<< @@ -13842,7 +13845,7 @@ static PyObject *__pyx_pw_5_cdec_7Lattice_17__iter__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":56 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":56 * return unicode(str(self), 'utf8') * * def __iter__(self): # <<<<<<<<<<<<<< @@ -13905,7 +13908,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_18generator14(__pyx_GeneratorObject *_ __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":58 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":58 * def __iter__(self): * cdef unsigned i * for i in range(len(self)): # <<<<<<<<<<<<<< @@ -13916,7 +13919,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_18generator14(__pyx_GeneratorObject *_ for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":59 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":59 * cdef unsigned i * for i in range(len(self)): * yield self[i] # <<<<<<<<<<<<<< @@ -13977,7 +13980,7 @@ static PyObject *__pyx_pw_5_cdec_7Lattice_5todot_1lines(PyObject *__pyx_self, CY return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":63 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":63 * def todot(self): * """lattice.todot() -> Representation of the lattice in GraphViz dot format.""" * def lines(): # <<<<<<<<<<<<<< @@ -14054,7 +14057,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":64 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":64 * """lattice.todot() -> Representation of the lattice in GraphViz dot format.""" * def lines(): * yield 'digraph lattice {' # <<<<<<<<<<<<<< @@ -14071,7 +14074,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __pyx_L4_resume_from_yield:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 64; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":65 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":65 * def lines(): * yield 'digraph lattice {' * yield 'rankdir = LR;' # <<<<<<<<<<<<<< @@ -14088,7 +14091,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __pyx_L5_resume_from_yield:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 65; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":66 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":66 * yield 'digraph lattice {' * yield 'rankdir = LR;' * yield 'node [shape=circle];' # <<<<<<<<<<<<<< @@ -14105,7 +14108,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __pyx_L6_resume_from_yield:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 66; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":67 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":67 * yield 'rankdir = LR;' * yield 'node [shape=circle];' * for i in range(len(self)): # <<<<<<<<<<<<<< @@ -14168,7 +14171,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __pyx_cur_scope->__pyx_v_i = __pyx_t_1; __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":68 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":68 * yield 'node [shape=circle];' * for i in range(len(self)): * for label, weight, delta in self[i]: # <<<<<<<<<<<<<< @@ -14239,8 +14242,11 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __Pyx_INCREF(__pyx_t_10); #else __pyx_t_8 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_8)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_8); __pyx_t_9 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_9)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_9); __pyx_t_10 = PySequence_ITEM(sequence, 2); if (unlikely(!__pyx_t_10)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 68; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_10); #endif __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; } else @@ -14283,7 +14289,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __pyx_cur_scope->__pyx_v_delta = __pyx_t_10; __pyx_t_10 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":69 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":69 * for i in range(len(self)): * for label, weight, delta in self[i]: * yield '%d -> %d [label="%s"];' % (i, i+delta, label.replace('"', '\\"')) # <<<<<<<<<<<<<< @@ -14343,7 +14349,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj } __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":70 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":70 * for label, weight, delta in self[i]: * yield '%d -> %d [label="%s"];' % (i, i+delta, label.replace('"', '\\"')) * yield '%d [shape=doublecircle]' % len(self) # <<<<<<<<<<<<<< @@ -14369,7 +14375,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj __pyx_L14_resume_from_yield:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 70; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":71 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":71 * yield '%d -> %d [label="%s"];' % (i, i+delta, label.replace('"', '\\"')) * yield '%d [shape=doublecircle]' % len(self) * yield '}' # <<<<<<<<<<<<<< @@ -14404,7 +14410,7 @@ static PyObject *__pyx_gb_5_cdec_7Lattice_5todot_2generator20(__pyx_GeneratorObj return NULL; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":61 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":61 * yield self[i] * * def todot(self): # <<<<<<<<<<<<<< @@ -14434,19 +14440,19 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_19todot(struct __pyx_obj_5_cdec_Lattic __Pyx_INCREF((PyObject *)__pyx_cur_scope->__pyx_v_self); __Pyx_GIVEREF((PyObject *)__pyx_cur_scope->__pyx_v_self); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":63 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":63 * def todot(self): * """lattice.todot() -> Representation of the lattice in GraphViz dot format.""" * def lines(): # <<<<<<<<<<<<<< * yield 'digraph lattice {' * yield 'rankdir = LR;' */ - __pyx_t_1 = __Pyx_CyFunction_NewEx(&__pyx_mdef_5_cdec_7Lattice_5todot_1lines, 0, ((PyObject*)__pyx_cur_scope), __pyx_n_s___cdec, ((PyObject *)__pyx_k_codeobj_36)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = __Pyx_CyFunction_NewEx(&__pyx_mdef_5_cdec_7Lattice_5todot_1lines, 0, __pyx_n_s_38, ((PyObject*)__pyx_cur_scope), __pyx_n_s___cdec, ((PyObject *)__pyx_k_codeobj_36)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); __pyx_v_lines = __pyx_t_1; __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":72 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":72 * yield '%d [shape=doublecircle]' % len(self) * yield '}' * return '\n'.join(lines()).encode('utf8') # <<<<<<<<<<<<<< @@ -14454,9 +14460,9 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_19todot(struct __pyx_obj_5_cdec_Lattic * def as_hypergraph(self): */ __Pyx_XDECREF(__pyx_r); - __pyx_t_1 = PyObject_GetAttr(((PyObject *)__pyx_kp_s_38), __pyx_n_s__join); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = PyObject_GetAttr(((PyObject *)__pyx_kp_s_39), __pyx_n_s__join); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); - __pyx_t_2 = PyObject_Call(__pyx_v_lines, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = __pyx_pf_5_cdec_7Lattice_5todot_lines(__pyx_v_lines); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); @@ -14470,7 +14476,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_19todot(struct __pyx_obj_5_cdec_Lattic __pyx_t_3 = PyObject_GetAttr(__pyx_t_2, __pyx_n_s__encode); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; - __pyx_t_2 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_k_tuple_39), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_k_tuple_40), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_r = __pyx_t_2; @@ -14505,7 +14511,7 @@ static PyObject *__pyx_pw_5_cdec_7Lattice_22as_hypergraph(PyObject *__pyx_v_self return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":74 +/* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":74 * return '\n'.join(lines()).encode('utf8') * * def as_hypergraph(self): # <<<<<<<<<<<<<< @@ -14526,7 +14532,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_21as_hypergraph(struct __pyx_obj_5_cde int __pyx_clineno = 0; __Pyx_RefNannySetupContext("as_hypergraph", 0); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":76 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":76 * def as_hypergraph(self): * """lattice.as_hypergraph() -> Hypergraph representation of the lattice.""" * cdef Hypergraph result = Hypergraph.__new__(Hypergraph) # <<<<<<<<<<<<<< @@ -14539,7 +14545,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_21as_hypergraph(struct __pyx_obj_5_cde __pyx_v_result = ((struct __pyx_obj_5_cdec_Hypergraph *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":77 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":77 * """lattice.as_hypergraph() -> Hypergraph representation of the lattice.""" * cdef Hypergraph result = Hypergraph.__new__(Hypergraph) * result.hg = new hypergraph.Hypergraph() # <<<<<<<<<<<<<< @@ -14548,7 +14554,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_21as_hypergraph(struct __pyx_obj_5_cde */ __pyx_v_result->hg = new Hypergraph(); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":78 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":78 * cdef Hypergraph result = Hypergraph.__new__(Hypergraph) * result.hg = new hypergraph.Hypergraph() * cdef bytes plf = str(self) # <<<<<<<<<<<<<< @@ -14567,7 +14573,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_21as_hypergraph(struct __pyx_obj_5_cde __pyx_v_plf = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":79 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":79 * result.hg = new hypergraph.Hypergraph() * cdef bytes plf = str(self) * hypergraph.ReadFromPLF(plf, result.hg) # <<<<<<<<<<<<<< @@ -14576,7 +14582,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_21as_hypergraph(struct __pyx_obj_5_cde __pyx_t_3 = __pyx_convert_string_from_py_(((PyObject *)__pyx_v_plf)); if (unlikely(PyErr_Occurred())) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} HypergraphIO::ReadFromPLF(__pyx_t_3, __pyx_v_result->hg); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":80 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":80 * cdef bytes plf = str(self) * hypergraph.ReadFromPLF(plf, result.hg) * return result # <<<<<<<<<<<<<< @@ -14601,7 +14607,7 @@ static PyObject *__pyx_pf_5_cdec_7Lattice_21as_hypergraph(struct __pyx_obj_5_cde return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":3 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":3 * cimport mteval * * cdef SufficientStats as_stats(x, y): # <<<<<<<<<<<<<< @@ -14613,29 +14619,27 @@ static struct __pyx_obj_5_cdec_SufficientStats *__pyx_f_5_cdec_as_stats(PyObject struct __pyx_obj_5_cdec_SufficientStats *__pyx_v_stats = NULL; struct __pyx_obj_5_cdec_SufficientStats *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; int __pyx_t_3; int __pyx_t_4; + EvaluationMetric *__pyx_t_5; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("as_stats", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":4 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":4 * * cdef SufficientStats as_stats(x, y): * if isinstance(x, SufficientStats): # <<<<<<<<<<<<<< * return x * elif x == 0 and isinstance(y, SufficientStats): */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SufficientStats)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_x, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_x, ((PyObject*)__pyx_ptype_5_cdec_SufficientStats)); + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":5 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":5 * cdef SufficientStats as_stats(x, y): * if isinstance(x, SufficientStats): * return x # <<<<<<<<<<<<<< @@ -14650,40 +14654,37 @@ static struct __pyx_obj_5_cdec_SufficientStats *__pyx_f_5_cdec_as_stats(PyObject goto __pyx_L3; } - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":6 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":6 * if isinstance(x, SufficientStats): * return x * elif x == 0 and isinstance(y, SufficientStats): # <<<<<<<<<<<<<< * stats = SufficientStats() * stats.stats = new mteval.SufficientStats() */ - __pyx_t_1 = PyObject_RichCompare(__pyx_v_x, __pyx_int_0, Py_EQ); __Pyx_XGOTREF(__pyx_t_1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_t_1); if (unlikely(__pyx_t_2 < 0)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SufficientStats)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_3 = __Pyx_TypeCheck(__pyx_v_y, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __pyx_t_2 = PyObject_RichCompare(__pyx_v_x, __pyx_int_0, Py_EQ); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + if (__pyx_t_1) { + __pyx_t_3 = __Pyx_TypeCheck(__pyx_v_y, ((PyObject*)__pyx_ptype_5_cdec_SufficientStats)); __pyx_t_4 = __pyx_t_3; } else { - __pyx_t_4 = __pyx_t_2; + __pyx_t_4 = __pyx_t_1; } if (__pyx_t_4) { - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":7 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":7 * return x * elif x == 0 and isinstance(y, SufficientStats): * stats = SufficientStats() # <<<<<<<<<<<<<< * stats.stats = new mteval.SufficientStats() * stats.metric = (<SufficientStats> y).metric */ - __pyx_t_1 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SufficientStats)), ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 7; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_v_stats = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_t_2 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SufficientStats)), ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 7; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_v_stats = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_t_2); + __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":8 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":8 * elif x == 0 and isinstance(y, SufficientStats): * stats = SufficientStats() * stats.stats = new mteval.SufficientStats() # <<<<<<<<<<<<<< @@ -14692,16 +14693,17 @@ static struct __pyx_obj_5_cdec_SufficientStats *__pyx_f_5_cdec_as_stats(PyObject */ __pyx_v_stats->stats = new SufficientStats(); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":9 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":9 * stats = SufficientStats() * stats.stats = new mteval.SufficientStats() * stats.metric = (<SufficientStats> y).metric # <<<<<<<<<<<<<< * return stats * */ - __pyx_v_stats->metric = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_v_y)->metric; + __pyx_t_5 = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_v_y)->metric; + __pyx_v_stats->metric = __pyx_t_5; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":10 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":10 * stats.stats = new mteval.SufficientStats() * stats.metric = (<SufficientStats> y).metric * return stats # <<<<<<<<<<<<<< @@ -14719,7 +14721,7 @@ static struct __pyx_obj_5_cdec_SufficientStats *__pyx_f_5_cdec_as_stats(PyObject __pyx_r = ((struct __pyx_obj_5_cdec_SufficientStats *)Py_None); __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_AddTraceback("_cdec.as_stats", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = 0; __pyx_L0:; @@ -14740,7 +14742,7 @@ static PyObject *__pyx_pw_5_cdec_9Candidate_5words_1__get__(PyObject *__pyx_v_se return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":17 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":17 * * property words: * def __get__(self): # <<<<<<<<<<<<<< @@ -14759,7 +14761,7 @@ static PyObject *__pyx_pf_5_cdec_9Candidate_5words___get__(struct __pyx_obj_5_cd int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":18 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":18 * property words: * def __get__(self): * return unicode(GetString(self.candidate.ewords).c_str(), encoding='utf8') # <<<<<<<<<<<<<< @@ -14810,7 +14812,7 @@ static PyObject *__pyx_pw_5_cdec_9Candidate_4fmap_1__get__(PyObject *__pyx_v_sel return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":21 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":21 * * property fmap: * def __get__(self): # <<<<<<<<<<<<<< @@ -14828,7 +14830,7 @@ static PyObject *__pyx_pf_5_cdec_9Candidate_4fmap___get__(struct __pyx_obj_5_cde int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":22 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":22 * property fmap: * def __get__(self): * cdef SparseVector fmap = SparseVector.__new__(SparseVector) # <<<<<<<<<<<<<< @@ -14841,7 +14843,7 @@ static PyObject *__pyx_pf_5_cdec_9Candidate_4fmap___get__(struct __pyx_obj_5_cde __pyx_v_fmap = ((struct __pyx_obj_5_cdec_SparseVector *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":23 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":23 * def __get__(self): * cdef SparseVector fmap = SparseVector.__new__(SparseVector) * fmap.vector = new FastSparseVector[weight_t](self.candidate.fmap) # <<<<<<<<<<<<<< @@ -14850,7 +14852,7 @@ static PyObject *__pyx_pf_5_cdec_9Candidate_4fmap___get__(struct __pyx_obj_5_cde */ __pyx_v_fmap->vector = new FastSparseVector<weight_t>(__pyx_v_self->candidate->fmap); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":24 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":24 * cdef SparseVector fmap = SparseVector.__new__(SparseVector) * fmap.vector = new FastSparseVector[weight_t](self.candidate.fmap) * return fmap # <<<<<<<<<<<<<< @@ -14886,7 +14888,7 @@ static PyObject *__pyx_pw_5_cdec_9Candidate_5score_1__get__(PyObject *__pyx_v_se return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":14 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":14 * cdef class Candidate: * cdef mteval.const_Candidate* candidate * cdef public float score # <<<<<<<<<<<<<< @@ -14962,7 +14964,7 @@ static void __pyx_pw_5_cdec_15SufficientStats_1__dealloc__(PyObject *__pyx_v_sel __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":30 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":30 * cdef mteval.EvaluationMetric* metric * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -14974,7 +14976,7 @@ static void __pyx_pf_5_cdec_15SufficientStats___dealloc__(CYTHON_UNUSED struct _ __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":31 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":31 * * def __dealloc__(self): * del self.stats # <<<<<<<<<<<<<< @@ -14997,7 +14999,7 @@ static PyObject *__pyx_pw_5_cdec_15SufficientStats_5score_1__get__(PyObject *__p return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":34 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":34 * * property score: * def __get__(self): # <<<<<<<<<<<<<< @@ -15014,7 +15016,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_5score___get__(struct __pyx_o int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":35 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":35 * property score: * def __get__(self): * return self.metric.ComputeScore(self.stats[0]) # <<<<<<<<<<<<<< @@ -15051,7 +15053,7 @@ static PyObject *__pyx_pw_5_cdec_15SufficientStats_6detail_1__get__(PyObject *__ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":38 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":38 * * property detail: * def __get__(self): # <<<<<<<<<<<<<< @@ -15069,7 +15071,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_6detail___get__(struct __pyx_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__get__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":39 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":39 * property detail: * def __get__(self): * return str(self.metric.DetailedScore(self.stats[0]).c_str()) # <<<<<<<<<<<<<< @@ -15115,7 +15117,7 @@ static Py_ssize_t __pyx_pw_5_cdec_15SufficientStats_3__len__(PyObject *__pyx_v_s return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":41 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":41 * return str(self.metric.DetailedScore(self.stats[0]).c_str()) * * def __len__(self): # <<<<<<<<<<<<<< @@ -15128,7 +15130,7 @@ static Py_ssize_t __pyx_pf_5_cdec_15SufficientStats_2__len__(struct __pyx_obj_5_ __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__len__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":42 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":42 * * def __len__(self): * return self.stats.size() # <<<<<<<<<<<<<< @@ -15156,7 +15158,7 @@ static PyObject *__pyx_pw_5_cdec_15SufficientStats_5__iter__(PyObject *__pyx_v_s return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":44 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":44 * return self.stats.size() * * def __iter__(self): # <<<<<<<<<<<<<< @@ -15220,7 +15222,7 @@ static PyObject *__pyx_gb_5_cdec_15SufficientStats_6generator15(__pyx_GeneratorO __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 44; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":45 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":45 * * def __iter__(self): * for i in range(len(self)): # <<<<<<<<<<<<<< @@ -15279,7 +15281,7 @@ static PyObject *__pyx_gb_5_cdec_15SufficientStats_6generator15(__pyx_GeneratorO __pyx_cur_scope->__pyx_v_i = __pyx_t_2; __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":46 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":46 * def __iter__(self): * for i in range(len(self)): * yield self[i] # <<<<<<<<<<<<<< @@ -15343,7 +15345,7 @@ static PyObject *__pyx_pw_5_cdec_15SufficientStats_8__getitem__(PyObject *__pyx_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":48 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":48 * yield self[i] * * def __getitem__(self, int index): # <<<<<<<<<<<<<< @@ -15363,7 +15365,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_7__getitem__(struct __pyx_obj int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__getitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":49 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":49 * * def __getitem__(self, int index): * if not 0 <= index < len(self): # <<<<<<<<<<<<<< @@ -15378,14 +15380,14 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_7__getitem__(struct __pyx_obj __pyx_t_3 = (!__pyx_t_1); if (__pyx_t_3) { - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":50 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":50 * def __getitem__(self, int index): * if not 0 <= index < len(self): * raise IndexError('sufficient stats vector index out of range') # <<<<<<<<<<<<<< * return self.stats[0][index] * */ - __pyx_t_4 = PyObject_Call(__pyx_builtin_IndexError, ((PyObject *)__pyx_k_tuple_41), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 50; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_4 = PyObject_Call(__pyx_builtin_IndexError, ((PyObject *)__pyx_k_tuple_42), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 50; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_Raise(__pyx_t_4, 0, 0, 0); __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0; @@ -15394,7 +15396,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_7__getitem__(struct __pyx_obj } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":51 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":51 * if not 0 <= index < len(self): * raise IndexError('sufficient stats vector index out of range') * return self.stats[0][index] # <<<<<<<<<<<<<< @@ -15436,7 +15438,7 @@ static PyObject *__pyx_pw_5_cdec_15SufficientStats_10__iadd__(PyObject *__pyx_v_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":53 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":53 * return self.stats[0][index] * * def __iadd__(SufficientStats self, SufficientStats other): # <<<<<<<<<<<<<< @@ -15449,7 +15451,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_9__iadd__(struct __pyx_obj_5_ __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__iadd__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":54 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":54 * * def __iadd__(SufficientStats self, SufficientStats other): * self.stats[0] += other.stats[0] # <<<<<<<<<<<<<< @@ -15458,7 +15460,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_9__iadd__(struct __pyx_obj_5_ */ (__pyx_v_self->stats[0]) += (__pyx_v_other->stats[0]); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":55 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":55 * def __iadd__(SufficientStats self, SufficientStats other): * self.stats[0] += other.stats[0] * return self # <<<<<<<<<<<<<< @@ -15488,7 +15490,7 @@ static PyObject *__pyx_pw_5_cdec_15SufficientStats_12__add__(PyObject *__pyx_v_x return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":57 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":57 * return self * * def __add__(x, y): # <<<<<<<<<<<<<< @@ -15503,12 +15505,13 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_11__add__(PyObject *__pyx_v_x PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; + EvaluationMetric *__pyx_t_2; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__add__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":58 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":58 * * def __add__(x, y): * cdef SufficientStats sx = as_stats(x, y) # <<<<<<<<<<<<<< @@ -15520,7 +15523,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_11__add__(PyObject *__pyx_v_x __pyx_v_sx = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":59 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":59 * def __add__(x, y): * cdef SufficientStats sx = as_stats(x, y) * cdef SufficientStats sy = as_stats(y, x) # <<<<<<<<<<<<<< @@ -15532,7 +15535,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_11__add__(PyObject *__pyx_v_x __pyx_v_sy = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":60 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":60 * cdef SufficientStats sx = as_stats(x, y) * cdef SufficientStats sy = as_stats(y, x) * cdef SufficientStats result = SufficientStats() # <<<<<<<<<<<<<< @@ -15544,7 +15547,7 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_11__add__(PyObject *__pyx_v_x __pyx_v_result = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":61 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":61 * cdef SufficientStats sy = as_stats(y, x) * cdef SufficientStats result = SufficientStats() * result.stats = new mteval.SufficientStats(mteval.add(sx.stats[0], sy.stats[0])) # <<<<<<<<<<<<<< @@ -15553,16 +15556,17 @@ static PyObject *__pyx_pf_5_cdec_15SufficientStats_11__add__(PyObject *__pyx_v_x */ __pyx_v_result->stats = new SufficientStats(operator+((__pyx_v_sx->stats[0]), (__pyx_v_sy->stats[0]))); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":62 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":62 * cdef SufficientStats result = SufficientStats() * result.stats = new mteval.SufficientStats(mteval.add(sx.stats[0], sy.stats[0])) * result.metric = sx.metric # <<<<<<<<<<<<<< * return result * */ - __pyx_v_result->metric = __pyx_v_sx->metric; + __pyx_t_2 = __pyx_v_sx->metric; + __pyx_v_result->metric = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":63 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":63 * result.stats = new mteval.SufficientStats(mteval.add(sx.stats[0], sy.stats[0])) * result.metric = sx.metric * return result # <<<<<<<<<<<<<< @@ -15641,7 +15645,7 @@ static int __pyx_pw_5_cdec_12CandidateSet_1__cinit__(PyObject *__pyx_v_self, PyO return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":70 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":70 * cdef mteval.CandidateSet* cs * * def __cinit__(self, SegmentEvaluator evaluator): # <<<<<<<<<<<<<< @@ -15652,9 +15656,10 @@ static int __pyx_pw_5_cdec_12CandidateSet_1__cinit__(PyObject *__pyx_v_self, PyO static int __pyx_pf_5_cdec_12CandidateSet___cinit__(struct __pyx_obj_5_cdec_CandidateSet *__pyx_v_self, struct __pyx_obj_5_cdec_SegmentEvaluator *__pyx_v_evaluator) { int __pyx_r; __Pyx_RefNannyDeclarations + EvaluationMetric *__pyx_t_1; __Pyx_RefNannySetupContext("__cinit__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":71 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":71 * * def __cinit__(self, SegmentEvaluator evaluator): * self.scorer = new shared_ptr[mteval.SegmentEvaluator](evaluator.scorer[0]) # <<<<<<<<<<<<<< @@ -15663,16 +15668,17 @@ static int __pyx_pf_5_cdec_12CandidateSet___cinit__(struct __pyx_obj_5_cdec_Cand */ __pyx_v_self->scorer = new boost::shared_ptr<SegmentEvaluator>((__pyx_v_evaluator->scorer[0])); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":72 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":72 * def __cinit__(self, SegmentEvaluator evaluator): * self.scorer = new shared_ptr[mteval.SegmentEvaluator](evaluator.scorer[0]) * self.metric = evaluator.metric # <<<<<<<<<<<<<< * self.cs = new mteval.CandidateSet() * */ - __pyx_v_self->metric = __pyx_v_evaluator->metric; + __pyx_t_1 = __pyx_v_evaluator->metric; + __pyx_v_self->metric = __pyx_t_1; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":73 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":73 * self.scorer = new shared_ptr[mteval.SegmentEvaluator](evaluator.scorer[0]) * self.metric = evaluator.metric * self.cs = new mteval.CandidateSet() # <<<<<<<<<<<<<< @@ -15695,7 +15701,7 @@ static void __pyx_pw_5_cdec_12CandidateSet_3__dealloc__(PyObject *__pyx_v_self) __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":75 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":75 * self.cs = new mteval.CandidateSet() * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -15707,7 +15713,7 @@ static void __pyx_pf_5_cdec_12CandidateSet_2__dealloc__(CYTHON_UNUSED struct __p __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":76 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":76 * * def __dealloc__(self): * del self.scorer # <<<<<<<<<<<<<< @@ -15716,7 +15722,7 @@ static void __pyx_pf_5_cdec_12CandidateSet_2__dealloc__(CYTHON_UNUSED struct __p */ delete __pyx_v_self->scorer; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":77 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":77 * def __dealloc__(self): * del self.scorer * del self.cs # <<<<<<<<<<<<<< @@ -15739,7 +15745,7 @@ static Py_ssize_t __pyx_pw_5_cdec_12CandidateSet_5__len__(PyObject *__pyx_v_self return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":79 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":79 * del self.cs * * def __len__(self): # <<<<<<<<<<<<<< @@ -15752,7 +15758,7 @@ static Py_ssize_t __pyx_pf_5_cdec_12CandidateSet_4__len__(struct __pyx_obj_5_cde __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__len__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":80 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":80 * * def __len__(self): * return self.cs.size() # <<<<<<<<<<<<<< @@ -15789,7 +15795,7 @@ static PyObject *__pyx_pw_5_cdec_12CandidateSet_7__getitem__(PyObject *__pyx_v_s return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":82 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":82 * return self.cs.size() * * def __getitem__(self,int k): # <<<<<<<<<<<<<< @@ -15809,7 +15815,7 @@ static PyObject *__pyx_pf_5_cdec_12CandidateSet_6__getitem__(struct __pyx_obj_5_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__getitem__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":83 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":83 * * def __getitem__(self,int k): * if not 0 <= k < self.cs.size(): # <<<<<<<<<<<<<< @@ -15823,14 +15829,14 @@ static PyObject *__pyx_pf_5_cdec_12CandidateSet_6__getitem__(struct __pyx_obj_5_ __pyx_t_2 = (!__pyx_t_1); if (__pyx_t_2) { - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":84 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":84 * def __getitem__(self,int k): * if not 0 <= k < self.cs.size(): * raise IndexError('candidate set index out of range') # <<<<<<<<<<<<<< * cdef Candidate candidate = Candidate() * candidate.candidate = &self.cs[0][k] */ - __pyx_t_3 = PyObject_Call(__pyx_builtin_IndexError, ((PyObject *)__pyx_k_tuple_43), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 84; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyObject_Call(__pyx_builtin_IndexError, ((PyObject *)__pyx_k_tuple_44), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 84; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_Raise(__pyx_t_3, 0, 0, 0); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; @@ -15839,7 +15845,7 @@ static PyObject *__pyx_pf_5_cdec_12CandidateSet_6__getitem__(struct __pyx_obj_5_ } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":85 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":85 * if not 0 <= k < self.cs.size(): * raise IndexError('candidate set index out of range') * cdef Candidate candidate = Candidate() # <<<<<<<<<<<<<< @@ -15851,7 +15857,7 @@ static PyObject *__pyx_pf_5_cdec_12CandidateSet_6__getitem__(struct __pyx_obj_5_ __pyx_v_candidate = ((struct __pyx_obj_5_cdec_Candidate *)__pyx_t_3); __pyx_t_3 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":86 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":86 * raise IndexError('candidate set index out of range') * cdef Candidate candidate = Candidate() * candidate.candidate = &self.cs[0][k] # <<<<<<<<<<<<<< @@ -15860,7 +15866,7 @@ static PyObject *__pyx_pf_5_cdec_12CandidateSet_6__getitem__(struct __pyx_obj_5_ */ __pyx_v_candidate->candidate = (&((__pyx_v_self->cs[0])[__pyx_v_k])); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":87 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":87 * cdef Candidate candidate = Candidate() * candidate.candidate = &self.cs[0][k] * candidate.score = self.metric.ComputeScore(self.cs[0][k].eval_feats) # <<<<<<<<<<<<<< @@ -15869,7 +15875,7 @@ static PyObject *__pyx_pf_5_cdec_12CandidateSet_6__getitem__(struct __pyx_obj_5_ */ __pyx_v_candidate->score = __pyx_v_self->metric->ComputeScore(((__pyx_v_self->cs[0])[__pyx_v_k]).eval_feats); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":88 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":88 * candidate.candidate = &self.cs[0][k] * candidate.score = self.metric.ComputeScore(self.cs[0][k].eval_feats) * return candidate # <<<<<<<<<<<<<< @@ -15906,7 +15912,7 @@ static PyObject *__pyx_pw_5_cdec_12CandidateSet_9__iter__(PyObject *__pyx_v_self return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":90 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":90 * return candidate * * def __iter__(self): # <<<<<<<<<<<<<< @@ -15969,7 +15975,7 @@ static PyObject *__pyx_gb_5_cdec_12CandidateSet_10generator16(__pyx_GeneratorObj __pyx_L3_first_run:; if (unlikely(!__pyx_sent_value)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 90; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":92 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":92 * def __iter__(self): * cdef unsigned i * for i in range(len(self)): # <<<<<<<<<<<<<< @@ -15980,7 +15986,7 @@ static PyObject *__pyx_gb_5_cdec_12CandidateSet_10generator16(__pyx_GeneratorObj for (__pyx_t_2 = 0; __pyx_t_2 < __pyx_t_1; __pyx_t_2+=1) { __pyx_cur_scope->__pyx_v_i = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":93 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":93 * cdef unsigned i * for i in range(len(self)): * yield self[i] # <<<<<<<<<<<<<< @@ -16078,7 +16084,7 @@ static PyObject *__pyx_pw_5_cdec_12CandidateSet_12add_kbest(PyObject *__pyx_v_se return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":95 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":95 * yield self[i] * * def add_kbest(self, Hypergraph hypergraph, unsigned k): # <<<<<<<<<<<<<< @@ -16091,7 +16097,7 @@ static PyObject *__pyx_pf_5_cdec_12CandidateSet_11add_kbest(struct __pyx_obj_5_c __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("add_kbest", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":98 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":98 * """cs.add_kbest(Hypergraph hypergraph, int k) -> Extract K-best hypotheses * from the hypergraph and add them to the candidate set.""" * self.cs.AddKBestCandidates(hypergraph.hg[0], k, self.scorer.get()) # <<<<<<<<<<<<<< @@ -16115,7 +16121,7 @@ static void __pyx_pw_5_cdec_16SegmentEvaluator_1__dealloc__(PyObject *__pyx_v_se __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":104 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":104 * cdef mteval.EvaluationMetric* metric * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -16127,7 +16133,7 @@ static void __pyx_pf_5_cdec_16SegmentEvaluator___dealloc__(CYTHON_UNUSED struct __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":105 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":105 * * def __dealloc__(self): * del self.scorer # <<<<<<<<<<<<<< @@ -16151,7 +16157,7 @@ static PyObject *__pyx_pw_5_cdec_16SegmentEvaluator_3evaluate(PyObject *__pyx_v_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":107 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":107 * del self.scorer * * def evaluate(self, sentence): # <<<<<<<<<<<<<< @@ -16165,14 +16171,15 @@ static PyObject *__pyx_pf_5_cdec_16SegmentEvaluator_2evaluate(struct __pyx_obj_5 PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations PyObject *__pyx_t_1 = NULL; - PyObject *__pyx_t_2 = NULL; - std::string __pyx_t_3; + EvaluationMetric *__pyx_t_2; + PyObject *__pyx_t_3 = NULL; + std::string __pyx_t_4; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("evaluate", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":110 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":110 * """se.evaluate(sentence) -> SufficientStats for the given hypothesis.""" * cdef vector[WordID] hyp * cdef SufficientStats sf = SufficientStats() # <<<<<<<<<<<<<< @@ -16184,16 +16191,17 @@ static PyObject *__pyx_pf_5_cdec_16SegmentEvaluator_2evaluate(struct __pyx_obj_5 __pyx_v_sf = ((struct __pyx_obj_5_cdec_SufficientStats *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":111 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":111 * cdef vector[WordID] hyp * cdef SufficientStats sf = SufficientStats() * sf.metric = self.metric # <<<<<<<<<<<<<< * sf.stats = new mteval.SufficientStats() * ConvertSentence(as_str(sentence.strip()), &hyp) */ - __pyx_v_sf->metric = __pyx_v_self->metric; + __pyx_t_2 = __pyx_v_self->metric; + __pyx_v_sf->metric = __pyx_t_2; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":112 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":112 * cdef SufficientStats sf = SufficientStats() * sf.metric = self.metric * sf.stats = new mteval.SufficientStats() # <<<<<<<<<<<<<< @@ -16202,7 +16210,7 @@ static PyObject *__pyx_pf_5_cdec_16SegmentEvaluator_2evaluate(struct __pyx_obj_5 */ __pyx_v_sf->stats = new SufficientStats(); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":113 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":113 * sf.metric = self.metric * sf.stats = new mteval.SufficientStats() * ConvertSentence(as_str(sentence.strip()), &hyp) # <<<<<<<<<<<<<< @@ -16211,17 +16219,17 @@ static PyObject *__pyx_pf_5_cdec_16SegmentEvaluator_2evaluate(struct __pyx_obj_5 */ __pyx_t_1 = PyObject_GetAttr(__pyx_v_sentence, __pyx_n_s__strip); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 113; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); - __pyx_t_2 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 113; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_2); + __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 113; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - __pyx_t_1 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_2, NULL)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 113; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_3, NULL)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 113; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; - __pyx_t_3 = __pyx_convert_string_from_py_(__pyx_t_1); if (unlikely(PyErr_Occurred())) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 113; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; + __pyx_t_4 = __pyx_convert_string_from_py_(__pyx_t_1); if (unlikely(PyErr_Occurred())) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 113; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - TD::ConvertSentence(__pyx_t_3, (&__pyx_v_hyp)); + TD::ConvertSentence(__pyx_t_4, (&__pyx_v_hyp)); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":114 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":114 * sf.stats = new mteval.SufficientStats() * ConvertSentence(as_str(sentence.strip()), &hyp) * self.scorer.get().Evaluate(hyp, sf.stats) # <<<<<<<<<<<<<< @@ -16230,7 +16238,7 @@ static PyObject *__pyx_pf_5_cdec_16SegmentEvaluator_2evaluate(struct __pyx_obj_5 */ __pyx_v_self->scorer->get()->Evaluate(__pyx_v_hyp, __pyx_v_sf->stats); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":115 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":115 * ConvertSentence(as_str(sentence.strip()), &hyp) * self.scorer.get().Evaluate(hyp, sf.stats) * return sf # <<<<<<<<<<<<<< @@ -16246,7 +16254,7 @@ static PyObject *__pyx_pf_5_cdec_16SegmentEvaluator_2evaluate(struct __pyx_obj_5 goto __pyx_L0; __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_1); - __Pyx_XDECREF(__pyx_t_2); + __Pyx_XDECREF(__pyx_t_3); __Pyx_AddTraceback("_cdec.SegmentEvaluator.evaluate", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; __pyx_L0:; @@ -16268,7 +16276,7 @@ static PyObject *__pyx_pw_5_cdec_16SegmentEvaluator_5candidate_set(PyObject *__p return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":117 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":117 * return sf * * def candidate_set(self): # <<<<<<<<<<<<<< @@ -16286,7 +16294,7 @@ static PyObject *__pyx_pf_5_cdec_16SegmentEvaluator_4candidate_set(struct __pyx_ int __pyx_clineno = 0; __Pyx_RefNannySetupContext("candidate_set", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":119 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":119 * def candidate_set(self): * """se.candidate_set() -> Candidate set using this segment evaluator for scoring.""" * return CandidateSet(self) # <<<<<<<<<<<<<< @@ -16330,7 +16338,7 @@ static int __pyx_pw_5_cdec_6Scorer_1__cinit__(PyObject *__pyx_v_self, PyObject * static PyObject **__pyx_pyargnames[] = {&__pyx_n_s__name,0}; PyObject* values[1] = {0}; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":125 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":125 * cdef mteval.EvaluationMetric* metric * * def __cinit__(self, bytes name=None): # <<<<<<<<<<<<<< @@ -16395,7 +16403,7 @@ static int __pyx_pf_5_cdec_6Scorer___cinit__(struct __pyx_obj_5_cdec_Scorer *__p int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__cinit__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":126 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":126 * * def __cinit__(self, bytes name=None): * if name: # <<<<<<<<<<<<<< @@ -16405,7 +16413,7 @@ static int __pyx_pf_5_cdec_6Scorer___cinit__(struct __pyx_obj_5_cdec_Scorer *__p __pyx_t_1 = (((PyObject *)__pyx_v_name) != Py_None) && (PyBytes_GET_SIZE(((PyObject *)__pyx_v_name)) != 0); if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":127 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":127 * def __cinit__(self, bytes name=None): * if name: * self.name = new string(name) # <<<<<<<<<<<<<< @@ -16413,10 +16421,15 @@ static int __pyx_pf_5_cdec_6Scorer___cinit__(struct __pyx_obj_5_cdec_Scorer *__p * */ __pyx_t_2 = PyBytes_AsString(((PyObject *)__pyx_v_name)); if (unlikely((!__pyx_t_2) && PyErr_Occurred())) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - try {__pyx_t_3 = new std::string(__pyx_t_2);} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[5]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_3 = new std::string(__pyx_t_2); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[5]; __pyx_lineno = 127; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_v_self->name = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":128 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":128 * if name: * self.name = new string(name) * self.metric = mteval.MetricInstance(self.name[0]) # <<<<<<<<<<<<<< @@ -16447,7 +16460,7 @@ static void __pyx_pw_5_cdec_6Scorer_3__dealloc__(PyObject *__pyx_v_self) { __Pyx_RefNannyFinishContext(); } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":130 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":130 * self.metric = mteval.MetricInstance(self.name[0]) * * def __dealloc__(self): # <<<<<<<<<<<<<< @@ -16459,7 +16472,7 @@ static void __pyx_pf_5_cdec_6Scorer_2__dealloc__(CYTHON_UNUSED struct __pyx_obj_ __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("__dealloc__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":131 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":131 * * def __dealloc__(self): * del self.name # <<<<<<<<<<<<<< @@ -16518,7 +16531,7 @@ static PyObject *__pyx_pw_5_cdec_6Scorer_5__call__(PyObject *__pyx_v_self, PyObj return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":133 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":133 * del self.name * * def __call__(self, refs): # <<<<<<<<<<<<<< @@ -16533,8 +16546,8 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score struct __pyx_obj_5_cdec_SegmentEvaluator *__pyx_v_evaluator = 0; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; std::vector<std::vector<WordID> > *__pyx_t_3; Py_ssize_t __pyx_t_4; PyObject *(*__pyx_t_5)(PyObject *); @@ -16542,55 +16555,58 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score std::vector<WordID> *__pyx_t_7; PyObject *__pyx_t_8 = NULL; std::string __pyx_t_9; + EvaluationMetric *__pyx_t_10; int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__call__", 0); __Pyx_INCREF(__pyx_v_refs); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":134 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":134 * * def __call__(self, refs): * if isinstance(refs, basestring): # <<<<<<<<<<<<<< * refs = [refs] * cdef vector[vector[WordID]]* refsv = new vector[vector[WordID]]() */ - __pyx_t_1 = __pyx_builtin_basestring; - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = PyObject_IsInstance(__pyx_v_refs, __pyx_t_1); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyObject_IsInstance(__pyx_v_refs, __pyx_builtin_basestring); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 134; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (__pyx_t_1) { - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":135 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":135 * def __call__(self, refs): * if isinstance(refs, basestring): * refs = [refs] # <<<<<<<<<<<<<< * cdef vector[vector[WordID]]* refsv = new vector[vector[WordID]]() * cdef vector[WordID]* refv */ - __pyx_t_1 = PyList_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 135; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + __pyx_t_2 = PyList_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 135; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_INCREF(__pyx_v_refs); - PyList_SET_ITEM(__pyx_t_1, 0, __pyx_v_refs); + PyList_SET_ITEM(__pyx_t_2, 0, __pyx_v_refs); __Pyx_GIVEREF(__pyx_v_refs); __Pyx_DECREF(__pyx_v_refs); - __pyx_v_refs = ((PyObject *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_v_refs = ((PyObject *)__pyx_t_2); + __pyx_t_2 = 0; goto __pyx_L3; } __pyx_L3:; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":136 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":136 * if isinstance(refs, basestring): * refs = [refs] * cdef vector[vector[WordID]]* refsv = new vector[vector[WordID]]() # <<<<<<<<<<<<<< * cdef vector[WordID]* refv * for ref in refs: */ - try {__pyx_t_3 = new std::vector<std::vector<WordID> >();} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[5]; __pyx_lineno = 136; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_3 = new std::vector<std::vector<WordID> >(); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[5]; __pyx_lineno = 136; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_v_refsv = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":138 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":138 * cdef vector[vector[WordID]]* refsv = new vector[vector[WordID]]() * cdef vector[WordID]* refv * for ref in refs: # <<<<<<<<<<<<<< @@ -16598,30 +16614,30 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score * ConvertSentence(as_str(ref.strip()), refv) */ if (PyList_CheckExact(__pyx_v_refs) || PyTuple_CheckExact(__pyx_v_refs)) { - __pyx_t_1 = __pyx_v_refs; __Pyx_INCREF(__pyx_t_1); __pyx_t_4 = 0; + __pyx_t_2 = __pyx_v_refs; __Pyx_INCREF(__pyx_t_2); __pyx_t_4 = 0; __pyx_t_5 = NULL; } else { - __pyx_t_4 = -1; __pyx_t_1 = PyObject_GetIter(__pyx_v_refs); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_t_5 = Py_TYPE(__pyx_t_1)->tp_iternext; + __pyx_t_4 = -1; __pyx_t_2 = PyObject_GetIter(__pyx_v_refs); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_t_5 = Py_TYPE(__pyx_t_2)->tp_iternext; } for (;;) { - if (!__pyx_t_5 && PyList_CheckExact(__pyx_t_1)) { - if (__pyx_t_4 >= PyList_GET_SIZE(__pyx_t_1)) break; + if (!__pyx_t_5 && PyList_CheckExact(__pyx_t_2)) { + if (__pyx_t_4 >= PyList_GET_SIZE(__pyx_t_2)) break; #if CYTHON_COMPILING_IN_CPYTHON - __pyx_t_6 = PyList_GET_ITEM(__pyx_t_1, __pyx_t_4); __Pyx_INCREF(__pyx_t_6); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PyList_GET_ITEM(__pyx_t_2, __pyx_t_4); __Pyx_INCREF(__pyx_t_6); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #else - __pyx_t_6 = PySequence_ITEM(__pyx_t_1, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PySequence_ITEM(__pyx_t_2, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif - } else if (!__pyx_t_5 && PyTuple_CheckExact(__pyx_t_1)) { - if (__pyx_t_4 >= PyTuple_GET_SIZE(__pyx_t_1)) break; + } else if (!__pyx_t_5 && PyTuple_CheckExact(__pyx_t_2)) { + if (__pyx_t_4 >= PyTuple_GET_SIZE(__pyx_t_2)) break; #if CYTHON_COMPILING_IN_CPYTHON - __pyx_t_6 = PyTuple_GET_ITEM(__pyx_t_1, __pyx_t_4); __Pyx_INCREF(__pyx_t_6); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PyTuple_GET_ITEM(__pyx_t_2, __pyx_t_4); __Pyx_INCREF(__pyx_t_6); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #else - __pyx_t_6 = PySequence_ITEM(__pyx_t_1, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PySequence_ITEM(__pyx_t_2, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 138; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif } else { - __pyx_t_6 = __pyx_t_5(__pyx_t_1); + __pyx_t_6 = __pyx_t_5(__pyx_t_2); if (unlikely(!__pyx_t_6)) { if (PyErr_Occurred()) { if (likely(PyErr_ExceptionMatches(PyExc_StopIteration))) PyErr_Clear(); @@ -16635,17 +16651,22 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score __pyx_v_ref = __pyx_t_6; __pyx_t_6 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":139 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":139 * cdef vector[WordID]* refv * for ref in refs: * refv = new vector[WordID]() # <<<<<<<<<<<<<< * ConvertSentence(as_str(ref.strip()), refv) * refsv.push_back(refv[0]) */ - try {__pyx_t_7 = new std::vector<WordID>();} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[5]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_7 = new std::vector<WordID>(); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[5]; __pyx_lineno = 139; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_v_refv = __pyx_t_7; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":140 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":140 * for ref in refs: * refv = new vector[WordID]() * ConvertSentence(as_str(ref.strip()), refv) # <<<<<<<<<<<<<< @@ -16664,7 +16685,7 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; TD::ConvertSentence(__pyx_t_9, __pyx_v_refv); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":141 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":141 * refv = new vector[WordID]() * ConvertSentence(as_str(ref.strip()), refv) * refsv.push_back(refv[0]) # <<<<<<<<<<<<<< @@ -16673,7 +16694,7 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score */ __pyx_v_refsv->push_back((__pyx_v_refv[0])); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":142 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":142 * ConvertSentence(as_str(ref.strip()), refv) * refsv.push_back(refv[0]) * del refv # <<<<<<<<<<<<<< @@ -16682,30 +16703,31 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score */ delete __pyx_v_refv; } - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":144 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":144 * del refv * cdef unsigned i * cdef SegmentEvaluator evaluator = SegmentEvaluator() # <<<<<<<<<<<<<< * evaluator.metric = self.metric * evaluator.scorer = new shared_ptr[mteval.SegmentEvaluator]( */ - __pyx_t_1 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SegmentEvaluator)), ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 144; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_v_evaluator = ((struct __pyx_obj_5_cdec_SegmentEvaluator *)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_t_2 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SegmentEvaluator)), ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 144; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_v_evaluator = ((struct __pyx_obj_5_cdec_SegmentEvaluator *)__pyx_t_2); + __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":145 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":145 * cdef unsigned i * cdef SegmentEvaluator evaluator = SegmentEvaluator() * evaluator.metric = self.metric # <<<<<<<<<<<<<< * evaluator.scorer = new shared_ptr[mteval.SegmentEvaluator]( * self.metric.CreateSegmentEvaluator(refsv[0])) */ - __pyx_v_evaluator->metric = __pyx_v_self->metric; + __pyx_t_10 = __pyx_v_self->metric; + __pyx_v_evaluator->metric = __pyx_t_10; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":146 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":146 * cdef SegmentEvaluator evaluator = SegmentEvaluator() * evaluator.metric = self.metric * evaluator.scorer = new shared_ptr[mteval.SegmentEvaluator]( # <<<<<<<<<<<<<< @@ -16714,7 +16736,7 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score */ __pyx_v_evaluator->scorer = new boost::shared_ptr<SegmentEvaluator>(__pyx_v_self->metric->CreateSegmentEvaluator((__pyx_v_refsv[0]))); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":148 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":148 * evaluator.scorer = new shared_ptr[mteval.SegmentEvaluator]( * self.metric.CreateSegmentEvaluator(refsv[0])) * del refsv # in theory should not delete but store in SegmentEvaluator # <<<<<<<<<<<<<< @@ -16723,7 +16745,7 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score */ delete __pyx_v_refsv; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":149 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":149 * self.metric.CreateSegmentEvaluator(refsv[0])) * del refsv # in theory should not delete but store in SegmentEvaluator * return evaluator # <<<<<<<<<<<<<< @@ -16738,7 +16760,7 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_4__call__(struct __pyx_obj_5_cdec_Score __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_6); __Pyx_XDECREF(__pyx_t_8); __Pyx_AddTraceback("_cdec.Scorer.__call__", __pyx_clineno, __pyx_lineno, __pyx_filename); @@ -16763,7 +16785,7 @@ static PyObject *__pyx_pw_5_cdec_6Scorer_7__str__(PyObject *__pyx_v_self) { return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":151 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":151 * return evaluator * * def __str__(self): # <<<<<<<<<<<<<< @@ -16781,7 +16803,7 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_6__str__(struct __pyx_obj_5_cdec_Scorer int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__str__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":152 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":152 * * def __str__(self): * return str(self.name.c_str()) # <<<<<<<<<<<<<< @@ -16816,7 +16838,7 @@ static PyObject *__pyx_pf_5_cdec_6Scorer_6__str__(struct __pyx_obj_5_cdec_Scorer return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":154 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":154 * return str(self.name.c_str()) * * cdef float _compute_score(void* metric_, mteval.SufficientStats* stats): # <<<<<<<<<<<<<< @@ -16842,7 +16864,7 @@ static float __pyx_f_5_cdec__compute_score(void *__pyx_v_metric_, SufficientStat int __pyx_clineno = 0; __Pyx_RefNannySetupContext("_compute_score", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":155 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":155 * * cdef float _compute_score(void* metric_, mteval.SufficientStats* stats): * cdef Metric metric = <Metric> metric_ # <<<<<<<<<<<<<< @@ -16852,7 +16874,7 @@ static float __pyx_f_5_cdec__compute_score(void *__pyx_v_metric_, SufficientStat __Pyx_INCREF(((PyObject *)((struct __pyx_obj_5_cdec_Metric *)__pyx_v_metric_))); __pyx_v_metric = ((struct __pyx_obj_5_cdec_Metric *)__pyx_v_metric_); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":156 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":156 * cdef float _compute_score(void* metric_, mteval.SufficientStats* stats): * cdef Metric metric = <Metric> metric_ * cdef list ss = [] # <<<<<<<<<<<<<< @@ -16861,10 +16883,10 @@ static float __pyx_f_5_cdec__compute_score(void *__pyx_v_metric_, SufficientStat */ __pyx_t_1 = PyList_New(0); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 156; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); - __pyx_v_ss = __pyx_t_1; + __pyx_v_ss = ((PyObject*)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":158 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":158 * cdef list ss = [] * cdef unsigned i * for i in range(stats.size()): # <<<<<<<<<<<<<< @@ -16875,7 +16897,7 @@ static float __pyx_f_5_cdec__compute_score(void *__pyx_v_metric_, SufficientStat for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_2; __pyx_t_3+=1) { __pyx_v_i = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":159 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":159 * cdef unsigned i * for i in range(stats.size()): * ss.append(stats[0][i]) # <<<<<<<<<<<<<< @@ -16888,7 +16910,7 @@ static float __pyx_f_5_cdec__compute_score(void *__pyx_v_metric_, SufficientStat __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; } - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":160 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":160 * for i in range(stats.size()): * ss.append(stats[0][i]) * return metric.score(ss) # <<<<<<<<<<<<<< @@ -16926,7 +16948,7 @@ static float __pyx_f_5_cdec__compute_score(void *__pyx_v_metric_, SufficientStat return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":162 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":162 * return metric.score(ss) * * cdef void _compute_sufficient_stats(void* metric_, # <<<<<<<<<<<<<< @@ -16953,7 +16975,7 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: int __pyx_clineno = 0; __Pyx_RefNannySetupContext("_compute_sufficient_stats", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":166 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":166 * vector[string]* refs, * mteval.SufficientStats* out): * cdef Metric metric = <Metric> metric_ # <<<<<<<<<<<<<< @@ -16963,7 +16985,7 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: __Pyx_INCREF(((PyObject *)((struct __pyx_obj_5_cdec_Metric *)__pyx_v_metric_))); __pyx_v_metric = ((struct __pyx_obj_5_cdec_Metric *)__pyx_v_metric_); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":167 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":167 * mteval.SufficientStats* out): * cdef Metric metric = <Metric> metric_ * cdef list refs_ = [] # <<<<<<<<<<<<<< @@ -16972,10 +16994,10 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: */ __pyx_t_1 = PyList_New(0); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 167; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_1); - __pyx_v_refs_ = __pyx_t_1; + __pyx_v_refs_ = ((PyObject*)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":169 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":169 * cdef list refs_ = [] * cdef unsigned i * for i in range(refs.size()): # <<<<<<<<<<<<<< @@ -16986,7 +17008,7 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_2; __pyx_t_3+=1) { __pyx_v_i = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":170 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":170 * cdef unsigned i * for i in range(refs.size()): * refs_.append(str(refs[0][i].c_str())) # <<<<<<<<<<<<<< @@ -17007,7 +17029,7 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; } - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":171 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":171 * for i in range(refs.size()): * refs_.append(str(refs[0][i].c_str())) * cdef list ss = metric.evaluate(str(hyp.c_str()), refs_) # <<<<<<<<<<<<<< @@ -17042,7 +17064,7 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: __pyx_v_ss = ((PyObject*)__pyx_t_4); __pyx_t_4 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":172 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":172 * refs_.append(str(refs[0][i].c_str())) * cdef list ss = metric.evaluate(str(hyp.c_str()), refs_) * out.fields.resize(len(ss)) # <<<<<<<<<<<<<< @@ -17056,7 +17078,7 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: __pyx_t_7 = PyList_GET_SIZE(((PyObject *)__pyx_v_ss)); if (unlikely(__pyx_t_7 == -1)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 172; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_out->fields.resize(__pyx_t_7); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":173 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":173 * cdef list ss = metric.evaluate(str(hyp.c_str()), refs_) * out.fields.resize(len(ss)) * for i in range(len(ss)): # <<<<<<<<<<<<<< @@ -17071,7 +17093,7 @@ static void __pyx_f_5_cdec__compute_sufficient_stats(void *__pyx_v_metric_, std: for (__pyx_t_3 = 0; __pyx_t_3 < __pyx_t_7; __pyx_t_3+=1) { __pyx_v_i = __pyx_t_3; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":174 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":174 * out.fields.resize(len(ss)) * for i in range(len(ss)): * out.fields[i] = ss[i] # <<<<<<<<<<<<<< @@ -17116,7 +17138,7 @@ static int __pyx_pw_5_cdec_6Metric_1__cinit__(PyObject *__pyx_v_self, PyObject * return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":178 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":178 * cdef class Metric: * cdef Scorer scorer * def __cinit__(self): # <<<<<<<<<<<<<< @@ -17137,7 +17159,7 @@ static int __pyx_pf_5_cdec_6Metric___cinit__(struct __pyx_obj_5_cdec_Metric *__p int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__cinit__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":179 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":179 * cdef Scorer scorer * def __cinit__(self): * self.scorer = Scorer() # <<<<<<<<<<<<<< @@ -17152,7 +17174,7 @@ static int __pyx_pf_5_cdec_6Metric___cinit__(struct __pyx_obj_5_cdec_Metric *__p __pyx_v_self->scorer = ((struct __pyx_obj_5_cdec_Scorer *)__pyx_t_1); __pyx_t_1 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":180 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":180 * def __cinit__(self): * self.scorer = Scorer() * cdef bytes class_name = self.__class__.__name__ # <<<<<<<<<<<<<< @@ -17168,7 +17190,7 @@ static int __pyx_pf_5_cdec_6Metric___cinit__(struct __pyx_obj_5_cdec_Metric *__p __pyx_v_class_name = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":181 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":181 * self.scorer = Scorer() * cdef bytes class_name = self.__class__.__name__ * self.scorer.name = new string(class_name) # <<<<<<<<<<<<<< @@ -17176,10 +17198,15 @@ static int __pyx_pf_5_cdec_6Metric___cinit__(struct __pyx_obj_5_cdec_Metric *__p * <void*> self, _compute_sufficient_stats, _compute_score) */ __pyx_t_3 = PyBytes_AsString(((PyObject *)__pyx_v_class_name)); if (unlikely((!__pyx_t_3) && PyErr_Occurred())) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 181; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - try {__pyx_t_4 = new std::string(__pyx_t_3);} catch(...) {__Pyx_CppExn2PyErr(); {__pyx_filename = __pyx_f[5]; __pyx_lineno = 181; __pyx_clineno = __LINE__; goto __pyx_L1_error;}} + try { + __pyx_t_4 = new std::string(__pyx_t_3); + } catch(...) { + __Pyx_CppExn2PyErr(); + {__pyx_filename = __pyx_f[5]; __pyx_lineno = 181; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } __pyx_v_self->scorer->name = __pyx_t_4; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":182 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":182 * cdef bytes class_name = self.__class__.__name__ * self.scorer.name = new string(class_name) * self.scorer.metric = mteval.PyMetricInstance(self.scorer.name[0], # <<<<<<<<<<<<<< @@ -17248,7 +17275,7 @@ static PyObject *__pyx_pw_5_cdec_6Metric_3__call__(PyObject *__pyx_v_self, PyObj return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":185 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":185 * <void*> self, _compute_sufficient_stats, _compute_score) * * def __call__(self, refs): # <<<<<<<<<<<<<< @@ -17266,7 +17293,7 @@ static PyObject *__pyx_pf_5_cdec_6Metric_2__call__(struct __pyx_obj_5_cdec_Metri int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__call__", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":186 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":186 * * def __call__(self, refs): * return self.scorer(refs) # <<<<<<<<<<<<<< @@ -17310,7 +17337,7 @@ static PyObject *__pyx_pw_5_cdec_6Metric_5score(PyObject *__pyx_v_stats, CYTHON_ return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":188 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":188 * return self.scorer(refs) * * def score(SufficientStats stats): # <<<<<<<<<<<<<< @@ -17323,7 +17350,7 @@ static PyObject *__pyx_pf_5_cdec_6Metric_4score(CYTHON_UNUSED struct __pyx_obj_5 __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("score", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":189 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":189 * * def score(SufficientStats stats): * return 0 # <<<<<<<<<<<<<< @@ -17398,7 +17425,7 @@ static PyObject *__pyx_pw_5_cdec_6Metric_7evaluate(PyObject *__pyx_v_self, PyObj return __pyx_r; } -/* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":191 +/* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":191 * return 0 * * def evaluate(self, hyp, refs): # <<<<<<<<<<<<<< @@ -17415,7 +17442,7 @@ static PyObject *__pyx_pf_5_cdec_6Metric_6evaluate(CYTHON_UNUSED struct __pyx_ob int __pyx_clineno = 0; __Pyx_RefNannySetupContext("evaluate", 0); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":192 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":192 * * def evaluate(self, hyp, refs): * return [] # <<<<<<<<<<<<<< @@ -17650,7 +17677,9 @@ static PyObject *__pyx_gb_5_cdec_6generator17(__pyx_GeneratorObject *__pyx_gener __Pyx_INCREF(__pyx_t_6); #else __pyx_t_5 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 33; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_5); __pyx_t_6 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 33; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_6); #endif __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; } else @@ -17693,10 +17722,7 @@ static PyObject *__pyx_gb_5_cdec_6generator17(__pyx_GeneratorObject *__pyx_gener * for name, info in value.items(): * yield key, '%s %s' % (name, info) */ - __pyx_t_2 = ((PyObject *)((PyObject*)(&PyDict_Type))); - __Pyx_INCREF(__pyx_t_2); - __pyx_t_9 = __Pyx_TypeCheck(__pyx_cur_scope->__pyx_v_value, __pyx_t_2); - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_9 = PyDict_Check(__pyx_cur_scope->__pyx_v_value); if (__pyx_t_9) { /* "_cdec.pyx":35 @@ -17770,7 +17796,9 @@ static PyObject *__pyx_gb_5_cdec_6generator17(__pyx_GeneratorObject *__pyx_gener __Pyx_INCREF(__pyx_t_7); #else __pyx_t_5 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 35; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_5); __pyx_t_7 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_7)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 35; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_7); #endif __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; } else @@ -17821,7 +17849,7 @@ static PyObject *__pyx_gb_5_cdec_6generator17(__pyx_GeneratorObject *__pyx_gener __Pyx_INCREF(__pyx_cur_scope->__pyx_v_info); PyTuple_SET_ITEM(__pyx_t_6, 1, __pyx_cur_scope->__pyx_v_info); __Pyx_GIVEREF(__pyx_cur_scope->__pyx_v_info); - __pyx_t_7 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_44), ((PyObject *)__pyx_t_6)); if (unlikely(!__pyx_t_7)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 36; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_7 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_45), ((PyObject *)__pyx_t_6)); if (unlikely(!__pyx_t_7)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 36; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_7)); __Pyx_DECREF(((PyObject *)__pyx_t_6)); __pyx_t_6 = 0; __pyx_t_6 = PyTuple_New(2); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 36; __pyx_clineno = __LINE__; goto __pyx_L1_error;} @@ -17871,10 +17899,7 @@ static PyObject *__pyx_gb_5_cdec_6generator17(__pyx_GeneratorObject *__pyx_gener * for name in value: * yield key, name */ - __pyx_t_2 = ((PyObject *)((PyObject*)(&PyList_Type))); - __Pyx_INCREF(__pyx_t_2); - __pyx_t_9 = __Pyx_TypeCheck(__pyx_cur_scope->__pyx_v_value, __pyx_t_2); - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_9 = PyList_Check(__pyx_cur_scope->__pyx_v_value); if (__pyx_t_9) { /* "_cdec.pyx":38 @@ -18174,9 +18199,9 @@ static PyObject *__pyx_gb_5_cdec_7Decoder_8__init___2generator21(__pyx_Generator if (unlikely(!__pyx_cur_scope->__pyx_outer_scope->__pyx_v_config)) { __Pyx_RaiseClosureNameError("config"); {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); - __Pyx_INCREF(__pyx_cur_scope->__pyx_outer_scope->__pyx_v_config); - PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_cur_scope->__pyx_outer_scope->__pyx_v_config); - __Pyx_GIVEREF(__pyx_cur_scope->__pyx_outer_scope->__pyx_v_config); + __Pyx_INCREF(((PyObject *)__pyx_cur_scope->__pyx_outer_scope->__pyx_v_config)); + PyTuple_SET_ITEM(__pyx_t_2, 0, ((PyObject *)__pyx_cur_scope->__pyx_outer_scope->__pyx_v_config)); + __Pyx_GIVEREF(((PyObject *)__pyx_cur_scope->__pyx_outer_scope->__pyx_v_config)); __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; @@ -18221,7 +18246,7 @@ static PyObject *__pyx_gb_5_cdec_7Decoder_8__init___2generator21(__pyx_Generator __Pyx_GIVEREF(__pyx_t_3); __pyx_cur_scope->__pyx_v_kv = __pyx_t_3; __pyx_t_3 = 0; - __pyx_t_3 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_45), __pyx_cur_scope->__pyx_v_kv); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_46), __pyx_cur_scope->__pyx_v_kv); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_3)); __pyx_r = ((PyObject *)__pyx_t_3); __pyx_t_3 = 0; @@ -18311,13 +18336,10 @@ static int __pyx_pf_5_cdec_7Decoder___init__(struct __pyx_obj_5_cdec_Decoder *__ * if formalism not in ('scfg', 'fst', 'lextrans', 'pb', * 'csplit', 'tagger', 'lexalign'): */ - __pyx_t_2 = PyObject_GetAttr(__pyx_cur_scope->__pyx_v_config, __pyx_n_s__get); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 52; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = __Pyx_PyDict_GetItemDefault(((PyObject *)__pyx_cur_scope->__pyx_v_config), ((PyObject *)__pyx_n_s__formalism), Py_None); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 52; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); - __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_k_tuple_46), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 52; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; - __pyx_v_formalism = __pyx_t_3; - __pyx_t_3 = 0; + __pyx_v_formalism = __pyx_t_2; + __pyx_t_2 = 0; /* "_cdec.pyx":53 * if config_str is None: @@ -18327,59 +18349,59 @@ static int __pyx_pf_5_cdec_7Decoder___init__(struct __pyx_obj_5_cdec_Decoder *__ * raise InvalidConfig('formalism "%s" unknown' % formalism) */ __Pyx_INCREF(__pyx_v_formalism); - __pyx_t_3 = __pyx_v_formalism; - __pyx_t_2 = PyObject_RichCompare(__pyx_t_3, ((PyObject *)__pyx_n_s__scfg), Py_NE); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_1 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_2 = __pyx_v_formalism; + __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, ((PyObject *)__pyx_n_s__scfg), Py_NE); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely((__pyx_t_1 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; if (((int)__pyx_t_1)) { - __pyx_t_2 = PyObject_RichCompare(__pyx_t_3, ((PyObject *)__pyx_n_s__fst), Py_NE); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_4 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, ((PyObject *)__pyx_n_s__fst), Py_NE); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely((__pyx_t_4 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_5 = ((int)__pyx_t_4); } else { __pyx_t_5 = ((int)__pyx_t_1); } if (__pyx_t_5) { - __pyx_t_2 = PyObject_RichCompare(__pyx_t_3, ((PyObject *)__pyx_n_s__lextrans), Py_NE); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_1 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, ((PyObject *)__pyx_n_s__lextrans), Py_NE); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely((__pyx_t_1 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_4 = ((int)__pyx_t_1); } else { __pyx_t_4 = __pyx_t_5; } if (__pyx_t_4) { - __pyx_t_2 = PyObject_RichCompare(__pyx_t_3, ((PyObject *)__pyx_n_s__pb), Py_NE); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_5 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_5 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, ((PyObject *)__pyx_n_s__pb), Py_NE); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_5 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely((__pyx_t_5 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_1 = ((int)__pyx_t_5); } else { __pyx_t_1 = __pyx_t_4; } if (__pyx_t_1) { - __pyx_t_2 = PyObject_RichCompare(__pyx_t_3, ((PyObject *)__pyx_n_s__csplit), Py_NE); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_4 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, ((PyObject *)__pyx_n_s__csplit), Py_NE); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely((__pyx_t_4 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_5 = ((int)__pyx_t_4); } else { __pyx_t_5 = __pyx_t_1; } if (__pyx_t_5) { - __pyx_t_2 = PyObject_RichCompare(__pyx_t_3, ((PyObject *)__pyx_n_s__tagger), Py_NE); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_1 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, ((PyObject *)__pyx_n_s__tagger), Py_NE); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely((__pyx_t_1 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_4 = ((int)__pyx_t_1); } else { __pyx_t_4 = __pyx_t_5; } if (__pyx_t_4) { - __pyx_t_2 = PyObject_RichCompare(__pyx_t_3, ((PyObject *)__pyx_n_s__lexalign), Py_NE); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_5 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_5 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, ((PyObject *)__pyx_n_s__lexalign), Py_NE); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_5 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely((__pyx_t_5 == (int)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 53; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __pyx_t_1 = ((int)__pyx_t_5); } else { __pyx_t_1 = __pyx_t_4; } - __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __pyx_t_4 = __pyx_t_1; if (__pyx_t_4) { @@ -18390,21 +18412,21 @@ static int __pyx_pf_5_cdec_7Decoder___init__(struct __pyx_obj_5_cdec_Decoder *__ * config_str = '\n'.join('%s = %s' % kv for kv in _make_config(config)) * cdef istringstream* config_stream = new istringstream(config_str) */ - __pyx_t_3 = __Pyx_GetName(__pyx_m, __pyx_n_s__InvalidConfig); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_3); - __pyx_t_2 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_47), __pyx_v_formalism); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(((PyObject *)__pyx_t_2)); + __pyx_t_2 = __Pyx_GetName(__pyx_m, __pyx_n_s__InvalidConfig); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_t_3 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_47), __pyx_v_formalism); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(((PyObject *)__pyx_t_3)); __pyx_t_6 = PyTuple_New(1); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_6); - PyTuple_SET_ITEM(__pyx_t_6, 0, ((PyObject *)__pyx_t_2)); - __Pyx_GIVEREF(((PyObject *)__pyx_t_2)); - __pyx_t_2 = 0; - __pyx_t_2 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_6), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_2); - __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - __Pyx_DECREF(((PyObject *)__pyx_t_6)); __pyx_t_6 = 0; - __Pyx_Raise(__pyx_t_2, 0, 0, 0); + PyTuple_SET_ITEM(__pyx_t_6, 0, ((PyObject *)__pyx_t_3)); + __Pyx_GIVEREF(((PyObject *)__pyx_t_3)); + __pyx_t_3 = 0; + __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_6), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __Pyx_DECREF(((PyObject *)__pyx_t_6)); __pyx_t_6 = 0; + __Pyx_Raise(__pyx_t_3, 0, 0, 0); + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 55; __pyx_clineno = __LINE__; goto __pyx_L1_error;} goto __pyx_L4; } @@ -18417,19 +18439,19 @@ static int __pyx_pf_5_cdec_7Decoder___init__(struct __pyx_obj_5_cdec_Decoder *__ * cdef istringstream* config_stream = new istringstream(config_str) * self.dec = new decoder.Decoder(config_stream) */ - __pyx_t_2 = PyObject_GetAttr(((PyObject *)__pyx_kp_s_38), __pyx_n_s__join); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_2); + __pyx_t_3 = PyObject_GetAttr(((PyObject *)__pyx_kp_s_39), __pyx_n_s__join); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); __pyx_t_6 = __pyx_pf_5_cdec_7Decoder_8__init___genexpr(((PyObject*)__pyx_cur_scope)); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_6); - __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_3); - PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_6); + __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_6); __Pyx_GIVEREF(__pyx_t_6); __pyx_t_6 = 0; - __pyx_t_6 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_6 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 56; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_6); - __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; - __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; + __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; __Pyx_DECREF(__pyx_v_config_str); __pyx_v_config_str = __pyx_t_6; __pyx_t_6 = 0; @@ -18615,8 +18637,8 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De PyObject *__pyx_v_fval = NULL; int __pyx_r; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; Py_ssize_t __pyx_t_4; PyObject *(*__pyx_t_5)(PyObject *); @@ -18636,11 +18658,8 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De * self.weights.vector[0] = (<DenseVector> weights).vector[0] * elif isinstance(weights, SparseVector): */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_DenseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_weights, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_weights, ((PyObject*)__pyx_ptype_5_cdec_DenseVector)); + if (__pyx_t_1) { /* "_cdec.pyx":73 * def __set__(self, weights): @@ -18660,11 +18679,8 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De * self.weights.vector.clear() * ((<SparseVector> weights).vector[0]).init_vector(self.weights.vector) */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_SparseVector)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_weights, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_weights, ((PyObject*)__pyx_ptype_5_cdec_SparseVector)); + if (__pyx_t_1) { /* "_cdec.pyx":75 * self.weights.vector[0] = (<DenseVector> weights).vector[0] @@ -18693,11 +18709,8 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De * self.weights.vector.clear() * for fname, fval in weights.items(): */ - __pyx_t_1 = ((PyObject *)((PyObject*)(&PyDict_Type))); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_weights, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyDict_Check(__pyx_v_weights); + if (__pyx_t_1) { /* "_cdec.pyx":78 * ((<SparseVector> weights).vector[0]).init_vector(self.weights.vector) @@ -18715,37 +18728,37 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De * self.weights[fname] = fval * else: */ - __pyx_t_1 = PyObject_GetAttr(__pyx_v_weights, __pyx_n_s__items); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = PyObject_GetAttr(__pyx_v_weights, __pyx_n_s__items); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; if (PyList_CheckExact(__pyx_t_3) || PyTuple_CheckExact(__pyx_t_3)) { - __pyx_t_1 = __pyx_t_3; __Pyx_INCREF(__pyx_t_1); __pyx_t_4 = 0; + __pyx_t_2 = __pyx_t_3; __Pyx_INCREF(__pyx_t_2); __pyx_t_4 = 0; __pyx_t_5 = NULL; } else { - __pyx_t_4 = -1; __pyx_t_1 = PyObject_GetIter(__pyx_t_3); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_t_5 = Py_TYPE(__pyx_t_1)->tp_iternext; + __pyx_t_4 = -1; __pyx_t_2 = PyObject_GetIter(__pyx_t_3); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_t_5 = Py_TYPE(__pyx_t_2)->tp_iternext; } __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; for (;;) { - if (!__pyx_t_5 && PyList_CheckExact(__pyx_t_1)) { - if (__pyx_t_4 >= PyList_GET_SIZE(__pyx_t_1)) break; + if (!__pyx_t_5 && PyList_CheckExact(__pyx_t_2)) { + if (__pyx_t_4 >= PyList_GET_SIZE(__pyx_t_2)) break; #if CYTHON_COMPILING_IN_CPYTHON - __pyx_t_3 = PyList_GET_ITEM(__pyx_t_1, __pyx_t_4); __Pyx_INCREF(__pyx_t_3); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyList_GET_ITEM(__pyx_t_2, __pyx_t_4); __Pyx_INCREF(__pyx_t_3); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #else - __pyx_t_3 = PySequence_ITEM(__pyx_t_1, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PySequence_ITEM(__pyx_t_2, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif - } else if (!__pyx_t_5 && PyTuple_CheckExact(__pyx_t_1)) { - if (__pyx_t_4 >= PyTuple_GET_SIZE(__pyx_t_1)) break; + } else if (!__pyx_t_5 && PyTuple_CheckExact(__pyx_t_2)) { + if (__pyx_t_4 >= PyTuple_GET_SIZE(__pyx_t_2)) break; #if CYTHON_COMPILING_IN_CPYTHON - __pyx_t_3 = PyTuple_GET_ITEM(__pyx_t_1, __pyx_t_4); __Pyx_INCREF(__pyx_t_3); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyTuple_GET_ITEM(__pyx_t_2, __pyx_t_4); __Pyx_INCREF(__pyx_t_3); __pyx_t_4++; if (unlikely(0 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #else - __pyx_t_3 = PySequence_ITEM(__pyx_t_1, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PySequence_ITEM(__pyx_t_2, __pyx_t_4); __pyx_t_4++; if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #endif } else { - __pyx_t_3 = __pyx_t_5(__pyx_t_1); + __pyx_t_3 = __pyx_t_5(__pyx_t_2); if (unlikely(!__pyx_t_3)) { if (PyErr_Occurred()) { if (likely(PyErr_ExceptionMatches(PyExc_StopIteration))) PyErr_Clear(); @@ -18779,7 +18792,9 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De __Pyx_INCREF(__pyx_t_7); #else __pyx_t_6 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_6); __pyx_t_7 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_7)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 79; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_7); #endif __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; } else @@ -18820,7 +18835,7 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De */ if (PyObject_SetItem(((PyObject *)__pyx_v_self->weights), __pyx_v_fname, __pyx_v_fval) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 80; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; goto __pyx_L3; } /*else*/ { @@ -18832,18 +18847,18 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De * * property formalism: */ - __pyx_t_1 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_48), ((PyObject *)Py_TYPE(__pyx_v_weights))); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 82; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(((PyObject *)__pyx_t_1)); + __pyx_t_2 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_48), ((PyObject *)Py_TYPE(__pyx_v_weights))); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 82; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(((PyObject *)__pyx_t_2)); __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 82; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_t_1)); - __Pyx_GIVEREF(((PyObject *)__pyx_t_1)); - __pyx_t_1 = 0; - __pyx_t_1 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 82; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_t_2)); + __Pyx_GIVEREF(((PyObject *)__pyx_t_2)); + __pyx_t_2 = 0; + __pyx_t_2 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 82; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; - __Pyx_Raise(__pyx_t_1, 0, 0, 0); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_Raise(__pyx_t_2, 0, 0, 0); + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 82; __pyx_clineno = __LINE__; goto __pyx_L1_error;} } __pyx_L3:; @@ -18851,7 +18866,7 @@ static int __pyx_pf_5_cdec_7Decoder_7weights_2__set__(struct __pyx_obj_5_cdec_De __pyx_r = 0; goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_XDECREF(__pyx_t_6); __Pyx_XDECREF(__pyx_t_7); @@ -19129,7 +19144,9 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_4read_weights(struct __pyx_obj_5_cdec_ __Pyx_INCREF(__pyx_t_11); #else __pyx_t_1 = PySequence_ITEM(sequence, 0); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 94; __pyx_clineno = __LINE__; goto __pyx_L7_error;} + __Pyx_GOTREF(__pyx_t_1); __pyx_t_11 = PySequence_ITEM(sequence, 1); if (unlikely(!__pyx_t_11)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 94; __pyx_clineno = __LINE__; goto __pyx_L7_error;} + __Pyx_GOTREF(__pyx_t_11); #endif __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; } else @@ -19368,8 +19385,8 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec struct __pyx_obj_5_cdec_Hypergraph *__pyx_v_hg = 0; PyObject *__pyx_r = NULL; __Pyx_RefNannyDeclarations - PyObject *__pyx_t_1 = NULL; - int __pyx_t_2; + int __pyx_t_1; + PyObject *__pyx_t_2 = NULL; PyObject *__pyx_t_3 = NULL; std::string __pyx_t_4; int __pyx_lineno = 0; @@ -19384,11 +19401,8 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * input_str = as_str(sentence.strip()) * elif isinstance(sentence, Lattice): */ - __pyx_t_1 = __pyx_builtin_basestring; - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = PyObject_IsInstance(__pyx_v_sentence, __pyx_t_1); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyObject_IsInstance(__pyx_v_sentence, __pyx_builtin_basestring); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 101; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (__pyx_t_1) { /* "_cdec.pyx":102 * cdef bytes input_str @@ -19397,16 +19411,16 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * elif isinstance(sentence, Lattice): * input_str = str(sentence) # PLF format */ - __pyx_t_1 = PyObject_GetAttr(__pyx_v_sentence, __pyx_n_s__strip); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 102; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 102; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = PyObject_GetAttr(__pyx_v_sentence, __pyx_n_s__strip); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 102; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 102; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - __pyx_t_1 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_3, NULL)); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 102; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; + __pyx_t_2 = ((PyObject *)__pyx_f_5_cdec_as_str(__pyx_t_3, NULL)); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 102; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - __pyx_v_input_str = ((PyObject*)__pyx_t_1); - __pyx_t_1 = 0; + __pyx_v_input_str = ((PyObject*)__pyx_t_2); + __pyx_t_2 = 0; goto __pyx_L3; } @@ -19417,11 +19431,8 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * input_str = str(sentence) # PLF format * else: */ - __pyx_t_1 = ((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Lattice)); - __Pyx_INCREF(__pyx_t_1); - __pyx_t_2 = __Pyx_TypeCheck(__pyx_v_sentence, __pyx_t_1); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_TypeCheck(__pyx_v_sentence, ((PyObject*)__pyx_ptype_5_cdec_Lattice)); + if (__pyx_t_1) { /* "_cdec.pyx":104 * input_str = as_str(sentence.strip()) @@ -19430,14 +19441,14 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * else: * raise TypeError('Cannot translate input type %s' % type(sentence)) */ - __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 104; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 104; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_INCREF(__pyx_v_sentence); - PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_v_sentence); + PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_v_sentence); __Pyx_GIVEREF(__pyx_v_sentence); - __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)(&PyString_Type))), ((PyObject *)__pyx_t_1), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 104; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)(&PyString_Type))), ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 104; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; + __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; if (!(likely(PyBytes_CheckExact(__pyx_t_3))||((__pyx_t_3) == Py_None)||(PyErr_Format(PyExc_TypeError, "Expected bytes, got %.200s", Py_TYPE(__pyx_t_3)->tp_name), 0))) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 104; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_v_input_str = ((PyObject*)__pyx_t_3); __pyx_t_3 = 0; @@ -19454,14 +19465,14 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec */ __pyx_t_3 = PyNumber_Remainder(((PyObject *)__pyx_kp_s_52), ((PyObject *)Py_TYPE(__pyx_v_sentence))); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(((PyObject *)__pyx_t_3)); - __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - PyTuple_SET_ITEM(__pyx_t_1, 0, ((PyObject *)__pyx_t_3)); + __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + PyTuple_SET_ITEM(__pyx_t_2, 0, ((PyObject *)__pyx_t_3)); __Pyx_GIVEREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; - __pyx_t_3 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_1), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyObject_Call(__pyx_builtin_TypeError, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(((PyObject *)__pyx_t_1)); __pyx_t_1 = 0; + __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; __Pyx_Raise(__pyx_t_3, 0, 0, 0); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 106; __pyx_clineno = __LINE__; goto __pyx_L1_error;} @@ -19475,8 +19486,8 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * if isinstance(grammar, basestring): * self.dec.AddSupplementalGrammarFromString(as_str(grammar)) */ - __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_v_grammar); if (unlikely(__pyx_t_2 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 107; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - if (__pyx_t_2) { + __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_v_grammar); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 107; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (__pyx_t_1) { /* "_cdec.pyx":108 * raise TypeError('Cannot translate input type %s' % type(sentence)) @@ -19485,11 +19496,8 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * self.dec.AddSupplementalGrammarFromString(as_str(grammar)) * else: */ - __pyx_t_3 = __pyx_builtin_basestring; - __Pyx_INCREF(__pyx_t_3); - __pyx_t_2 = PyObject_IsInstance(__pyx_v_grammar, __pyx_t_3); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - if (__pyx_t_2) { + __pyx_t_1 = PyObject_IsInstance(__pyx_v_grammar, __pyx_builtin_basestring); if (unlikely(__pyx_t_1 == -1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 108; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (__pyx_t_1) { /* "_cdec.pyx":109 * if grammar: @@ -19519,11 +19527,11 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec __Pyx_INCREF(__pyx_v_grammar); PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_v_grammar); __Pyx_GIVEREF(__pyx_v_grammar); - __pyx_t_1 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_TextGrammar)), ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 111; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); + __pyx_t_2 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_TextGrammar)), ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 111; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; - __pyx_v_self->dec->AddSupplementalGrammar((((struct __pyx_obj_5_cdec_TextGrammar *)__pyx_t_1)->__pyx_base.grammar[0])); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __pyx_v_self->dec->AddSupplementalGrammar((((struct __pyx_obj_5_cdec_TextGrammar *)__pyx_t_2)->__pyx_base.grammar[0])); + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; } __pyx_L5:; goto __pyx_L4; @@ -19556,8 +19564,8 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * raise ParseFailed() * cdef Hypergraph hg = Hypergraph() */ - __pyx_t_2 = (__pyx_v_observer.hypergraph == NULL); - if (__pyx_t_2) { + __pyx_t_1 = (__pyx_v_observer.hypergraph == NULL); + if (__pyx_t_1) { /* "_cdec.pyx":115 * self.dec.Decode(input_str, &observer) @@ -19566,11 +19574,11 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec * cdef Hypergraph hg = Hypergraph() * hg.hg = new hypergraph.Hypergraph(observer.hypergraph[0]) */ - __pyx_t_1 = __Pyx_GetName(__pyx_m, __pyx_n_s__ParseFailed); if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 115; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_t_1); - __pyx_t_3 = PyObject_Call(__pyx_t_1, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 115; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = __Pyx_GetName(__pyx_m, __pyx_n_s__ParseFailed); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 115; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_2); + __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_empty_tuple), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 115; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; + __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_Raise(__pyx_t_3, 0, 0, 0); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; {__pyx_filename = __pyx_f[0]; __pyx_lineno = 115; __pyx_clineno = __LINE__; goto __pyx_L1_error;} @@ -19611,7 +19619,7 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0; __pyx_L1_error:; - __Pyx_XDECREF(__pyx_t_1); + __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_3); __Pyx_AddTraceback("_cdec.Decoder.translate", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; @@ -19623,7 +19631,7 @@ static PyObject *__pyx_pf_5_cdec_7Decoder_6translate(struct __pyx_obj_5_cdec_Dec return __pyx_r; } -/* "string.from_py":11 +/* "string.from_py":12 * * @cname("__pyx_convert_string_from_py_") * cdef string __pyx_convert_string_from_py_(object o) except *: # <<<<<<<<<<<<<< @@ -19641,15 +19649,15 @@ static std::string __pyx_convert_string_from_py_(PyObject *__pyx_v_o) { int __pyx_clineno = 0; __Pyx_RefNannySetupContext("__pyx_convert_string_from_py_", 0); - /* "string.from_py":12 + /* "string.from_py":13 * @cname("__pyx_convert_string_from_py_") * cdef string __pyx_convert_string_from_py_(object o) except *: * return string(<char*>o, len(o)) # <<<<<<<<<<<<<< * * */ - __pyx_t_1 = PyBytes_AsString(__pyx_v_o); if (unlikely((!__pyx_t_1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[6]; __pyx_lineno = 12; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __pyx_t_2 = PyObject_Length(__pyx_v_o); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[6]; __pyx_lineno = 12; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_1 = PyBytes_AsString(__pyx_v_o); if (unlikely((!__pyx_t_1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[6]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = PyObject_Length(__pyx_v_o); if (unlikely(__pyx_t_2 == -1)) {__pyx_filename = __pyx_f[6]; __pyx_lineno = 13; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __pyx_r = std::string(((char *)__pyx_t_1), __pyx_t_2); goto __pyx_L0; @@ -20066,7 +20074,9 @@ static PyObject *__pyx_tp_new_5_cdec_NT(PyTypeObject *t, CYTHON_UNUSED PyObject static void __pyx_tp_dealloc_5_cdec_NT(PyObject *o) { struct __pyx_obj_5_cdec_NT *p = (struct __pyx_obj_5_cdec_NT *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->cat); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -21493,7 +21503,9 @@ static PyObject *__pyx_tp_new_5_cdec_HypergraphEdge(PyTypeObject *t, CYTHON_UNUS static void __pyx_tp_dealloc_5_cdec_HypergraphEdge(PyObject *o) { struct __pyx_obj_5_cdec_HypergraphEdge *p = (struct __pyx_obj_5_cdec_HypergraphEdge *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->trule); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -23074,7 +23086,9 @@ static PyObject *__pyx_tp_new_5_cdec_Metric(PyTypeObject *t, CYTHON_UNUSED PyObj static void __pyx_tp_dealloc_5_cdec_Metric(PyObject *o) { struct __pyx_obj_5_cdec_Metric *p = (struct __pyx_obj_5_cdec_Metric *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->scorer); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -23267,6 +23281,7 @@ static PyObject *__pyx_tp_new_5_cdec_Decoder(PyTypeObject *t, CYTHON_UNUSED PyOb static void __pyx_tp_dealloc_5_cdec_Decoder(PyObject *o) { struct __pyx_obj_5_cdec_Decoder *p = (struct __pyx_obj_5_cdec_Decoder *)o; + PyObject_GC_UnTrack(o); { PyObject *etype, *eval, *etb; PyErr_Fetch(&etype, &eval, &etb); @@ -23277,6 +23292,7 @@ static void __pyx_tp_dealloc_5_cdec_Decoder(PyObject *o) { PyErr_Restore(etype, eval, etb); } Py_CLEAR(p->weights); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -23493,7 +23509,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct____iter__(PyTypeObject * static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct____iter__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct____iter__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct____iter__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -23684,7 +23702,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_1___iter__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_1___iter__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_1___iter__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_1___iter__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -23875,7 +23895,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_2__phrase(PyTypeObject * static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_2__phrase(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_2__phrase *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_2__phrase *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_phrase); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -24068,9 +24090,11 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_3_genexpr(PyTypeObject * static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_3_genexpr(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_3_genexpr *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_3_genexpr *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_outer_scope); Py_CLEAR(p->__pyx_v_w); Py_CLEAR(p->__pyx_t_0); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -24273,7 +24297,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_4___get__(PyTypeObject * static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_4___get__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_4___get__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_4___get__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -24464,7 +24490,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_5___str__(PyTypeObject * static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_5___str__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_5___str__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_5___str__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -24657,9 +24685,11 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_6_genexpr(PyTypeObject * static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_6_genexpr(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_6_genexpr *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_6_genexpr *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_outer_scope); Py_CLEAR(p->__pyx_v_feat); Py_CLEAR(p->__pyx_t_0); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -24863,8 +24893,10 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_7___iter__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_7___iter__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_7___iter__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_7___iter__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); Py_CLEAR(p->__pyx_v_trule); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -25062,8 +25094,10 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_8_kbest(PyTypeObject *t, static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_8_kbest(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_8_kbest *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_8_kbest *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); Py_CLEAR(p->__pyx_v_size); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -25263,10 +25297,12 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_9_kbest_trees(PyTypeObje static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_9_kbest_trees(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_9_kbest_trees *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_9_kbest_trees *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_e_tree); Py_CLEAR(p->__pyx_v_f_tree); Py_CLEAR(p->__pyx_v_self); Py_CLEAR(p->__pyx_v_size); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -25477,9 +25513,11 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_10_kbest_features(PyType static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_10_kbest_features(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_10_kbest_features *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_10_kbest_features *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_fmap); Py_CLEAR(p->__pyx_v_self); Py_CLEAR(p->__pyx_v_size); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -25682,7 +25720,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_11_sample(PyTypeObject * static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_11_sample(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_11_sample *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_11_sample *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -25873,7 +25913,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_12_sample_trees(PyTypeOb static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_12_sample_trees(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_12_sample_trees *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_12_sample_trees *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -26064,7 +26106,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_13___get__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_13___get__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_13___get__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_13___get__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -26255,7 +26299,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_14___get__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_14___get__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_14___get__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_14___get__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -26446,7 +26492,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_15___get__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_15___get__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_15___get__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_15___get__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -26637,7 +26685,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_16___get__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_16___get__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_16___get__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_16___get__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -26828,7 +26878,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_17___get__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_17___get__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_17___get__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_17___get__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -27019,7 +27071,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_18___iter__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_18___iter__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_18___iter__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_18___iter__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -27210,7 +27264,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_19_todot(PyTypeObject *t static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_19_todot(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_19_todot *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_19_todot *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -27407,6 +27463,7 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_20_lines(PyTypeObject *t static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_20_lines(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_20_lines *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_20_lines *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_outer_scope); Py_CLEAR(p->__pyx_v_delta); Py_CLEAR(p->__pyx_v_i); @@ -27414,6 +27471,7 @@ static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_20_lines(PyObject *o) { Py_CLEAR(p->__pyx_v_weight); Py_CLEAR(p->__pyx_t_1); Py_CLEAR(p->__pyx_t_3); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -27642,9 +27700,11 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_21___iter__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_21___iter__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_21___iter__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_21___iter__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_i); Py_CLEAR(p->__pyx_v_self); Py_CLEAR(p->__pyx_t_1); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -27847,7 +27907,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_22___iter__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_22___iter__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_22___iter__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_22___iter__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_self); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -28044,6 +28106,7 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_23__make_config(PyTypeOb static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_23__make_config(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_23__make_config *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_23__make_config *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_config); Py_CLEAR(p->__pyx_v_info); Py_CLEAR(p->__pyx_v_key); @@ -28051,6 +28114,7 @@ static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_23__make_config(PyObject Py_CLEAR(p->__pyx_v_value); Py_CLEAR(p->__pyx_t_0); Py_CLEAR(p->__pyx_t_1); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -28277,7 +28341,9 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_24___init__(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_24___init__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_24___init__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_24___init__ *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_v_config); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -28294,7 +28360,7 @@ static int __pyx_tp_clear_5_cdec___pyx_scope_struct_24___init__(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_24___init__ *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_24___init__ *)o; PyObject* tmp; tmp = ((PyObject*)p->__pyx_v_config); - p->__pyx_v_config = Py_None; Py_INCREF(Py_None); + p->__pyx_v_config = ((PyObject*)Py_None); Py_INCREF(Py_None); Py_XDECREF(tmp); return 0; } @@ -28470,9 +28536,11 @@ static PyObject *__pyx_tp_new_5_cdec___pyx_scope_struct_25_genexpr(PyTypeObject static void __pyx_tp_dealloc_5_cdec___pyx_scope_struct_25_genexpr(PyObject *o) { struct __pyx_obj_5_cdec___pyx_scope_struct_25_genexpr *p = (struct __pyx_obj_5_cdec___pyx_scope_struct_25_genexpr *)o; + PyObject_GC_UnTrack(o); Py_CLEAR(p->__pyx_outer_scope); Py_CLEAR(p->__pyx_v_kv); Py_CLEAR(p->__pyx_t_0); + PyObject_GC_Track(o); (*Py_TYPE(o)->tp_free)(o); } @@ -28670,7 +28738,11 @@ static PyMethodDef __pyx_methods[] = { #if PY_MAJOR_VERSION >= 3 static struct PyModuleDef __pyx_moduledef = { + #if PY_VERSION_HEX < 0x03020000 + { PyObject_HEAD_INIT(NULL) NULL, 0, NULL }, + #else PyModuleDef_HEAD_INIT, + #endif __Pyx_NAMESTR("_cdec"), 0, /* m_doc */ -1, /* m_size */ @@ -28704,12 +28776,13 @@ static __Pyx_StringTabEntry __pyx_string_tab[] = { {&__pyx_kp_s_33, __pyx_k_33, sizeof(__pyx_k_33), 0, 0, 1, 0}, {&__pyx_kp_s_34, __pyx_k_34, sizeof(__pyx_k_34), 0, 0, 1, 0}, {&__pyx_kp_s_37, __pyx_k_37, sizeof(__pyx_k_37), 0, 0, 1, 0}, - {&__pyx_kp_s_38, __pyx_k_38, sizeof(__pyx_k_38), 0, 0, 1, 0}, + {&__pyx_n_s_38, __pyx_k_38, sizeof(__pyx_k_38), 0, 0, 1, 1}, + {&__pyx_kp_s_39, __pyx_k_39, sizeof(__pyx_k_39), 0, 0, 1, 0}, {&__pyx_kp_s_4, __pyx_k_4, sizeof(__pyx_k_4), 0, 0, 1, 0}, - {&__pyx_kp_s_40, __pyx_k_40, sizeof(__pyx_k_40), 0, 0, 1, 0}, - {&__pyx_kp_s_42, __pyx_k_42, sizeof(__pyx_k_42), 0, 0, 1, 0}, - {&__pyx_kp_s_44, __pyx_k_44, sizeof(__pyx_k_44), 0, 0, 1, 0}, + {&__pyx_kp_s_41, __pyx_k_41, sizeof(__pyx_k_41), 0, 0, 1, 0}, + {&__pyx_kp_s_43, __pyx_k_43, sizeof(__pyx_k_43), 0, 0, 1, 0}, {&__pyx_kp_s_45, __pyx_k_45, sizeof(__pyx_k_45), 0, 0, 1, 0}, + {&__pyx_kp_s_46, __pyx_k_46, sizeof(__pyx_k_46), 0, 0, 1, 0}, {&__pyx_kp_s_47, __pyx_k_47, sizeof(__pyx_k_47), 0, 0, 1, 0}, {&__pyx_kp_s_48, __pyx_k_48, sizeof(__pyx_k_48), 0, 0, 1, 0}, {&__pyx_kp_s_49, __pyx_k_49, sizeof(__pyx_k_49), 0, 0, 1, 0}, @@ -28717,7 +28790,7 @@ static __Pyx_StringTabEntry __pyx_string_tab[] = { {&__pyx_n_s_53, __pyx_k_53, sizeof(__pyx_k_53), 0, 0, 1, 1}, {&__pyx_n_s_54, __pyx_k_54, sizeof(__pyx_k_54), 0, 0, 1, 1}, {&__pyx_kp_s_57, __pyx_k_57, sizeof(__pyx_k_57), 0, 0, 1, 0}, - {&__pyx_kp_s_63, __pyx_k_63, sizeof(__pyx_k_63), 0, 0, 1, 0}, + {&__pyx_kp_s_65, __pyx_k_65, sizeof(__pyx_k_65), 0, 0, 1, 0}, {&__pyx_kp_s_7, __pyx_k_7, sizeof(__pyx_k_7), 0, 0, 1, 0}, {&__pyx_kp_s_8, __pyx_k_8, sizeof(__pyx_k_8), 0, 0, 1, 0}, {&__pyx_kp_s_9, __pyx_k_9, sizeof(__pyx_k_9), 0, 0, 1, 0}, @@ -28730,6 +28803,9 @@ static __Pyx_StringTabEntry __pyx_string_tab[] = { {&__pyx_n_s__KeyError, __pyx_k__KeyError, sizeof(__pyx_k__KeyError), 0, 0, 1, 1}, {&__pyx_n_s__NotImplemented, __pyx_k__NotImplemented, sizeof(__pyx_k__NotImplemented), 0, 0, 1, 1}, {&__pyx_n_s__ParseFailed, __pyx_k__ParseFailed, sizeof(__pyx_k__ParseFailed), 0, 0, 1, 1}, + {&__pyx_n_s__QCRI, __pyx_k__QCRI, sizeof(__pyx_k__QCRI), 0, 0, 1, 1}, + {&__pyx_n_s__QCRI_BLEU, __pyx_k__QCRI_BLEU, sizeof(__pyx_k__QCRI_BLEU), 0, 0, 1, 1}, + {&__pyx_n_s__SSK, __pyx_k__SSK, sizeof(__pyx_k__SSK), 0, 0, 1, 1}, {&__pyx_n_s__TER, __pyx_k__TER, sizeof(__pyx_k__TER), 0, 0, 1, 1}, {&__pyx_n_s__TypeError, __pyx_k__TypeError, sizeof(__pyx_k__TypeError), 0, 0, 1, 1}, {&__pyx_n_s__ValueError, __pyx_k__ValueError, sizeof(__pyx_k__ValueError), 0, 0, 1, 1}, @@ -28767,7 +28843,6 @@ static __Pyx_StringTabEntry __pyx_string_tab[] = { {&__pyx_n_s__format, __pyx_k__format, sizeof(__pyx_k__format), 0, 0, 1, 1}, {&__pyx_n_s__fst, __pyx_k__fst, sizeof(__pyx_k__fst), 0, 0, 1, 1}, {&__pyx_n_s__genexpr, __pyx_k__genexpr, sizeof(__pyx_k__genexpr), 0, 0, 1, 1}, - {&__pyx_n_s__get, __pyx_k__get, sizeof(__pyx_k__get), 0, 0, 1, 1}, {&__pyx_n_s__grammar, __pyx_k__grammar, sizeof(__pyx_k__grammar), 0, 0, 1, 1}, {&__pyx_n_s__hyp, __pyx_k__hyp, sizeof(__pyx_k__hyp), 0, 0, 1, 1}, {&__pyx_n_s__hypergraph, __pyx_k__hypergraph, sizeof(__pyx_k__hypergraph), 0, 0, 1, 1}, @@ -28842,207 +28917,139 @@ static int __Pyx_InitCachedConstants(void) { * elif isinstance(data, str): * ret = data */ - __pyx_k_tuple_2 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_2 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__utf8)); if (unlikely(!__pyx_k_tuple_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 9; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_2); - __Pyx_INCREF(((PyObject *)__pyx_n_s__utf8)); - PyTuple_SET_ITEM(__pyx_k_tuple_2, 0, ((PyObject *)__pyx_n_s__utf8)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__utf8)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_2)); - /* "/home/pauldb/workspace/cdec/python/src/vectors.pxi":95 + /* "/usr0/home/cdyer/cdec-merge/python/src/vectors.pxi":95 * elif op == 3: # != * return not (x == y) * raise NotImplemented('comparison not implemented for SparseVector') # <<<<<<<<<<<<<< * * def __len__(self): */ - __pyx_k_tuple_5 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_5)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 95; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_5 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_4)); if (unlikely(!__pyx_k_tuple_5)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 95; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_5); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_4)); - PyTuple_SET_ITEM(__pyx_k_tuple_5, 0, ((PyObject *)__pyx_kp_s_4)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_4)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_5)); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":6 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":6 * * def _phrase(phrase): * return ' '.join(w.encode('utf8') if isinstance(w, unicode) else str(w) for w in phrase) # <<<<<<<<<<<<<< * * cdef class NT: */ - __pyx_k_tuple_6 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_6 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__utf8)); if (unlikely(!__pyx_k_tuple_6)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 6; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_6); - __Pyx_INCREF(((PyObject *)__pyx_n_s__utf8)); - PyTuple_SET_ITEM(__pyx_k_tuple_6, 0, ((PyObject *)__pyx_n_s__utf8)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__utf8)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_6)); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":226 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":226 * trule = convert_rule(trule) * elif not isinstance(trule, TRule): * raise ValueError('the grammar should contain TRule objects') # <<<<<<<<<<<<<< * _g.AddRule((<TRule> trule).rule[0]) */ - __pyx_k_tuple_14 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_14)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 226; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_14 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_13)); if (unlikely(!__pyx_k_tuple_14)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 226; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_14); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_13)); - PyTuple_SET_ITEM(__pyx_k_tuple_14, 0, ((PyObject *)__pyx_kp_s_13)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_13)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_14)); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":244 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":244 * elif op == 3: # != * return not (x == y) * raise NotImplemented('comparison not implemented for HypergraphEdge') # <<<<<<<<<<<<<< * * cdef class HypergraphNode: */ - __pyx_k_tuple_19 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_19)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 244; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_19 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_18)); if (unlikely(!__pyx_k_tuple_19)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 244; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_19); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_18)); - PyTuple_SET_ITEM(__pyx_k_tuple_19, 0, ((PyObject *)__pyx_kp_s_18)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_18)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_19)); - /* "/home/pauldb/workspace/cdec/python/src/hypergraph.pxi":281 + /* "/usr0/home/cdyer/cdec-merge/python/src/hypergraph.pxi":281 * elif op == 3: # != * return not (x == y) * raise NotImplemented('comparison not implemented for HypergraphNode') # <<<<<<<<<<<<<< */ - __pyx_k_tuple_21 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_21)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 281; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_21 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_20)); if (unlikely(!__pyx_k_tuple_21)) {__pyx_filename = __pyx_f[3]; __pyx_lineno = 281; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_21); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_20)); - PyTuple_SET_ITEM(__pyx_k_tuple_21, 0, ((PyObject *)__pyx_kp_s_20)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_20)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_21)); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":26 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":26 * def __getitem__(self, int index): * if not 0 <= index < len(self): * raise IndexError('lattice index out of range') # <<<<<<<<<<<<<< * arcs = [] * cdef vector[lattice.LatticeArc] arc_vector = self.lattice[0][index] */ - __pyx_k_tuple_24 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_24)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 26; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_24 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_23)); if (unlikely(!__pyx_k_tuple_24)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 26; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_24); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_23)); - PyTuple_SET_ITEM(__pyx_k_tuple_24, 0, ((PyObject *)__pyx_kp_s_23)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_23)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_24)); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":39 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":39 * def __setitem__(self, int index, tuple arcs): * if not 0 <= index < len(self): * raise IndexError('lattice index out of range') # <<<<<<<<<<<<<< * cdef lattice.LatticeArc* arc * for (label, cost, dist2next) in arcs: */ - __pyx_k_tuple_25 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_25)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 39; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_25 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_23)); if (unlikely(!__pyx_k_tuple_25)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 39; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_25); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_23)); - PyTuple_SET_ITEM(__pyx_k_tuple_25, 0, ((PyObject *)__pyx_kp_s_23)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_23)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_25)); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":69 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":69 * for i in range(len(self)): * for label, weight, delta in self[i]: * yield '%d -> %d [label="%s"];' % (i, i+delta, label.replace('"', '\\"')) # <<<<<<<<<<<<<< * yield '%d [shape=doublecircle]' % len(self) * yield '}' */ - __pyx_k_tuple_32 = PyTuple_New(2); if (unlikely(!__pyx_k_tuple_32)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 69; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_32 = PyTuple_Pack(2, ((PyObject *)__pyx_kp_s_30), ((PyObject *)__pyx_kp_s_31)); if (unlikely(!__pyx_k_tuple_32)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 69; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_32); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_30)); - PyTuple_SET_ITEM(__pyx_k_tuple_32, 0, ((PyObject *)__pyx_kp_s_30)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_30)); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_31)); - PyTuple_SET_ITEM(__pyx_k_tuple_32, 1, ((PyObject *)__pyx_kp_s_31)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_31)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_32)); - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":63 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":63 * def todot(self): * """lattice.todot() -> Representation of the lattice in GraphViz dot format.""" * def lines(): # <<<<<<<<<<<<<< * yield 'digraph lattice {' * yield 'rankdir = LR;' */ - __pyx_k_tuple_35 = PyTuple_New(4); if (unlikely(!__pyx_k_tuple_35)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_35 = PyTuple_Pack(4, ((PyObject *)__pyx_n_s__i), ((PyObject *)__pyx_n_s__label), ((PyObject *)__pyx_n_s__weight), ((PyObject *)__pyx_n_s__delta)); if (unlikely(!__pyx_k_tuple_35)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_35); - __Pyx_INCREF(((PyObject *)__pyx_n_s__i)); - PyTuple_SET_ITEM(__pyx_k_tuple_35, 0, ((PyObject *)__pyx_n_s__i)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__i)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__label)); - PyTuple_SET_ITEM(__pyx_k_tuple_35, 1, ((PyObject *)__pyx_n_s__label)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__label)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__weight)); - PyTuple_SET_ITEM(__pyx_k_tuple_35, 2, ((PyObject *)__pyx_n_s__weight)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__weight)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__delta)); - PyTuple_SET_ITEM(__pyx_k_tuple_35, 3, ((PyObject *)__pyx_n_s__delta)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__delta)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_35)); __pyx_k_codeobj_36 = (PyObject*)__Pyx_PyCode_New(0, 0, 4, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_35, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_37, __pyx_n_s__lines, 63, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_36)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 63; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/lattice.pxi":72 + /* "/usr0/home/cdyer/cdec-merge/python/src/lattice.pxi":72 * yield '%d [shape=doublecircle]' % len(self) * yield '}' * return '\n'.join(lines()).encode('utf8') # <<<<<<<<<<<<<< * * def as_hypergraph(self): */ - __pyx_k_tuple_39 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_39)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_k_tuple_39); - __Pyx_INCREF(((PyObject *)__pyx_n_s__utf8)); - PyTuple_SET_ITEM(__pyx_k_tuple_39, 0, ((PyObject *)__pyx_n_s__utf8)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__utf8)); - __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_39)); + __pyx_k_tuple_40 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__utf8)); if (unlikely(!__pyx_k_tuple_40)) {__pyx_filename = __pyx_f[4]; __pyx_lineno = 72; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_k_tuple_40); + __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_40)); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":50 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":50 * def __getitem__(self, int index): * if not 0 <= index < len(self): * raise IndexError('sufficient stats vector index out of range') # <<<<<<<<<<<<<< * return self.stats[0][index] * */ - __pyx_k_tuple_41 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_41)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 50; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_k_tuple_41); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_40)); - PyTuple_SET_ITEM(__pyx_k_tuple_41, 0, ((PyObject *)__pyx_kp_s_40)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_40)); - __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_41)); + __pyx_k_tuple_42 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_41)); if (unlikely(!__pyx_k_tuple_42)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 50; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_k_tuple_42); + __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_42)); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":84 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":84 * def __getitem__(self,int k): * if not 0 <= k < self.cs.size(): * raise IndexError('candidate set index out of range') # <<<<<<<<<<<<<< * cdef Candidate candidate = Candidate() * candidate.candidate = &self.cs[0][k] */ - __pyx_k_tuple_43 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_43)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 84; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_k_tuple_43); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_42)); - PyTuple_SET_ITEM(__pyx_k_tuple_43, 0, ((PyObject *)__pyx_kp_s_42)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_42)); - __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_43)); - - /* "_cdec.pyx":52 - * Create a decoder using a given configuration. Formalism is required.""" - * if config_str is None: - * formalism = config.get('formalism', None) # <<<<<<<<<<<<<< - * if formalism not in ('scfg', 'fst', 'lextrans', 'pb', - * 'csplit', 'tagger', 'lexalign'): - */ - __pyx_k_tuple_46 = PyTuple_New(2); if (unlikely(!__pyx_k_tuple_46)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 52; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_k_tuple_46); - __Pyx_INCREF(((PyObject *)__pyx_n_s__formalism)); - PyTuple_SET_ITEM(__pyx_k_tuple_46, 0, ((PyObject *)__pyx_n_s__formalism)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__formalism)); - __Pyx_INCREF(Py_None); - PyTuple_SET_ITEM(__pyx_k_tuple_46, 1, Py_None); - __Pyx_GIVEREF(Py_None); - __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_46)); + __pyx_k_tuple_44 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_43)); if (unlikely(!__pyx_k_tuple_44)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 84; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_k_tuple_44); + __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_44)); /* "_cdec.pyx":93 * with open(weights) as fp: @@ -29051,11 +29058,8 @@ static int __Pyx_InitCachedConstants(void) { * fname, value = line.split() * self.weights[fname.strip()] = float(value) */ - __pyx_k_tuple_50 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_50)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_50 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_49)); if (unlikely(!__pyx_k_tuple_50)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 93; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_50); - __Pyx_INCREF(((PyObject *)__pyx_kp_s_49)); - PyTuple_SET_ITEM(__pyx_k_tuple_50, 0, ((PyObject *)__pyx_kp_s_49)); - __Pyx_GIVEREF(((PyObject *)__pyx_kp_s_49)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_50)); /* "_cdec.pyx":91 @@ -29065,79 +29069,74 @@ static int __Pyx_InitCachedConstants(void) { * for line in fp: * if line.strip().startswith('#'): continue */ - __pyx_k_tuple_51 = PyTuple_New(3); if (unlikely(!__pyx_k_tuple_51)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_51 = PyTuple_Pack(3, Py_None, Py_None, Py_None); if (unlikely(!__pyx_k_tuple_51)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 91; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_51); - __Pyx_INCREF(Py_None); - PyTuple_SET_ITEM(__pyx_k_tuple_51, 0, Py_None); - __Pyx_GIVEREF(Py_None); - __Pyx_INCREF(Py_None); - PyTuple_SET_ITEM(__pyx_k_tuple_51, 1, Py_None); - __Pyx_GIVEREF(Py_None); - __Pyx_INCREF(Py_None); - PyTuple_SET_ITEM(__pyx_k_tuple_51, 2, Py_None); - __Pyx_GIVEREF(Py_None); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_51)); - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":5 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":5 * import cdec.sa._sa as _sa * * def _phrase(phrase): # <<<<<<<<<<<<<< * return ' '.join(w.encode('utf8') if isinstance(w, unicode) else str(w) for w in phrase) * */ - __pyx_k_tuple_55 = PyTuple_New(3); if (unlikely(!__pyx_k_tuple_55)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 5; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_55 = PyTuple_Pack(3, ((PyObject *)__pyx_n_s__phrase), ((PyObject *)__pyx_n_s__genexpr), ((PyObject *)__pyx_n_s__genexpr)); if (unlikely(!__pyx_k_tuple_55)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 5; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_55); - __Pyx_INCREF(((PyObject *)__pyx_n_s__phrase)); - PyTuple_SET_ITEM(__pyx_k_tuple_55, 0, ((PyObject *)__pyx_n_s__phrase)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__phrase)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__genexpr)); - PyTuple_SET_ITEM(__pyx_k_tuple_55, 1, ((PyObject *)__pyx_n_s__genexpr)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__genexpr)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__genexpr)); - PyTuple_SET_ITEM(__pyx_k_tuple_55, 2, ((PyObject *)__pyx_n_s__genexpr)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__genexpr)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_55)); __pyx_k_codeobj_56 = (PyObject*)__Pyx_PyCode_New(1, 0, 3, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_55, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_57, __pyx_n_s___phrase, 5, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_56)) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 5; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":194 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":194 * return [] * * BLEU = Scorer('IBM_BLEU') # <<<<<<<<<<<<<< + * QCRI = Scorer('QCRI_BLEU') * TER = Scorer('TER') - * CER = Scorer('CER') */ - __pyx_k_tuple_58 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_58)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 194; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_58 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__IBM_BLEU)); if (unlikely(!__pyx_k_tuple_58)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 194; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_58); - __Pyx_INCREF(((PyObject *)__pyx_n_s__IBM_BLEU)); - PyTuple_SET_ITEM(__pyx_k_tuple_58, 0, ((PyObject *)__pyx_n_s__IBM_BLEU)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__IBM_BLEU)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_58)); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":195 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":195 * * BLEU = Scorer('IBM_BLEU') - * TER = Scorer('TER') # <<<<<<<<<<<<<< + * QCRI = Scorer('QCRI_BLEU') # <<<<<<<<<<<<<< + * TER = Scorer('TER') * CER = Scorer('CER') */ - __pyx_k_tuple_59 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_59)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 195; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_59 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__QCRI_BLEU)); if (unlikely(!__pyx_k_tuple_59)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 195; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_59); - __Pyx_INCREF(((PyObject *)__pyx_n_s__TER)); - PyTuple_SET_ITEM(__pyx_k_tuple_59, 0, ((PyObject *)__pyx_n_s__TER)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__TER)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_59)); - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":196 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":196 * BLEU = Scorer('IBM_BLEU') - * TER = Scorer('TER') - * CER = Scorer('CER') # <<<<<<<<<<<<<< + * QCRI = Scorer('QCRI_BLEU') + * TER = Scorer('TER') # <<<<<<<<<<<<<< + * CER = Scorer('CER') + * SSK = Scorer('SSK') */ - __pyx_k_tuple_60 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_60)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 196; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_60 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__TER)); if (unlikely(!__pyx_k_tuple_60)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 196; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_k_tuple_60); - __Pyx_INCREF(((PyObject *)__pyx_n_s__CER)); - PyTuple_SET_ITEM(__pyx_k_tuple_60, 0, ((PyObject *)__pyx_n_s__CER)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__CER)); __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_60)); + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":197 + * QCRI = Scorer('QCRI_BLEU') + * TER = Scorer('TER') + * CER = Scorer('CER') # <<<<<<<<<<<<<< + * SSK = Scorer('SSK') + */ + __pyx_k_tuple_61 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__CER)); if (unlikely(!__pyx_k_tuple_61)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 197; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_k_tuple_61); + __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_61)); + + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":198 + * TER = Scorer('TER') + * CER = Scorer('CER') + * SSK = Scorer('SSK') # <<<<<<<<<<<<<< + */ + __pyx_k_tuple_62 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__SSK)); if (unlikely(!__pyx_k_tuple_62)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 198; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_k_tuple_62); + __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_62)); + /* "_cdec.pyx":28 * class ParseFailed(Exception): pass * @@ -29145,13 +29144,10 @@ static int __Pyx_InitCachedConstants(void) { * """set_silent(bool): Configure the verbosity of cdec.""" * SetSilent(yn) */ - __pyx_k_tuple_61 = PyTuple_New(1); if (unlikely(!__pyx_k_tuple_61)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_k_tuple_61); - __Pyx_INCREF(((PyObject *)__pyx_n_s__yn)); - PyTuple_SET_ITEM(__pyx_k_tuple_61, 0, ((PyObject *)__pyx_n_s__yn)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__yn)); - __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_61)); - __pyx_k_codeobj_62 = (PyObject*)__Pyx_PyCode_New(1, 0, 1, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_61, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_63, __pyx_n_s__set_silent, 28, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_62)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_63 = PyTuple_Pack(1, ((PyObject *)__pyx_n_s__yn)); if (unlikely(!__pyx_k_tuple_63)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_k_tuple_63); + __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_63)); + __pyx_k_codeobj_64 = (PyObject*)__Pyx_PyCode_New(1, 0, 1, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_63, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_65, __pyx_n_s__set_silent, 28, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_64)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 28; __pyx_clineno = __LINE__; goto __pyx_L1_error;} /* "_cdec.pyx":32 * SetSilent(yn) @@ -29160,25 +29156,10 @@ static int __Pyx_InitCachedConstants(void) { * for key, value in config.items(): * if isinstance(value, dict): */ - __pyx_k_tuple_64 = PyTuple_New(5); if (unlikely(!__pyx_k_tuple_64)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 32; __pyx_clineno = __LINE__; goto __pyx_L1_error;} - __Pyx_GOTREF(__pyx_k_tuple_64); - __Pyx_INCREF(((PyObject *)__pyx_n_s__config)); - PyTuple_SET_ITEM(__pyx_k_tuple_64, 0, ((PyObject *)__pyx_n_s__config)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__config)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__key)); - PyTuple_SET_ITEM(__pyx_k_tuple_64, 1, ((PyObject *)__pyx_n_s__key)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__key)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__value)); - PyTuple_SET_ITEM(__pyx_k_tuple_64, 2, ((PyObject *)__pyx_n_s__value)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__value)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__name)); - PyTuple_SET_ITEM(__pyx_k_tuple_64, 3, ((PyObject *)__pyx_n_s__name)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__name)); - __Pyx_INCREF(((PyObject *)__pyx_n_s__info)); - PyTuple_SET_ITEM(__pyx_k_tuple_64, 4, ((PyObject *)__pyx_n_s__info)); - __Pyx_GIVEREF(((PyObject *)__pyx_n_s__info)); - __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_64)); - __pyx_k_codeobj_65 = (PyObject*)__Pyx_PyCode_New(1, 0, 5, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_64, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_63, __pyx_n_s___make_config, 32, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_65)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 32; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_k_tuple_66 = PyTuple_Pack(5, ((PyObject *)__pyx_n_s__config), ((PyObject *)__pyx_n_s__key), ((PyObject *)__pyx_n_s__value), ((PyObject *)__pyx_n_s__name), ((PyObject *)__pyx_n_s__info)); if (unlikely(!__pyx_k_tuple_66)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 32; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_k_tuple_66); + __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_66)); + __pyx_k_codeobj_67 = (PyObject*)__Pyx_PyCode_New(1, 0, 5, 0, 0, __pyx_empty_bytes, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_k_tuple_66, __pyx_empty_tuple, __pyx_empty_tuple, __pyx_kp_s_65, __pyx_n_s___make_config, 32, __pyx_empty_bytes); if (unlikely(!__pyx_k_codeobj_67)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 32; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_RefNannyFinishContext(); return 0; __pyx_L1_error:; @@ -29244,6 +29225,14 @@ PyMODINIT_FUNC PyInit__cdec(void) __pyx_m = PyModule_Create(&__pyx_moduledef); #endif if (unlikely(!__pyx_m)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + #if PY_MAJOR_VERSION >= 3 + { + PyObject *modules = PyImport_GetModuleDict(); if (unlikely(!modules)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (!PyDict_GetItemString(modules, "_cdec")) { + if (unlikely(PyDict_SetItemString(modules, "_cdec", __pyx_m) < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + } + } + #endif __pyx_b = PyImport_AddModule(__Pyx_NAMESTR(__Pyx_BUILTIN_MODULE_NAME)); if (unlikely(!__pyx_b)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} #if CYTHON_COMPILING_IN_PYPY Py_INCREF(__pyx_b); @@ -29491,7 +29480,7 @@ PyMODINIT_FUNC PyInit__cdec(void) Py_DECREF(__pyx_t_1); __pyx_t_1 = 0; /*--- Execution code ---*/ - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":3 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":3 * cimport grammar * cimport cdec.sa._sa as _sa * import cdec.sa._sa as _sa # <<<<<<<<<<<<<< @@ -29509,7 +29498,7 @@ PyMODINIT_FUNC PyInit__cdec(void) if (PyObject_SetAttr(__pyx_m, __pyx_n_s___sa, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 3; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - /* "/home/pauldb/workspace/cdec/python/src/grammar.pxi":5 + /* "/usr0/home/cdyer/cdec-merge/python/src/grammar.pxi":5 * import cdec.sa._sa as _sa * * def _phrase(phrase): # <<<<<<<<<<<<<< @@ -29521,37 +29510,61 @@ PyMODINIT_FUNC PyInit__cdec(void) if (PyObject_SetAttr(__pyx_m, __pyx_n_s___phrase, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[2]; __pyx_lineno = 5; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":194 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":194 * return [] * * BLEU = Scorer('IBM_BLEU') # <<<<<<<<<<<<<< + * QCRI = Scorer('QCRI_BLEU') * TER = Scorer('TER') - * CER = Scorer('CER') */ __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Scorer)), ((PyObject *)__pyx_k_tuple_58), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 194; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); if (PyObject_SetAttr(__pyx_m, __pyx_n_s__BLEU, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 194; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":195 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":195 * * BLEU = Scorer('IBM_BLEU') - * TER = Scorer('TER') # <<<<<<<<<<<<<< + * QCRI = Scorer('QCRI_BLEU') # <<<<<<<<<<<<<< + * TER = Scorer('TER') * CER = Scorer('CER') */ __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Scorer)), ((PyObject *)__pyx_k_tuple_59), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 195; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - if (PyObject_SetAttr(__pyx_m, __pyx_n_s__TER, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 195; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (PyObject_SetAttr(__pyx_m, __pyx_n_s__QCRI, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 195; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; - /* "/home/pauldb/workspace/cdec/python/src/mteval.pxi":196 + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":196 * BLEU = Scorer('IBM_BLEU') + * QCRI = Scorer('QCRI_BLEU') + * TER = Scorer('TER') # <<<<<<<<<<<<<< + * CER = Scorer('CER') + * SSK = Scorer('SSK') + */ + __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Scorer)), ((PyObject *)__pyx_k_tuple_60), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 196; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); + if (PyObject_SetAttr(__pyx_m, __pyx_n_s__TER, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 196; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; + + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":197 + * QCRI = Scorer('QCRI_BLEU') * TER = Scorer('TER') * CER = Scorer('CER') # <<<<<<<<<<<<<< + * SSK = Scorer('SSK') */ - __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Scorer)), ((PyObject *)__pyx_k_tuple_60), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 196; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Scorer)), ((PyObject *)__pyx_k_tuple_61), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 197; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_GOTREF(__pyx_t_3); + if (PyObject_SetAttr(__pyx_m, __pyx_n_s__CER, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 197; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; + + /* "/usr0/home/cdyer/cdec-merge/python/src/mteval.pxi":198 + * TER = Scorer('TER') + * CER = Scorer('CER') + * SSK = Scorer('SSK') # <<<<<<<<<<<<<< + */ + __pyx_t_3 = PyObject_Call(((PyObject *)((PyObject*)__pyx_ptype_5_cdec_Scorer)), ((PyObject *)__pyx_k_tuple_62), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 198; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_3); - if (PyObject_SetAttr(__pyx_m, __pyx_n_s__CER, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 196; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + if (PyObject_SetAttr(__pyx_m, __pyx_n_s__SSK, __pyx_t_3) < 0) {__pyx_filename = __pyx_f[5]; __pyx_lineno = 198; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; /* "_cdec.pyx":22 @@ -29586,7 +29599,7 @@ PyMODINIT_FUNC PyInit__cdec(void) __Pyx_INCREF(__pyx_builtin_Exception); PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_builtin_Exception); __Pyx_GIVEREF(__pyx_builtin_Exception); - __pyx_t_4 = __Pyx_CreateClass(((PyObject *)__pyx_t_2), ((PyObject *)__pyx_t_3), __pyx_n_s__InvalidConfig, __pyx_n_s___cdec); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 25; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_4 = __Pyx_CreateClass(((PyObject *)__pyx_t_2), ((PyObject *)__pyx_t_3), __pyx_n_s__InvalidConfig, __pyx_n_s__InvalidConfig, __pyx_n_s___cdec); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 25; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_4); __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0; if (PyObject_SetAttr(__pyx_m, __pyx_n_s__InvalidConfig, __pyx_t_4) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 25; __pyx_clineno = __LINE__; goto __pyx_L1_error;} @@ -29607,7 +29620,7 @@ PyMODINIT_FUNC PyInit__cdec(void) __Pyx_INCREF(__pyx_builtin_Exception); PyTuple_SET_ITEM(__pyx_t_4, 0, __pyx_builtin_Exception); __Pyx_GIVEREF(__pyx_builtin_Exception); - __pyx_t_2 = __Pyx_CreateClass(((PyObject *)__pyx_t_4), ((PyObject *)__pyx_t_3), __pyx_n_s__ParseFailed, __pyx_n_s___cdec); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 26; __pyx_clineno = __LINE__; goto __pyx_L1_error;} + __pyx_t_2 = __Pyx_CreateClass(((PyObject *)__pyx_t_4), ((PyObject *)__pyx_t_3), __pyx_n_s__ParseFailed, __pyx_n_s__ParseFailed, __pyx_n_s___cdec); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 26; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_GOTREF(__pyx_t_2); __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0; if (PyObject_SetAttr(__pyx_m, __pyx_n_s__ParseFailed, __pyx_t_2) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 26; __pyx_clineno = __LINE__; goto __pyx_L1_error;} @@ -29648,7 +29661,7 @@ PyMODINIT_FUNC PyInit__cdec(void) if (PyObject_SetAttr(__pyx_m, __pyx_n_s____test__, ((PyObject *)__pyx_t_3)) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 1; __pyx_clineno = __LINE__; goto __pyx_L1_error;} __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0; - /* "string.from_py":11 + /* "string.from_py":12 * * @cname("__pyx_convert_string_from_py_") * cdef string __pyx_convert_string_from_py_(object o) except *: # <<<<<<<<<<<<<< @@ -30315,16 +30328,18 @@ static void __Pyx_ExceptionReset(PyObject *type, PyObject *value, PyObject *tb) #endif } -static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, long level) { - PyObject *py_import = 0; +static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, int level) { PyObject *empty_list = 0; PyObject *module = 0; PyObject *global_dict = 0; PyObject *empty_dict = 0; PyObject *list; + #if PY_VERSION_HEX < 0x03030000 + PyObject *py_import = 0; py_import = __Pyx_GetAttrString(__pyx_b, "__import__"); if (!py_import) goto bad; + #endif if (from_list) list = from_list; else { @@ -30344,13 +30359,17 @@ static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, long level) { #if PY_MAJOR_VERSION >= 3 if (level == -1) { if (strchr(__Pyx_MODULE_NAME, '.')) { - /* try package relative import first */ + #if PY_VERSION_HEX < 0x03030000 PyObject *py_level = PyInt_FromLong(1); if (!py_level) goto bad; module = PyObject_CallFunctionObjArgs(py_import, name, global_dict, empty_dict, list, py_level, NULL); Py_DECREF(py_level); + #else + module = PyImport_ImportModuleLevelObject( + name, global_dict, empty_dict, list, 1); + #endif if (!module) { if (!PyErr_ExceptionMatches(PyExc_ImportError)) goto bad; @@ -30361,12 +30380,17 @@ static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, long level) { } #endif if (!module) { + #if PY_VERSION_HEX < 0x03030000 PyObject *py_level = PyInt_FromLong(level); if (!py_level) goto bad; module = PyObject_CallFunctionObjArgs(py_import, name, global_dict, empty_dict, list, py_level, NULL); Py_DECREF(py_level); + #else + module = PyImport_ImportModuleLevelObject( + name, global_dict, empty_dict, list, level); + #endif } } #else @@ -30378,8 +30402,10 @@ static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, long level) { name, global_dict, empty_dict, list, NULL); #endif bad: - Py_XDECREF(empty_list); + #if PY_VERSION_HEX < 0x03030000 Py_XDECREF(py_import); + #endif + Py_XDECREF(empty_list); Py_XDECREF(empty_dict); return module; } @@ -30410,11 +30436,13 @@ static PyObject *__Pyx_FindPy2Metaclass(PyObject *bases) { } static PyObject *__Pyx_CreateClass(PyObject *bases, PyObject *dict, PyObject *name, - PyObject *modname) { + PyObject *qualname, PyObject *modname) { PyObject *result; PyObject *metaclass; if (PyDict_SetItemString(dict, "__module__", modname) < 0) return NULL; + if (PyDict_SetItemString(dict, "__qualname__", qualname) < 0) + return NULL; metaclass = PyDict_GetItemString(dict, "__metaclass__"); if (metaclass) { Py_INCREF(metaclass); @@ -30455,6 +30483,10 @@ static CYTHON_INLINE WordID __Pyx_PyInt_from_py_WordID(PyObject* x) { else return (WordID)__Pyx_PyInt_AsSignedLongLong(x); } else { + #if CYTHON_COMPILING_IN_PYPY && !defined(_PyLong_AsByteArray) + PyErr_SetString(PyExc_RuntimeError, + "_PyLong_AsByteArray() not available in PyPy, cannot convert large numbers"); + #else WordID val; PyObject *v = __Pyx_PyNumber_Int(x); #if PY_VERSION_HEX < 0x03000000 @@ -30474,6 +30506,7 @@ static CYTHON_INLINE WordID __Pyx_PyInt_from_py_WordID(PyObject* x) { if (likely(!ret)) return val; } + #endif return (WordID)-1; } } @@ -30529,16 +30562,19 @@ bad: static PyObject * __Pyx_CyFunction_get_doc(__pyx_CyFunctionObject *op, CYTHON_UNUSED void *closure) { - if (op->func_doc == NULL && op->func.m_ml->ml_doc) { + if (unlikely(op->func_doc == NULL)) { + if (op->func.m_ml->ml_doc) { #if PY_MAJOR_VERSION >= 3 - op->func_doc = PyUnicode_FromString(op->func.m_ml->ml_doc); + op->func_doc = PyUnicode_FromString(op->func.m_ml->ml_doc); #else - op->func_doc = PyString_FromString(op->func.m_ml->ml_doc); + op->func_doc = PyString_FromString(op->func.m_ml->ml_doc); #endif - } - if (op->func_doc == 0) { - Py_INCREF(Py_None); - return Py_None; + if (unlikely(op->func_doc == NULL)) + return NULL; + } else { + Py_INCREF(Py_None); + return Py_None; + } } Py_INCREF(op->func_doc); return op->func_doc; @@ -30548,22 +30584,23 @@ __Pyx_CyFunction_set_doc(__pyx_CyFunctionObject *op, PyObject *value) { PyObject *tmp = op->func_doc; if (value == NULL) - op->func_doc = Py_None; /* Mark as deleted */ - else - op->func_doc = value; - Py_INCREF(op->func_doc); + value = Py_None; /* Mark as deleted */ + Py_INCREF(value); + op->func_doc = value; Py_XDECREF(tmp); return 0; } static PyObject * __Pyx_CyFunction_get_name(__pyx_CyFunctionObject *op) { - if (op->func_name == NULL) { + if (unlikely(op->func_name == NULL)) { #if PY_MAJOR_VERSION >= 3 op->func_name = PyUnicode_InternFromString(op->func.m_ml->ml_name); #else op->func_name = PyString_InternFromString(op->func.m_ml->ml_name); #endif + if (unlikely(op->func_name == NULL)) + return NULL; } Py_INCREF(op->func_name); return op->func_name; @@ -30573,9 +30610,9 @@ __Pyx_CyFunction_set_name(__pyx_CyFunctionObject *op, PyObject *value) { PyObject *tmp; #if PY_MAJOR_VERSION >= 3 - if (value == NULL || !PyUnicode_Check(value)) { + if (unlikely(value == NULL || !PyUnicode_Check(value))) { #else - if (value == NULL || !PyString_Check(value)) { + if (unlikely(value == NULL || !PyString_Check(value))) { #endif PyErr_SetString(PyExc_TypeError, "__name__ must be set to a string object"); @@ -30588,6 +30625,31 @@ __Pyx_CyFunction_set_name(__pyx_CyFunctionObject *op, PyObject *value) return 0; } static PyObject * +__Pyx_CyFunction_get_qualname(__pyx_CyFunctionObject *op) +{ + Py_INCREF(op->func_qualname); + return op->func_qualname; +} +static int +__Pyx_CyFunction_set_qualname(__pyx_CyFunctionObject *op, PyObject *value) +{ + PyObject *tmp; +#if PY_MAJOR_VERSION >= 3 + if (unlikely(value == NULL || !PyUnicode_Check(value))) { +#else + if (unlikely(value == NULL || !PyString_Check(value))) { +#endif + PyErr_SetString(PyExc_TypeError, + "__qualname__ must be set to a string object"); + return -1; + } + tmp = op->func_qualname; + Py_INCREF(value); + op->func_qualname = value; + Py_XDECREF(tmp); + return 0; +} +static PyObject * __Pyx_CyFunction_get_self(__pyx_CyFunctionObject *m, CYTHON_UNUSED void *closure) { PyObject *self; @@ -30600,9 +30662,9 @@ __Pyx_CyFunction_get_self(__pyx_CyFunctionObject *m, CYTHON_UNUSED void *closure static PyObject * __Pyx_CyFunction_get_dict(__pyx_CyFunctionObject *op) { - if (op->func_dict == NULL) { + if (unlikely(op->func_dict == NULL)) { op->func_dict = PyDict_New(); - if (op->func_dict == NULL) + if (unlikely(op->func_dict == NULL)) return NULL; } Py_INCREF(op->func_dict); @@ -30612,12 +30674,12 @@ static int __Pyx_CyFunction_set_dict(__pyx_CyFunctionObject *op, PyObject *value) { PyObject *tmp; - if (value == NULL) { + if (unlikely(value == NULL)) { PyErr_SetString(PyExc_TypeError, "function's dictionary may not be deleted"); return -1; } - if (!PyDict_Check(value)) { + if (unlikely(!PyDict_Check(value))) { PyErr_SetString(PyExc_TypeError, "setting function's dictionary to a non-dict"); return -1; @@ -30657,7 +30719,7 @@ __Pyx_CyFunction_get_defaults(__pyx_CyFunctionObject *op) } if (op->defaults_getter) { PyObject *res = op->defaults_getter((PyObject *) op); - if (res) { + if (likely(res)) { Py_INCREF(res); op->defaults_tuple = res; } @@ -30671,6 +30733,7 @@ static PyGetSetDef __pyx_CyFunction_getsets[] = { {(char *) "__doc__", (getter)__Pyx_CyFunction_get_doc, (setter)__Pyx_CyFunction_set_doc, 0, 0}, {(char *) "func_name", (getter)__Pyx_CyFunction_get_name, (setter)__Pyx_CyFunction_set_name, 0, 0}, {(char *) "__name__", (getter)__Pyx_CyFunction_get_name, (setter)__Pyx_CyFunction_set_name, 0, 0}, + {(char *) "__qualname__", (getter)__Pyx_CyFunction_get_qualname, (setter)__Pyx_CyFunction_set_qualname, 0, 0}, {(char *) "__self__", (getter)__Pyx_CyFunction_get_self, 0, 0, 0}, {(char *) "func_dict", (getter)__Pyx_CyFunction_get_dict, (setter)__Pyx_CyFunction_set_dict, 0, 0}, {(char *) "__dict__", (getter)__Pyx_CyFunction_get_dict, (setter)__Pyx_CyFunction_set_dict, 0, 0}, @@ -30704,7 +30767,7 @@ static PyMethodDef __pyx_CyFunction_methods[] = { {__Pyx_NAMESTR("__reduce__"), (PyCFunction)__Pyx_CyFunction_reduce, METH_VARARGS, 0}, {0, 0, 0, 0} }; -static PyObject *__Pyx_CyFunction_New(PyTypeObject *type, PyMethodDef *ml, int flags, +static PyObject *__Pyx_CyFunction_New(PyTypeObject *type, PyMethodDef *ml, int flags, PyObject* qualname, PyObject *closure, PyObject *module, PyObject* code) { __pyx_CyFunctionObject *op = PyObject_GC_New(__pyx_CyFunctionObject, type); if (op == NULL) @@ -30719,6 +30782,8 @@ static PyObject *__Pyx_CyFunction_New(PyTypeObject *type, PyMethodDef *ml, int f op->func.m_module = module; op->func_dict = NULL; op->func_name = NULL; + Py_INCREF(qualname); + op->func_qualname = qualname; op->func_doc = NULL; op->func_classobj = NULL; Py_XINCREF(code); @@ -30737,6 +30802,7 @@ __Pyx_CyFunction_clear(__pyx_CyFunctionObject *m) Py_CLEAR(m->func.m_module); Py_CLEAR(m->func_dict); Py_CLEAR(m->func_name); + Py_CLEAR(m->func_qualname); Py_CLEAR(m->func_doc); Py_CLEAR(m->func_code); Py_CLEAR(m->func_classobj); @@ -30765,6 +30831,7 @@ static int __Pyx_CyFunction_traverse(__pyx_CyFunctionObject *m, visitproc visit, Py_VISIT(m->func.m_module); Py_VISIT(m->func_dict); Py_VISIT(m->func_name); + Py_VISIT(m->func_qualname); Py_VISIT(m->func_doc); Py_VISIT(m->func_code); Py_VISIT(m->func_classobj); @@ -30797,13 +30864,12 @@ static PyObject *__Pyx_CyFunction_descr_get(PyObject *func, PyObject *obj, PyObj static PyObject* __Pyx_CyFunction_repr(__pyx_CyFunctionObject *op) { - PyObject *func_name = __Pyx_CyFunction_get_name(op); #if PY_MAJOR_VERSION >= 3 return PyUnicode_FromFormat("<cyfunction %U at %p>", - func_name, (void *)op); + op->func_qualname, (void *)op); #else return PyString_FromFormat("<cyfunction %s at %p>", - PyString_AsString(func_name), (void *)op); + PyString_AsString(op->func_qualname), (void *)op); #endif } #if CYTHON_COMPILING_IN_PYPY @@ -31936,6 +32002,10 @@ static PyTypeObject *__Pyx_ImportType(const char *module_name, const char *class PyObject *result = 0; PyObject *py_name = 0; char warning[200]; + Py_ssize_t basicsize; +#ifdef Py_LIMITED_API + PyObject *py_basicsize; +#endif py_module = __Pyx_ImportModule(module_name); if (!py_module) goto bad; @@ -31955,7 +32025,19 @@ static PyTypeObject *__Pyx_ImportType(const char *module_name, const char *class module_name, class_name); goto bad; } - if (!strict && (size_t)((PyTypeObject *)result)->tp_basicsize > size) { +#ifndef Py_LIMITED_API + basicsize = ((PyTypeObject *)result)->tp_basicsize; +#else + py_basicsize = PyObject_GetAttrString(result, "__basicsize__"); + if (!py_basicsize) + goto bad; + basicsize = PyLong_AsSsize_t(py_basicsize); + Py_DECREF(py_basicsize); + py_basicsize = 0; + if (basicsize == (Py_ssize_t)-1 && PyErr_Occurred()) + goto bad; +#endif + if (!strict && (size_t)basicsize > size) { PyOS_snprintf(warning, sizeof(warning), "%s.%s size changed, may indicate binary incompatibility", module_name, class_name); @@ -31965,7 +32047,7 @@ static PyTypeObject *__Pyx_ImportType(const char *module_name, const char *class if (PyErr_WarnEx(NULL, warning, 0) < 0) goto bad; #endif } - else if ((size_t)((PyTypeObject *)result)->tp_basicsize != size) { + else if ((size_t)basicsize != size) { PyErr_Format(PyExc_ValueError, "%s.%s has the wrong size, try recompiling", module_name, class_name); @@ -32243,15 +32325,11 @@ static int __Pyx_InitStrings(__Pyx_StringTabEntry *t) { return 0; } - -/* Type Conversion Functions */ - static CYTHON_INLINE int __Pyx_PyObject_IsTrue(PyObject* x) { int is_true = x == Py_True; if (is_true | (x == Py_False) | (x == Py_None)) return is_true; else return PyObject_IsTrue(x); } - static CYTHON_INLINE PyObject* __Pyx_PyNumber_Int(PyObject* x) { PyNumberMethods *m; const char *name = NULL; @@ -32297,7 +32375,6 @@ static CYTHON_INLINE PyObject* __Pyx_PyNumber_Int(PyObject* x) { } return res; } - static CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject* b) { Py_ssize_t ival; PyObject* x = PyNumber_Index(b); @@ -32306,7 +32383,6 @@ static CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject* b) { Py_DECREF(x); return ival; } - static CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t ival) { #if PY_VERSION_HEX < 0x02050000 if (ival <= LONG_MAX) @@ -32320,7 +32396,6 @@ static CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t ival) { return PyInt_FromSize_t(ival); #endif } - static CYTHON_INLINE size_t __Pyx_PyInt_AsSize_t(PyObject* x) { unsigned PY_LONG_LONG val = __Pyx_PyInt_AsUnsignedLongLong(x); if (unlikely(val == (unsigned PY_LONG_LONG)-1 && PyErr_Occurred())) { diff --git a/python/src/mteval.pxi b/python/src/mteval.pxi index f3bec393..436a1e01 100644 --- a/python/src/mteval.pxi +++ b/python/src/mteval.pxi @@ -192,5 +192,7 @@ cdef class Metric: return [] BLEU = Scorer('IBM_BLEU') +QCRI = Scorer('QCRI_BLEU') TER = Scorer('TER') CER = Scorer('CER') +SSK = Scorer('SSK') diff --git a/training/Makefile.am b/training/Makefile.am index e95e045f..8ef3c939 100644 --- a/training/Makefile.am +++ b/training/Makefile.am @@ -6,6 +6,7 @@ SUBDIRS = \ dpmert \ pro \ dtrain \ + latent_svm \ mira \ rampion diff --git a/training/dtrain/README.md b/training/dtrain/README.md index 7edabbf1..2ab2f232 100644 --- a/training/dtrain/README.md +++ b/training/dtrain/README.md @@ -13,36 +13,18 @@ Builds when building cdec, see ../BUILDING . To build only parts needed for dtrain do ``` autoreconf -ifv - ./configure [--disable-gtest] - cd dtrain/; make + ./configure + cd training/dtrain/; make ``` Running ------- -To run this on a dev set locally: -``` - #define DTRAIN_LOCAL -``` -otherwise remove that line or undef, then recompile. You need a single -grammar file or input annotated with per-sentence grammars (psg) as you -would use with cdec. Additionally you need to give dtrain a file with -references (--refs) when running locally. - -The input for use with hadoop streaming looks like this: -``` - <sid>\t<source>\t<ref>\t<grammar rules separated by \t> -``` -To convert a psg to this format you need to replace all "\n" -by "\t". Make sure there are no tabs in your data. - -For an example of local usage (with the 'distributed' format) -the see test/example/ . This expects dtrain to be built without -DTRAIN_LOCAL. +See directories under test/ . Legal ----- -Copyright (c) 2012 by Patrick Simianer <p@simianer.de> +Copyright (c) 2012-2013 by Patrick Simianer <p@simianer.de> -See the file ../LICENSE.txt for the licensing terms that this software is +See the file LICENSE.txt in the root folder for the licensing terms that this software is released under. diff --git a/training/dtrain/dtrain.cc b/training/dtrain/dtrain.cc index 18286668..149f87d4 100644 --- a/training/dtrain/dtrain.cc +++ b/training/dtrain/dtrain.cc @@ -6,15 +6,14 @@ dtrain_init(int argc, char** argv, po::variables_map* cfg) { po::options_description ini("Configuration File Options"); ini.add_options() - ("input", po::value<string>()->default_value("-"), "input file") + ("input", po::value<string>()->default_value("-"), "input file (src)") + ("refs,r", po::value<string>(), "references") ("output", po::value<string>()->default_value("-"), "output weights file, '-' for STDOUT") ("input_weights", po::value<string>(), "input weights file (e.g. from previous iteration)") ("decoder_config", po::value<string>(), "configuration file for cdec") ("print_weights", po::value<string>(), "weights to print on each iteration") ("stop_after", po::value<unsigned>()->default_value(0), "stop after X input sentences") - ("tmp", po::value<string>()->default_value("/tmp"), "temp dir to use") ("keep", po::value<bool>()->zero_tokens(), "keep weights files for each iteration") - ("hstreaming", po::value<string>(), "run in hadoop streaming mode, arg is a task id") ("epochs", po::value<unsigned>()->default_value(10), "# of iterations T (per shard)") ("k", po::value<unsigned>()->default_value(100), "how many translations to sample") ("sample_from", po::value<string>()->default_value("kbest"), "where to sample translations from: 'kbest', 'forest'") @@ -28,16 +27,13 @@ dtrain_init(int argc, char** argv, po::variables_map* cfg) ("gamma", po::value<weight_t>()->default_value(0.), "gamma for SVM (0 for perceptron)") ("select_weights", po::value<string>()->default_value("last"), "output best, last, avg weights ('VOID' to throw away)") ("rescale", po::value<bool>()->zero_tokens(), "rescale weight vector after each input") - ("l1_reg", po::value<string>()->default_value("none"), "apply l1 regularization as in 'Tsuroka et al' (2010)") + ("l1_reg", po::value<string>()->default_value("none"), "apply l1 regularization as in 'Tsuroka et al' (2010) UNTESTED") ("l1_reg_strength", po::value<weight_t>(), "l1 regularization strength") ("fselect", po::value<weight_t>()->default_value(-1), "select top x percent (or by threshold) of features after each epoch NOT IMPLEMENTED") // TODO ("approx_bleu_d", po::value<score_t>()->default_value(0.9), "discount for approx. BLEU") ("scale_bleu_diff", po::value<bool>()->zero_tokens(), "learning rate <- bleu diff of a misranked pair") ("loss_margin", po::value<weight_t>()->default_value(0.), "update if no error in pref pair but model scores this near") ("max_pairs", po::value<unsigned>()->default_value(std::numeric_limits<unsigned>::max()), "max. # of pairs per Sent.") -#ifdef DTRAIN_LOCAL - ("refs,r", po::value<string>(), "references in local mode") -#endif ("noup", po::value<bool>()->zero_tokens(), "do not update weights"); po::options_description cl("Command Line Options"); cl.add_options() @@ -55,16 +51,6 @@ dtrain_init(int argc, char** argv, po::variables_map* cfg) cerr << cl << endl; return false; } - if (cfg->count("hstreaming") && (*cfg)["output"].as<string>() != "-") { - cerr << "When using 'hstreaming' the 'output' param should be '-'." << endl; - return false; - } -#ifdef DTRAIN_LOCAL - if ((*cfg)["input"].as<string>() == "-") { - cerr << "Can't use stdin as input with this binary. Recompile without DTRAIN_LOCAL" << endl; - return false; - } -#endif if ((*cfg)["sample_from"].as<string>() != "kbest" && (*cfg)["sample_from"].as<string>() != "forest") { cerr << "Wrong 'sample_from' param: '" << (*cfg)["sample_from"].as<string>() << "', use 'kbest' or 'forest'." << endl; @@ -111,17 +97,8 @@ main(int argc, char** argv) if (cfg.count("verbose")) verbose = true; bool noup = false; if (cfg.count("noup")) noup = true; - bool hstreaming = false; - string task_id; - if (cfg.count("hstreaming")) { - hstreaming = true; - quiet = true; - task_id = cfg["hstreaming"].as<string>(); - cerr.precision(17); - } bool rescale = false; if (cfg.count("rescale")) rescale = true; - HSReporter rep(task_id); bool keep = false; if (cfg.count("keep")) keep = true; @@ -148,6 +125,7 @@ main(int argc, char** argv) if (cfg.count("print_weights")) boost::split(print_weights, cfg["print_weights"].as<string>(), boost::is_any_of(" ")); + // setup decoder register_feature_functions(); SetSilent(true); @@ -163,6 +141,8 @@ main(int argc, char** argv) scorer = dynamic_cast<BleuScorer*>(new BleuScorer); } else if (scorer_str == "stupid_bleu") { scorer = dynamic_cast<StupidBleuScorer*>(new StupidBleuScorer); + } else if (scorer_str == "fixed_stupid_bleu") { + scorer = dynamic_cast<FixedStupidBleuScorer*>(new FixedStupidBleuScorer); } else if (scorer_str == "smooth_bleu") { scorer = dynamic_cast<SmoothBleuScorer*>(new SmoothBleuScorer); } else if (scorer_str == "sum_bleu") { @@ -201,6 +181,11 @@ main(int argc, char** argv) weight_t eta = cfg["learning_rate"].as<weight_t>(); weight_t gamma = cfg["gamma"].as<weight_t>(); + // faster perceptron: consider only misranked pairs, see + // DO NOT ENABLE WITH SVM (gamma > 0) OR loss_margin! + bool faster_perceptron = false; + if (gamma==0 && loss_margin==0) faster_perceptron = true; + // l1 regularization bool l1naive = false; bool l1clip = false; @@ -222,16 +207,8 @@ main(int argc, char** argv) // buffer input for t > 0 vector<string> src_str_buf; // source strings (decoder takes only strings) vector<vector<WordID> > ref_ids_buf; // references as WordID vecs - // where temp files go - string tmp_path = cfg["tmp"].as<string>(); -#ifdef DTRAIN_LOCAL string refs_fn = cfg["refs"].as<string>(); ReadFile refs(refs_fn); -#else - string grammar_buf_fn = gettmpf(tmp_path, "dtrain-grammars"); - ogzstream grammar_buf_out; - grammar_buf_out.open(grammar_buf_fn.c_str()); -#endif unsigned in_sz = std::numeric_limits<unsigned>::max(); // input index, input size vector<pair<score_t, score_t> > all_scores; @@ -246,7 +223,7 @@ main(int argc, char** argv) cerr << setw(25) << "k " << k << endl; cerr << setw(25) << "N " << N << endl; cerr << setw(25) << "T " << T << endl; - cerr << setw(25) << "scorer '" << scorer_str << "'" << endl; + cerr << setw(26) << "scorer '" << scorer_str << "'" << endl; if (scorer_str == "approx_bleu") cerr << setw(25) << "approx. B discount " << approx_bleu_d << endl; cerr << setw(25) << "sample from " << "'" << sample_from << "'" << endl; @@ -256,6 +233,7 @@ main(int argc, char** argv) else cerr << setw(25) << "learning rate " << "bleu diff" << endl; cerr << setw(25) << "gamma " << gamma << endl; cerr << setw(25) << "loss margin " << loss_margin << endl; + cerr << setw(25) << "faster perceptron " << faster_perceptron << endl; cerr << setw(25) << "pairs " << "'" << pair_sampling << "'" << endl; if (pair_sampling == "XYX") cerr << setw(25) << "hi lo " << hi_lo << endl; @@ -268,9 +246,7 @@ main(int argc, char** argv) cerr << setw(25) << "max pairs " << max_pairs << endl; cerr << setw(25) << "cdec cfg " << "'" << cfg["decoder_config"].as<string>() << "'" << endl; cerr << setw(25) << "input " << "'" << input_fn << "'" << endl; -#ifdef DTRAIN_LOCAL cerr << setw(25) << "refs " << "'" << refs_fn << "'" << endl; -#endif cerr << setw(25) << "output " << "'" << output_fn << "'" << endl; if (cfg.count("input_weights")) cerr << setw(25) << "weights in " << "'" << cfg["input_weights"].as<string>() << "'" << endl; @@ -283,14 +259,8 @@ main(int argc, char** argv) for (unsigned t = 0; t < T; t++) // T epochs { - if (hstreaming) cerr << "reporter:status:Iteration #" << t+1 << " of " << T << endl; - time_t start, end; time(&start); -#ifndef DTRAIN_LOCAL - igzstream grammar_buf_in; - if (t > 0) grammar_buf_in.open(grammar_buf_fn.c_str()); -#endif score_t score_sum = 0.; score_t model_sum(0); unsigned ii = 0, rank_errors = 0, margin_violations = 0, npairs = 0, f_count = 0, list_sz = 0; @@ -338,52 +308,6 @@ main(int argc, char** argv) // getting input vector<WordID> ref_ids; // reference as vector<WordID> -#ifndef DTRAIN_LOCAL - vector<string> in_split; // input: sid\tsrc\tref\tpsg - if (t == 0) { - // handling input - split_in(in, in_split); - if (hstreaming && ii == 0) cerr << "reporter:counter:" << task_id << ",First ID," << in_split[0] << endl; - // getting reference - vector<string> ref_tok; - boost::split(ref_tok, in_split[2], boost::is_any_of(" ")); - register_and_convert(ref_tok, ref_ids); - ref_ids_buf.push_back(ref_ids); - // process and set grammar - bool broken_grammar = true; // ignore broken grammars - for (string::iterator it = in.begin(); it != in.end(); it++) { - if (!isspace(*it)) { - broken_grammar = false; - break; - } - } - if (broken_grammar) { - cerr << "Broken grammar for " << ii+1 << "! Ignoring this input." << endl; - continue; - } - boost::replace_all(in, "\t", "\n"); - in += "\n"; - grammar_buf_out << in << DTRAIN_GRAMMAR_DELIM << " " << in_split[0] << endl; - decoder.AddSupplementalGrammarFromString(in); - src_str_buf.push_back(in_split[1]); - // decode - observer->SetRef(ref_ids); - decoder.Decode(in_split[1], observer); - } else { - // get buffered grammar - string grammar_str; - while (true) { - string rule; - getline(grammar_buf_in, rule); - if (boost::starts_with(rule, DTRAIN_GRAMMAR_DELIM)) break; - grammar_str += rule + "\n"; - } - decoder.AddSupplementalGrammarFromString(grammar_str); - // decode - observer->SetRef(ref_ids_buf[ii]); - decoder.Decode(src_str_buf[ii], observer); - } -#else if (t == 0) { string r_; getline(*refs, r_); @@ -400,7 +324,6 @@ main(int argc, char** argv) decoder.Decode(in, observer); else decoder.Decode(src_str_buf[ii], observer); -#endif // get (scored) samples vector<ScoredHyp>* samples = observer->GetSamples(); @@ -430,25 +353,26 @@ main(int argc, char** argv) // get pairs vector<pair<ScoredHyp,ScoredHyp> > pairs; if (pair_sampling == "all") - all_pairs(samples, pairs, pair_threshold, max_pairs); + all_pairs(samples, pairs, pair_threshold, max_pairs, faster_perceptron); if (pair_sampling == "XYX") - partXYX(samples, pairs, pair_threshold, max_pairs, hi_lo); + partXYX(samples, pairs, pair_threshold, max_pairs, faster_perceptron, hi_lo); if (pair_sampling == "PRO") PROsampling(samples, pairs, pair_threshold, max_pairs); npairs += pairs.size(); for (vector<pair<ScoredHyp,ScoredHyp> >::iterator it = pairs.begin(); it != pairs.end(); it++) { -#ifdef DTRAIN_FASTER_PERCEPTRON - bool rank_error = true; // pair sampling already did this for us - rank_errors++; - score_t margin = std::numeric_limits<float>::max(); -#else - bool rank_error = it->first.model <= it->second.model; + bool rank_error; + score_t margin; + if (faster_perceptron) { // we only have considering misranked pairs + rank_error = true; // pair sampling already did this for us + margin = std::numeric_limits<float>::max(); + } else { + rank_error = it->first.model <= it->second.model; + margin = fabs(fabs(it->first.model) - fabs(it->second.model)); + if (!rank_error && margin < loss_margin) margin_violations++; + } if (rank_error) rank_errors++; - score_t margin = fabs(fabs(it->first.model) - fabs(it->second.model)); - if (!rank_error && margin < loss_margin) margin_violations++; -#endif if (scale_bleu_diff) eta = it->first.score - it->second.score; if (rank_error || margin < loss_margin) { SparseVector<weight_t> diff_vec = it->first.f - it->second.f; @@ -459,35 +383,40 @@ main(int argc, char** argv) } // l1 regularization + // please note that this penalizes _all_ weights + // (contrary to only the ones changed by the last update) + // after a _sentence_ (not after each example/pair) if (l1naive) { - for (unsigned d = 0; d < lambdas.size(); d++) { - weight_t v = lambdas.get(d); - lambdas.set_value(d, v - sign(v) * l1_reg); + FastSparseVector<weight_t>::iterator it = lambdas.begin(); + for (; it != lambdas.end(); ++it) { + it->second -= sign(it->second) * l1_reg; } } else if (l1clip) { - for (unsigned d = 0; d < lambdas.size(); d++) { - if (lambdas.nonzero(d)) { - weight_t v = lambdas.get(d); + FastSparseVector<weight_t>::iterator it = lambdas.begin(); + for (; it != lambdas.end(); ++it) { + if (it->second != 0) { + weight_t v = it->second; if (v > 0) { - lambdas.set_value(d, max(0., v - l1_reg)); + it->second = max(0., v - l1_reg); } else { - lambdas.set_value(d, min(0., v + l1_reg)); + it->second = min(0., v + l1_reg); } } } } else if (l1cumul) { weight_t acc_penalty = (ii+1) * l1_reg; // ii is the index of the current input - for (unsigned d = 0; d < lambdas.size(); d++) { - if (lambdas.nonzero(d)) { - weight_t v = lambdas.get(d); - weight_t penalty = 0; + FastSparseVector<weight_t>::iterator it = lambdas.begin(); + for (; it != lambdas.end(); ++it) { + if (it->second != 0) { + weight_t v = it->second; + weight_t penalized = 0.; if (v > 0) { - penalty = max(0., v-(acc_penalty + cumulative_penalties.get(d))); + penalized = max(0., v-(acc_penalty + cumulative_penalties.get(it->first))); } else { - penalty = min(0., v+(acc_penalty - cumulative_penalties.get(d))); + penalized = min(0., v+(acc_penalty - cumulative_penalties.get(it->first))); } - lambdas.set_value(d, penalty); - cumulative_penalties.set_value(d, cumulative_penalties.get(d)+penalty); + it->second = penalized; + cumulative_penalties.set_value(it->first, cumulative_penalties.get(it->first)+penalized); } } } @@ -498,11 +427,6 @@ main(int argc, char** argv) ++ii; - if (hstreaming) { - rep.update_counter("Seen #"+boost::lexical_cast<string>(t+1), 1u); - rep.update_counter("Seen", 1u); - } - } // input loop if (average) w_average += lambdas; @@ -511,21 +435,8 @@ main(int argc, char** argv) if (t == 0) { in_sz = ii; // remember size of input (# lines) - if (hstreaming) { - rep.update_counter("|Input|", ii); - rep.update_gcounter("|Input|", ii); - rep.update_gcounter("Shards", 1u); - } } -#ifndef DTRAIN_LOCAL - if (t == 0) { - grammar_buf_out.close(); - } else { - grammar_buf_in.close(); - } -#endif - // print some stats score_t score_avg = score_sum/(score_t)in_sz; score_t model_avg = model_sum/(score_t)in_sz; @@ -539,7 +450,7 @@ main(int argc, char** argv) } unsigned nonz = 0; - if (!quiet || hstreaming) nonz = (unsigned)lambdas.num_nonzero(); + if (!quiet) nonz = (unsigned)lambdas.num_nonzero(); if (!quiet) { cerr << _p5 << _p << "WEIGHTS" << endl; @@ -552,28 +463,18 @@ main(int argc, char** argv) cerr << _np << " 1best avg model score: " << model_avg; cerr << _p << " (" << model_diff << ")" << endl; cerr << " avg # pairs: "; - cerr << _np << npairs/(float)in_sz << endl; + cerr << _np << npairs/(float)in_sz; + if (faster_perceptron) cerr << " (meaningless)"; + cerr << endl; cerr << " avg # rank err: "; cerr << rank_errors/(float)in_sz << endl; -#ifndef DTRAIN_FASTER_PERCEPTRON cerr << " avg # margin viol: "; cerr << margin_violations/(float)in_sz << endl; -#endif cerr << " non0 feature count: " << nonz << endl; cerr << " avg list sz: " << list_sz/(float)in_sz << endl; cerr << " avg f count: " << f_count/(float)list_sz << endl; } - if (hstreaming) { - rep.update_counter("Score 1best avg #"+boost::lexical_cast<string>(t+1), (unsigned)(score_avg*DTRAIN_SCALE)); - rep.update_counter("Model 1best avg #"+boost::lexical_cast<string>(t+1), (unsigned)(model_avg*DTRAIN_SCALE)); - rep.update_counter("Pairs avg #"+boost::lexical_cast<string>(t+1), (unsigned)((npairs/(weight_t)in_sz)*DTRAIN_SCALE)); - rep.update_counter("Rank errors avg #"+boost::lexical_cast<string>(t+1), (unsigned)((rank_errors/(weight_t)in_sz)*DTRAIN_SCALE)); - rep.update_counter("Margin violations avg #"+boost::lexical_cast<string>(t+1), (unsigned)((margin_violations/(weight_t)in_sz)*DTRAIN_SCALE)); - rep.update_counter("Non zero feature count #"+boost::lexical_cast<string>(t+1), nonz); - rep.update_gcounter("Non zero feature count #"+boost::lexical_cast<string>(t+1), nonz); - } - pair<score_t,score_t> remember; remember.first = score_avg; remember.second = model_avg; @@ -604,10 +505,6 @@ main(int argc, char** argv) if (average) w_average /= (weight_t)T; -#ifndef DTRAIN_LOCAL - unlink(grammar_buf_fn.c_str()); -#endif - if (!noup) { if (!quiet) cerr << endl << "Writing weights file to '" << output_fn << "' ..." << endl; if (select_weights == "last" || average) { // last, average @@ -644,7 +541,6 @@ main(int argc, char** argv) } } } - if (output_fn == "-" && hstreaming) cout << "__SHARD_COUNT__\t1" << endl; if (!quiet) cerr << "done" << endl; } diff --git a/training/dtrain/dtrain.h b/training/dtrain/dtrain.h index 4b6f415c..eb0b9f17 100644 --- a/training/dtrain/dtrain.h +++ b/training/dtrain/dtrain.h @@ -1,14 +1,9 @@ #ifndef _DTRAIN_H_ #define _DTRAIN_H_ -#undef DTRAIN_FASTER_PERCEPTRON // only look at misranked pairs - // DO NOT USE WITH SVM! -//#define DTRAIN_LOCAL #define DTRAIN_DOTS 10 // after how many inputs to display a '.' -#define DTRAIN_GRAMMAR_DELIM "########EOS########" #define DTRAIN_SCALE 100000 - #include <iomanip> #include <climits> #include <string.h> diff --git a/training/dtrain/examples/parallelized/README b/training/dtrain/examples/parallelized/README new file mode 100644 index 00000000..89715105 --- /dev/null +++ b/training/dtrain/examples/parallelized/README @@ -0,0 +1,5 @@ +run for example + ../../parallelize.rb ./dtrain.ini 4 false 2 2 ./in ./refs + +final weights will be in the file work/weights.3 + diff --git a/training/dtrain/test/parallelize/cdec.ini b/training/dtrain/examples/parallelized/cdec.ini index 72e99dc5..e43ba1c4 100644 --- a/training/dtrain/test/parallelize/cdec.ini +++ b/training/dtrain/examples/parallelized/cdec.ini @@ -4,7 +4,7 @@ intersection_strategy=cube_pruning cubepruning_pop_limit=200 scfg_max_span_limit=15 feature_function=WordPenalty -feature_function=KLanguageModel /stor/dat/wmt12/en/news_only/m/wmt12.news.en.3.kenv5 +feature_function=KLanguageModel ../example/nc-wmt11.en.srilm.gz #feature_function=ArityPenalty #feature_function=CMR2008ReorderingFeatures #feature_function=Dwarf diff --git a/training/dtrain/test/parallelize/dtrain.ini b/training/dtrain/examples/parallelized/dtrain.ini index 03f9d240..f19ef891 100644 --- a/training/dtrain/test/parallelize/dtrain.ini +++ b/training/dtrain/examples/parallelized/dtrain.ini @@ -2,7 +2,7 @@ k=100 N=4 learning_rate=0.0001 gamma=0 -loss_margin=0 +loss_margin=1.0 epochs=1 scorer=stupid_bleu sample_from=kbest @@ -11,5 +11,6 @@ pair_sampling=XYX hi_lo=0.1 select_weights=last print_weights=Glue WordPenalty LanguageModel LanguageModel_OOV PhraseModel_0 PhraseModel_1 PhraseModel_2 PhraseModel_3 PhraseModel_4 PhraseModel_5 PhraseModel_6 PassThrough -tmp=/tmp +# newer version of the grammar extractor use different feature names: +#print_weights=Glue WordPenalty LanguageModel LanguageModel_OOV PhraseModel_0 PhraseModel_1 PhraseModel_2 PhraseModel_3 PhraseModel_4 PhraseModel_5 PhraseModel_6 PassThrough decoder_config=cdec.ini diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.0.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.0.gz Binary files differnew file mode 100644 index 00000000..1e28a24b --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.0.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.1.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.1.gz Binary files differnew file mode 100644 index 00000000..372f5675 --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.1.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.2.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.2.gz Binary files differnew file mode 100644 index 00000000..145d0dc0 --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.2.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.3.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.3.gz Binary files differnew file mode 100644 index 00000000..105593ff --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.3.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.4.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.4.gz Binary files differnew file mode 100644 index 00000000..30781f48 --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.4.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.5.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.5.gz Binary files differnew file mode 100644 index 00000000..834ee759 --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.5.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.6.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.6.gz Binary files differnew file mode 100644 index 00000000..2e76f348 --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.6.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.7.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.7.gz Binary files differnew file mode 100644 index 00000000..3741a887 --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.7.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.8.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.8.gz Binary files differnew file mode 100644 index 00000000..ebf6bd0c --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.8.gz diff --git a/training/dtrain/examples/parallelized/grammar/grammar.out.9.gz b/training/dtrain/examples/parallelized/grammar/grammar.out.9.gz Binary files differnew file mode 100644 index 00000000..c1791059 --- /dev/null +++ b/training/dtrain/examples/parallelized/grammar/grammar.out.9.gz diff --git a/training/dtrain/examples/parallelized/in b/training/dtrain/examples/parallelized/in new file mode 100644 index 00000000..51d01fe7 --- /dev/null +++ b/training/dtrain/examples/parallelized/in @@ -0,0 +1,10 @@ +<seg grammar="grammar/grammar.out.0.gz" id="0">europas nach rassen geteiltes haus</seg> +<seg grammar="grammar/grammar.out.1.gz" id="1">ein gemeinsames merkmal aller extremen rechten in europa ist ihr rassismus und die tatsache , daß sie das einwanderungsproblem als politischen hebel benutzen .</seg> +<seg grammar="grammar/grammar.out.2.gz" id="2">der lega nord in italien , der vlaams block in den niederlanden , die anhänger von le pens nationaler front in frankreich , sind beispiele für parteien oder bewegungen , die sich um das gemeinsame thema : ablehnung der zuwanderung gebildet haben und um forderung nach einer vereinfachten politik , um sie zu regeln .</seg> +<seg grammar="grammar/grammar.out.3.gz" id="3">während individuen wie jörg haidar und jean @-@ marie le pen kommen und ( leider nicht zu bald ) wieder gehen mögen , wird die rassenfrage aus der europäischer politik nicht so bald verschwinden .</seg> +<seg grammar="grammar/grammar.out.4.gz" id="4">eine alternde einheimische bevölkerung und immer offenere grenzen vermehren die rassistische zersplitterung in den europäischen ländern .</seg> +<seg grammar="grammar/grammar.out.5.gz" id="5">die großen parteien der rechten und der linken mitte haben sich dem problem gestellt , in dem sie den kopf in den sand gesteckt und allen aussichten zuwider gehofft haben , es möge bald verschwinden .</seg> +<seg grammar="grammar/grammar.out.6.gz" id="6">das aber wird es nicht , wie die geschichte des rassismus in amerika deutlich zeigt .</seg> +<seg grammar="grammar/grammar.out.7.gz" id="7">die beziehungen zwischen den rassen standen in den usa über jahrzehnte - und tun das noch heute - im zentrum der politischen debatte . das ging so weit , daß rassentrennung genauso wichtig wie das einkommen wurde , - wenn nicht sogar noch wichtiger - um politische zuneigungen und einstellungen zu bestimmen .</seg> +<seg grammar="grammar/grammar.out.8.gz" id="8">der erste schritt , um mit der rassenfrage umzugehen ist , ursache und folgen rassistischer feindseligkeiten zu verstehen , auch dann , wenn das bedeutet , unangenehme tatsachen aufzudecken .</seg> +<seg grammar="grammar/grammar.out.9.gz" id="9">genau das haben in den usa eine große anzahl an forschungsvorhaben in wirtschaft , soziologie , psychologie und politikwissenschaft geleistet . diese forschungen zeigten , daß menschen unterschiedlicher rasse einander deutlich weniger vertrauen .</seg> diff --git a/training/dtrain/examples/parallelized/refs b/training/dtrain/examples/parallelized/refs new file mode 100644 index 00000000..632e27b0 --- /dev/null +++ b/training/dtrain/examples/parallelized/refs @@ -0,0 +1,10 @@ +europe 's divided racial house +a common feature of europe 's extreme right is its racism and use of the immigration issue as a political wedge . +the lega nord in italy , the vlaams blok in the netherlands , the supporters of le pen 's national front in france , are all examples of parties or movements formed on the common theme of aversion to immigrants and promotion of simplistic policies to control them . +while individuals like jorg haidar and jean @-@ marie le pen may come and ( never to soon ) go , the race question will not disappear from european politics anytime soon . +an aging population at home and ever more open borders imply increasing racial fragmentation in european countries . +mainstream parties of the center left and center right have confronted this prospect by hiding their heads in the ground , hoping against hope that the problem will disappear . +it will not , as america 's racial history clearly shows . +race relations in the us have been for decades - and remain - at the center of political debate , to the point that racial cleavages are as important as income , if not more , as determinants of political preferences and attitudes . +the first step to address racial politics is to understand the origin and consequences of racial animosity , even if it means uncovering unpleasant truths . +this is precisely what a large amount of research in economics , sociology , psychology and political science has done for the us . diff --git a/training/dtrain/examples/parallelized/work/out.0.0 b/training/dtrain/examples/parallelized/work/out.0.0 new file mode 100644 index 00000000..7a00ed0f --- /dev/null +++ b/training/dtrain/examples/parallelized/work/out.0.0 @@ -0,0 +1,61 @@ + cdec cfg 'cdec.ini' +Loading the LM will be faster if you build a binary file. +Reading ../example/nc-wmt11.en.srilm.gz +----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100 +**************************************************************************************************** +Seeding random number sequence to 3121929377 + +dtrain +Parameters: + k 100 + N 4 + T 1 + scorer 'stupid_bleu' + sample from 'kbest' + filter 'uniq' + learning rate 0.0001 + gamma 0 + loss margin 1 + pairs 'XYX' + hi lo 0.1 + pair threshold 0 + select weights 'last' + l1 reg 0 'none' + max pairs 4294967295 + cdec cfg 'cdec.ini' + input 'work/shard.0.0.in' + refs 'work/shard.0.0.refs' + output 'work/weights.0.0' +(a dot represents 10 inputs) +Iteration #1 of 1. + 5 +WEIGHTS + Glue = +0.2663 + WordPenalty = -0.0079042 + LanguageModel = +0.44782 + LanguageModel_OOV = -0.0401 + PhraseModel_0 = -0.193 + PhraseModel_1 = +0.71321 + PhraseModel_2 = +0.85196 + PhraseModel_3 = -0.43986 + PhraseModel_4 = -0.44803 + PhraseModel_5 = -0.0538 + PhraseModel_6 = -0.1788 + PassThrough = -0.1477 + --- + 1best avg score: 0.17521 (+0.17521) + 1best avg model score: 21.556 (+21.556) + avg # pairs: 1671.2 + avg # rank err: 1118.6 + avg # margin viol: 552.6 + non0 feature count: 12 + avg list sz: 100 + avg f count: 11.32 +(time 0.37 min, 4.4 s/S) + +Writing weights file to 'work/weights.0.0' ... +done + +--- +Best iteration: 1 [SCORE 'stupid_bleu'=0.17521]. +This took 0.36667 min. diff --git a/training/dtrain/examples/parallelized/work/out.0.1 b/training/dtrain/examples/parallelized/work/out.0.1 new file mode 100644 index 00000000..e2bd6649 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/out.0.1 @@ -0,0 +1,62 @@ + cdec cfg 'cdec.ini' +Loading the LM will be faster if you build a binary file. +Reading ../example/nc-wmt11.en.srilm.gz +----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100 +**************************************************************************************************** +Seeding random number sequence to 2767202922 + +dtrain +Parameters: + k 100 + N 4 + T 1 + scorer 'stupid_bleu' + sample from 'kbest' + filter 'uniq' + learning rate 0.0001 + gamma 0 + loss margin 1 + pairs 'XYX' + hi lo 0.1 + pair threshold 0 + select weights 'last' + l1 reg 0 'none' + max pairs 4294967295 + cdec cfg 'cdec.ini' + input 'work/shard.0.0.in' + refs 'work/shard.0.0.refs' + output 'work/weights.0.1' + weights in 'work/weights.0' +(a dot represents 10 inputs) +Iteration #1 of 1. + 5 +WEIGHTS + Glue = -0.2699 + WordPenalty = +0.080605 + LanguageModel = -0.026572 + LanguageModel_OOV = -0.30025 + PhraseModel_0 = -0.32076 + PhraseModel_1 = +0.67451 + PhraseModel_2 = +0.92 + PhraseModel_3 = -0.36402 + PhraseModel_4 = -0.592 + PhraseModel_5 = -0.0269 + PhraseModel_6 = -0.28755 + PassThrough = -0.33285 + --- + 1best avg score: 0.26638 (+0.26638) + 1best avg model score: 53.197 (+53.197) + avg # pairs: 2028.6 + avg # rank err: 998.2 + avg # margin viol: 918.8 + non0 feature count: 12 + avg list sz: 100 + avg f count: 10.496 +(time 0.32 min, 3.8 s/S) + +Writing weights file to 'work/weights.0.1' ... +done + +--- +Best iteration: 1 [SCORE 'stupid_bleu'=0.26638]. +This took 0.31667 min. diff --git a/training/dtrain/examples/parallelized/work/out.1.0 b/training/dtrain/examples/parallelized/work/out.1.0 new file mode 100644 index 00000000..6e790e38 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/out.1.0 @@ -0,0 +1,61 @@ + cdec cfg 'cdec.ini' +Loading the LM will be faster if you build a binary file. +Reading ../example/nc-wmt11.en.srilm.gz +----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100 +**************************************************************************************************** +Seeding random number sequence to 1432415010 + +dtrain +Parameters: + k 100 + N 4 + T 1 + scorer 'stupid_bleu' + sample from 'kbest' + filter 'uniq' + learning rate 0.0001 + gamma 0 + loss margin 1 + pairs 'XYX' + hi lo 0.1 + pair threshold 0 + select weights 'last' + l1 reg 0 'none' + max pairs 4294967295 + cdec cfg 'cdec.ini' + input 'work/shard.1.0.in' + refs 'work/shard.1.0.refs' + output 'work/weights.1.0' +(a dot represents 10 inputs) +Iteration #1 of 1. + 5 +WEIGHTS + Glue = -0.3815 + WordPenalty = +0.20064 + LanguageModel = +0.95304 + LanguageModel_OOV = -0.264 + PhraseModel_0 = -0.22362 + PhraseModel_1 = +0.12254 + PhraseModel_2 = +0.26328 + PhraseModel_3 = +0.38018 + PhraseModel_4 = -0.48654 + PhraseModel_5 = +0 + PhraseModel_6 = -0.3645 + PassThrough = -0.2216 + --- + 1best avg score: 0.10863 (+0.10863) + 1best avg model score: -4.9841 (-4.9841) + avg # pairs: 1345.4 + avg # rank err: 822.4 + avg # margin viol: 501 + non0 feature count: 11 + avg list sz: 100 + avg f count: 11.814 +(time 0.45 min, 5.4 s/S) + +Writing weights file to 'work/weights.1.0' ... +done + +--- +Best iteration: 1 [SCORE 'stupid_bleu'=0.10863]. +This took 0.45 min. diff --git a/training/dtrain/examples/parallelized/work/out.1.1 b/training/dtrain/examples/parallelized/work/out.1.1 new file mode 100644 index 00000000..0b984761 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/out.1.1 @@ -0,0 +1,62 @@ + cdec cfg 'cdec.ini' +Loading the LM will be faster if you build a binary file. +Reading ../example/nc-wmt11.en.srilm.gz +----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100 +**************************************************************************************************** +Seeding random number sequence to 1771918374 + +dtrain +Parameters: + k 100 + N 4 + T 1 + scorer 'stupid_bleu' + sample from 'kbest' + filter 'uniq' + learning rate 0.0001 + gamma 0 + loss margin 1 + pairs 'XYX' + hi lo 0.1 + pair threshold 0 + select weights 'last' + l1 reg 0 'none' + max pairs 4294967295 + cdec cfg 'cdec.ini' + input 'work/shard.1.0.in' + refs 'work/shard.1.0.refs' + output 'work/weights.1.1' + weights in 'work/weights.0' +(a dot represents 10 inputs) +Iteration #1 of 1. + 5 +WEIGHTS + Glue = -0.3178 + WordPenalty = +0.11092 + LanguageModel = +0.17269 + LanguageModel_OOV = -0.13485 + PhraseModel_0 = -0.45371 + PhraseModel_1 = +0.38789 + PhraseModel_2 = +0.75311 + PhraseModel_3 = -0.38163 + PhraseModel_4 = -0.58817 + PhraseModel_5 = -0.0269 + PhraseModel_6 = -0.27315 + PassThrough = -0.16745 + --- + 1best avg score: 0.13169 (+0.13169) + 1best avg model score: 24.226 (+24.226) + avg # pairs: 1951.2 + avg # rank err: 985.4 + avg # margin viol: 951 + non0 feature count: 12 + avg list sz: 100 + avg f count: 11.224 +(time 0.42 min, 5 s/S) + +Writing weights file to 'work/weights.1.1' ... +done + +--- +Best iteration: 1 [SCORE 'stupid_bleu'=0.13169]. +This took 0.41667 min. diff --git a/training/dtrain/examples/parallelized/work/shard.0.0.in b/training/dtrain/examples/parallelized/work/shard.0.0.in new file mode 100644 index 00000000..92f9c78e --- /dev/null +++ b/training/dtrain/examples/parallelized/work/shard.0.0.in @@ -0,0 +1,5 @@ +<seg grammar="grammar/grammar.out.0.gz" id="0">europas nach rassen geteiltes haus</seg> +<seg grammar="grammar/grammar.out.1.gz" id="1">ein gemeinsames merkmal aller extremen rechten in europa ist ihr rassismus und die tatsache , daß sie das einwanderungsproblem als politischen hebel benutzen .</seg> +<seg grammar="grammar/grammar.out.2.gz" id="2">der lega nord in italien , der vlaams block in den niederlanden , die anhänger von le pens nationaler front in frankreich , sind beispiele für parteien oder bewegungen , die sich um das gemeinsame thema : ablehnung der zuwanderung gebildet haben und um forderung nach einer vereinfachten politik , um sie zu regeln .</seg> +<seg grammar="grammar/grammar.out.3.gz" id="3">während individuen wie jörg haidar und jean @-@ marie le pen kommen und ( leider nicht zu bald ) wieder gehen mögen , wird die rassenfrage aus der europäischer politik nicht so bald verschwinden .</seg> +<seg grammar="grammar/grammar.out.4.gz" id="4">eine alternde einheimische bevölkerung und immer offenere grenzen vermehren die rassistische zersplitterung in den europäischen ländern .</seg> diff --git a/training/dtrain/examples/parallelized/work/shard.0.0.refs b/training/dtrain/examples/parallelized/work/shard.0.0.refs new file mode 100644 index 00000000..bef68fee --- /dev/null +++ b/training/dtrain/examples/parallelized/work/shard.0.0.refs @@ -0,0 +1,5 @@ +europe 's divided racial house +a common feature of europe 's extreme right is its racism and use of the immigration issue as a political wedge . +the lega nord in italy , the vlaams blok in the netherlands , the supporters of le pen 's national front in france , are all examples of parties or movements formed on the common theme of aversion to immigrants and promotion of simplistic policies to control them . +while individuals like jorg haidar and jean @-@ marie le pen may come and ( never to soon ) go , the race question will not disappear from european politics anytime soon . +an aging population at home and ever more open borders imply increasing racial fragmentation in european countries . diff --git a/training/dtrain/examples/parallelized/work/shard.1.0.in b/training/dtrain/examples/parallelized/work/shard.1.0.in new file mode 100644 index 00000000..b7695ce7 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/shard.1.0.in @@ -0,0 +1,5 @@ +<seg grammar="grammar/grammar.out.5.gz" id="5">die großen parteien der rechten und der linken mitte haben sich dem problem gestellt , in dem sie den kopf in den sand gesteckt und allen aussichten zuwider gehofft haben , es möge bald verschwinden .</seg> +<seg grammar="grammar/grammar.out.6.gz" id="6">das aber wird es nicht , wie die geschichte des rassismus in amerika deutlich zeigt .</seg> +<seg grammar="grammar/grammar.out.7.gz" id="7">die beziehungen zwischen den rassen standen in den usa über jahrzehnte - und tun das noch heute - im zentrum der politischen debatte . das ging so weit , daß rassentrennung genauso wichtig wie das einkommen wurde , - wenn nicht sogar noch wichtiger - um politische zuneigungen und einstellungen zu bestimmen .</seg> +<seg grammar="grammar/grammar.out.8.gz" id="8">der erste schritt , um mit der rassenfrage umzugehen ist , ursache und folgen rassistischer feindseligkeiten zu verstehen , auch dann , wenn das bedeutet , unangenehme tatsachen aufzudecken .</seg> +<seg grammar="grammar/grammar.out.9.gz" id="9">genau das haben in den usa eine große anzahl an forschungsvorhaben in wirtschaft , soziologie , psychologie und politikwissenschaft geleistet . diese forschungen zeigten , daß menschen unterschiedlicher rasse einander deutlich weniger vertrauen .</seg> diff --git a/training/dtrain/examples/parallelized/work/shard.1.0.refs b/training/dtrain/examples/parallelized/work/shard.1.0.refs new file mode 100644 index 00000000..6076f6d5 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/shard.1.0.refs @@ -0,0 +1,5 @@ +mainstream parties of the center left and center right have confronted this prospect by hiding their heads in the ground , hoping against hope that the problem will disappear . +it will not , as america 's racial history clearly shows . +race relations in the us have been for decades - and remain - at the center of political debate , to the point that racial cleavages are as important as income , if not more , as determinants of political preferences and attitudes . +the first step to address racial politics is to understand the origin and consequences of racial animosity , even if it means uncovering unpleasant truths . +this is precisely what a large amount of research in economics , sociology , psychology and political science has done for the us . diff --git a/training/dtrain/examples/parallelized/work/weights.0 b/training/dtrain/examples/parallelized/work/weights.0 new file mode 100644 index 00000000..ddd595a8 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/weights.0 @@ -0,0 +1,12 @@ +LanguageModel 0.7004298992212881 +PhraseModel_2 0.5576194336478857 +PhraseModel_1 0.41787318415343155 +PhraseModel_4 -0.46728502545635164 +PhraseModel_3 -0.029839521598455515 +Glue -0.05760000000000068 +PhraseModel_6 -0.2716499999999978 +PhraseModel_0 -0.20831031065605327 +LanguageModel_OOV -0.15205000000000077 +PassThrough -0.1846500000000006 +WordPenalty 0.09636994553433414 +PhraseModel_5 -0.026900000000000257 diff --git a/training/dtrain/examples/parallelized/work/weights.0.0 b/training/dtrain/examples/parallelized/work/weights.0.0 new file mode 100644 index 00000000..c9370b18 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/weights.0.0 @@ -0,0 +1,12 @@ +WordPenalty -0.0079041595706392243 +LanguageModel 0.44781580828279532 +LanguageModel_OOV -0.04010000000000042 +Glue 0.26629999999999948 +PhraseModel_0 -0.19299677809125185 +PhraseModel_1 0.71321026861732773 +PhraseModel_2 0.85195540993310537 +PhraseModel_3 -0.43986310822842656 +PhraseModel_4 -0.44802855630415955 +PhraseModel_5 -0.053800000000000514 +PhraseModel_6 -0.17879999999999835 +PassThrough -0.14770000000000036 diff --git a/training/dtrain/examples/parallelized/work/weights.0.1 b/training/dtrain/examples/parallelized/work/weights.0.1 new file mode 100644 index 00000000..8fad3de8 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/weights.0.1 @@ -0,0 +1,12 @@ +WordPenalty 0.080605055841244472 +LanguageModel -0.026571720531022844 +LanguageModel_OOV -0.30024999999999141 +Glue -0.26989999999999842 +PhraseModel_2 0.92000295209089566 +PhraseModel_1 0.67450748692470841 +PhraseModel_4 -0.5920000014976784 +PhraseModel_3 -0.36402437203127397 +PhraseModel_6 -0.28754999999999603 +PhraseModel_0 -0.32076244202907672 +PassThrough -0.33284999999999004 +PhraseModel_5 -0.026900000000000257 diff --git a/training/dtrain/examples/parallelized/work/weights.1 b/training/dtrain/examples/parallelized/work/weights.1 new file mode 100644 index 00000000..03058a16 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/weights.1 @@ -0,0 +1,12 @@ +PhraseModel_2 0.8365578543552836 +PhraseModel_4 -0.5900840266009169 +PhraseModel_1 0.5312000609786991 +PhraseModel_0 -0.3872342271319619 +PhraseModel_3 -0.3728279676912084 +Glue -0.2938500000000036 +PhraseModel_6 -0.2803499999999967 +PassThrough -0.25014999999999626 +LanguageModel_OOV -0.21754999999999702 +LanguageModel 0.07306061161169894 +WordPenalty 0.09576193325966899 +PhraseModel_5 -0.026900000000000257 diff --git a/training/dtrain/examples/parallelized/work/weights.1.0 b/training/dtrain/examples/parallelized/work/weights.1.0 new file mode 100644 index 00000000..6a6a65c1 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/weights.1.0 @@ -0,0 +1,11 @@ +WordPenalty 0.20064405063930751 +LanguageModel 0.9530439901597807 +LanguageModel_OOV -0.26400000000000112 +Glue -0.38150000000000084 +PhraseModel_0 -0.22362384322085468 +PhraseModel_1 0.12253609968953538 +PhraseModel_2 0.26328345736266612 +PhraseModel_3 0.38018406503151553 +PhraseModel_4 -0.48654149460854373 +PhraseModel_6 -0.36449999999999722 +PassThrough -0.22160000000000085 diff --git a/training/dtrain/examples/parallelized/work/weights.1.1 b/training/dtrain/examples/parallelized/work/weights.1.1 new file mode 100644 index 00000000..f56ea4a2 --- /dev/null +++ b/training/dtrain/examples/parallelized/work/weights.1.1 @@ -0,0 +1,12 @@ +WordPenalty 0.1109188106780935 +LanguageModel 0.17269294375442074 +LanguageModel_OOV -0.13485000000000266 +Glue -0.3178000000000088 +PhraseModel_2 0.75311275661967159 +PhraseModel_1 0.38789263503268989 +PhraseModel_4 -0.58816805170415531 +PhraseModel_3 -0.38163156335114284 +PhraseModel_6 -0.27314999999999739 +PhraseModel_0 -0.45370601223484697 +PassThrough -0.16745000000000249 +PhraseModel_5 -0.026900000000000257 diff --git a/training/dtrain/examples/standard/README b/training/dtrain/examples/standard/README new file mode 100644 index 00000000..ce37d31a --- /dev/null +++ b/training/dtrain/examples/standard/README @@ -0,0 +1,2 @@ +Call `dtrain` from this folder with ../../dtrain -c dtrain.ini . + diff --git a/training/dtrain/test/example/cdec.ini b/training/dtrain/examples/standard/cdec.ini index d5955f0e..e1edc68d 100644 --- a/training/dtrain/test/example/cdec.ini +++ b/training/dtrain/examples/standard/cdec.ini @@ -2,9 +2,10 @@ formalism=scfg add_pass_through_rules=true scfg_max_span_limit=15 intersection_strategy=cube_pruning -cubepruning_pop_limit=30 +cubepruning_pop_limit=200 +grammar=nc-wmt11.grammar.gz feature_function=WordPenalty -feature_function=KLanguageModel test/example/nc-wmt11.en.srilm.gz +feature_function=KLanguageModel ./nc-wmt11.en.srilm.gz # all currently working feature functions for translation: # (with those features active that were used in the ACL paper) #feature_function=ArityPenalty diff --git a/training/dtrain/test/example/dtrain.ini b/training/dtrain/examples/standard/dtrain.ini index 72d50ca1..e1072d30 100644 --- a/training/dtrain/test/example/dtrain.ini +++ b/training/dtrain/examples/standard/dtrain.ini @@ -1,10 +1,12 @@ -input=test/example/nc-wmt11.1k.gz # use '-' for STDIN -output=- # a weights file (add .gz for gzip compression) or STDOUT '-' -select_weights=VOID # don't output weights -decoder_config=test/example/cdec.ini # config for cdec +input=./nc-wmt11.de.gz +refs=./nc-wmt11.en.gz +output=- # a weights file (add .gz for gzip compression) or STDOUT '-' +select_weights=VOID # output average (over epochs) weight vector +decoder_config=./cdec.ini # config for cdec # weights for these features will be printed on each iteration print_weights=Glue WordPenalty LanguageModel LanguageModel_OOV PhraseModel_0 PhraseModel_1 PhraseModel_2 PhraseModel_3 PhraseModel_4 PhraseModel_5 PhraseModel_6 PassThrough -tmp=/tmp +# newer version of the grammar extractor use different feature names: +#print_weights= EgivenFCoherent SampleCountF CountEF MaxLexFgivenE MaxLexEgivenF IsSingletonF IsSingletonFE Glue WordPenalty PassThrough LanguageModel LanguageModel_OOV stop_after=10 # stop epoch after 10 inputs # interesting stuff @@ -16,7 +18,7 @@ learning_rate=1.0 # learning rate, don't care if gamma=0 (perceptron) gamma=0 # use SVM reg sample_from=kbest # use kbest lists (as opposed to forest) filter=uniq # only unique entries in kbest (surface form) -pair_sampling=XYX +pair_sampling=XYX # hi_lo=0.1 # 10 vs 80 vs 10 and 80 vs 10 here -pair_threshold=0 # minimum distance in BLEU (this will still only use pairs with diff > 0) -loss_margin=0 +pair_threshold=0 # minimum distance in BLEU (here: > 0) +loss_margin=0 # update if correctly ranked, but within this margin diff --git a/training/dtrain/examples/standard/expected-output b/training/dtrain/examples/standard/expected-output new file mode 100644 index 00000000..7cd09dbf --- /dev/null +++ b/training/dtrain/examples/standard/expected-output @@ -0,0 +1,91 @@ + cdec cfg './cdec.ini' +Loading the LM will be faster if you build a binary file. +Reading ./nc-wmt11.en.srilm.gz +----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100 +**************************************************************************************************** + Example feature: Shape_S00000_T00000 +Seeding random number sequence to 2679584485 + +dtrain +Parameters: + k 100 + N 4 + T 2 + scorer 'stupid_bleu' + sample from 'kbest' + filter 'uniq' + learning rate 1 + gamma 0 + loss margin 0 + faster perceptron 1 + pairs 'XYX' + hi lo 0.1 + pair threshold 0 + select weights 'VOID' + l1 reg 0 'none' + max pairs 4294967295 + cdec cfg './cdec.ini' + input './nc-wmt11.de.gz' + refs './nc-wmt11.en.gz' + output '-' + stop_after 10 +(a dot represents 10 inputs) +Iteration #1 of 2. + . 10 +Stopping after 10 input sentences. +WEIGHTS + Glue = -576 + WordPenalty = +417.79 + LanguageModel = +5117.5 + LanguageModel_OOV = -1307 + PhraseModel_0 = -1612 + PhraseModel_1 = -2159.6 + PhraseModel_2 = -677.36 + PhraseModel_3 = +2663.8 + PhraseModel_4 = -1025.9 + PhraseModel_5 = -8 + PhraseModel_6 = +70 + PassThrough = -1455 + --- + 1best avg score: 0.27697 (+0.27697) + 1best avg model score: -47918 (-47918) + avg # pairs: 581.9 (meaningless) + avg # rank err: 581.9 + avg # margin viol: 0 + non0 feature count: 703 + avg list sz: 90.9 + avg f count: 100.09 +(time 0.25 min, 1.5 s/S) + +Iteration #2 of 2. + . 10 +WEIGHTS + Glue = -622 + WordPenalty = +898.56 + LanguageModel = +8066.2 + LanguageModel_OOV = -2590 + PhraseModel_0 = -4335.8 + PhraseModel_1 = -5864.4 + PhraseModel_2 = -1729.8 + PhraseModel_3 = +2831.9 + PhraseModel_4 = -5384.8 + PhraseModel_5 = +1449 + PhraseModel_6 = +480 + PassThrough = -2578 + --- + 1best avg score: 0.37119 (+0.094226) + 1best avg model score: -1.3174e+05 (-83822) + avg # pairs: 584.1 (meaningless) + avg # rank err: 584.1 + avg # margin viol: 0 + non0 feature count: 1115 + avg list sz: 91.3 + avg f count: 90.755 +(time 0.3 min, 1.8 s/S) + +Writing weights file to '-' ... +done + +--- +Best iteration: 2 [SCORE 'stupid_bleu'=0.37119]. +This took 0.55 min. diff --git a/training/dtrain/examples/standard/nc-wmt11.de.gz b/training/dtrain/examples/standard/nc-wmt11.de.gz Binary files differnew file mode 100644 index 00000000..0741fd92 --- /dev/null +++ b/training/dtrain/examples/standard/nc-wmt11.de.gz diff --git a/training/dtrain/examples/standard/nc-wmt11.en.gz b/training/dtrain/examples/standard/nc-wmt11.en.gz Binary files differnew file mode 100644 index 00000000..1c0bd401 --- /dev/null +++ b/training/dtrain/examples/standard/nc-wmt11.en.gz diff --git a/training/dtrain/examples/standard/nc-wmt11.en.srilm.gz b/training/dtrain/examples/standard/nc-wmt11.en.srilm.gz Binary files differnew file mode 100644 index 00000000..7ce81057 --- /dev/null +++ b/training/dtrain/examples/standard/nc-wmt11.en.srilm.gz diff --git a/training/dtrain/examples/standard/nc-wmt11.grammar.gz b/training/dtrain/examples/standard/nc-wmt11.grammar.gz Binary files differnew file mode 100644 index 00000000..ce4024a1 --- /dev/null +++ b/training/dtrain/examples/standard/nc-wmt11.grammar.gz diff --git a/training/dtrain/test/toy/cdec.ini b/training/dtrain/examples/toy/cdec.ini index 98b02d44..b14f4819 100644 --- a/training/dtrain/test/toy/cdec.ini +++ b/training/dtrain/examples/toy/cdec.ini @@ -1,2 +1,3 @@ formalism=scfg add_pass_through_rules=true +grammar=grammar.gz diff --git a/training/dtrain/test/toy/dtrain.ini b/training/dtrain/examples/toy/dtrain.ini index a091732f..cd715f26 100644 --- a/training/dtrain/test/toy/dtrain.ini +++ b/training/dtrain/examples/toy/dtrain.ini @@ -1,5 +1,6 @@ -decoder_config=test/toy/cdec.ini -input=test/toy/input +decoder_config=cdec.ini +input=src +refs=tgt output=- print_weights=logp shell_rule house_rule small_rule little_rule PassThrough k=4 diff --git a/training/dtrain/examples/toy/expected-output b/training/dtrain/examples/toy/expected-output new file mode 100644 index 00000000..1da2aadd --- /dev/null +++ b/training/dtrain/examples/toy/expected-output @@ -0,0 +1,77 @@ +Warning: hi_lo only works with pair_sampling XYX. + cdec cfg 'cdec.ini' +Seeding random number sequence to 1664825829 + +dtrain +Parameters: + k 4 + N 4 + T 2 + scorer 'bleu' + sample from 'kbest' + filter 'uniq' + learning rate 1 + gamma 0 + loss margin 0 + pairs 'all' + pair threshold 0 + select weights 'last' + l1 reg 0 'none' + max pairs 4294967295 + cdec cfg 'cdec.ini' + input 'src' + refs 'tgt' + output '-' +(a dot represents 10 inputs) +Iteration #1 of 2. + 2 +WEIGHTS + logp = +0 + shell_rule = -1 + house_rule = +2 + small_rule = -2 + little_rule = +3 + PassThrough = -5 + --- + 1best avg score: 0.5 (+0.5) + 1best avg model score: 2.5 (+2.5) + avg # pairs: 4 + avg # rank err: 1.5 + avg # margin viol: 0 + non0 feature count: 6 + avg list sz: 4 + avg f count: 2.875 +(time 0 min, 0 s/S) + +Iteration #2 of 2. + 2 +WEIGHTS + logp = +0 + shell_rule = -1 + house_rule = +2 + small_rule = -2 + little_rule = +3 + PassThrough = -5 + --- + 1best avg score: 1 (+0.5) + 1best avg model score: 5 (+2.5) + avg # pairs: 5 + avg # rank err: 0 + avg # margin viol: 0 + non0 feature count: 6 + avg list sz: 4 + avg f count: 3 +(time 0 min, 0 s/S) + +Writing weights file to '-' ... +house_rule 2 +little_rule 3 +Glue -4 +PassThrough -5 +small_rule -2 +shell_rule -1 +done + +--- +Best iteration: 2 [SCORE 'bleu'=1]. +This took 0 min. diff --git a/training/dtrain/examples/toy/grammar.gz b/training/dtrain/examples/toy/grammar.gz Binary files differnew file mode 100644 index 00000000..8eb0d29e --- /dev/null +++ b/training/dtrain/examples/toy/grammar.gz diff --git a/training/dtrain/examples/toy/src b/training/dtrain/examples/toy/src new file mode 100644 index 00000000..87e39ef2 --- /dev/null +++ b/training/dtrain/examples/toy/src @@ -0,0 +1,2 @@ +ich sah ein kleines haus +ich fand ein kleines haus diff --git a/training/dtrain/examples/toy/tgt b/training/dtrain/examples/toy/tgt new file mode 100644 index 00000000..174926b3 --- /dev/null +++ b/training/dtrain/examples/toy/tgt @@ -0,0 +1,2 @@ +i saw a little house +i found a little house diff --git a/training/dtrain/hstreaming/avg.rb b/training/dtrain/hstreaming/avg.rb deleted file mode 100755 index 2599c732..00000000 --- a/training/dtrain/hstreaming/avg.rb +++ /dev/null @@ -1,32 +0,0 @@ -#!/usr/bin/env ruby -# first arg may be an int of custom shard count - -shard_count_key = "__SHARD_COUNT__" - -STDIN.set_encoding 'utf-8' -STDOUT.set_encoding 'utf-8' - -w = {} -c = {} -w.default = 0 -c.default = 0 -while line = STDIN.gets - key, val = line.split /\s/ - w[key] += val.to_f - c[key] += 1 -end - -if ARGV.size == 0 - shard_count = w["__SHARD_COUNT__"] -else - shard_count = ARGV[0].to_f -end -w.each_key { |k| - if k == shard_count_key - next - else - puts "#{k}\t#{w[k]/shard_count}" - #puts "# #{c[k]}" - end -} - diff --git a/training/dtrain/hstreaming/cdec.ini b/training/dtrain/hstreaming/cdec.ini deleted file mode 100644 index d4f5cecd..00000000 --- a/training/dtrain/hstreaming/cdec.ini +++ /dev/null @@ -1,22 +0,0 @@ -formalism=scfg -add_pass_through_rules=true -scfg_max_span_limit=15 -intersection_strategy=cube_pruning -cubepruning_pop_limit=30 -feature_function=WordPenalty -feature_function=KLanguageModel nc-wmt11.en.srilm.gz -#feature_function=ArityPenalty -#feature_function=CMR2008ReorderingFeatures -#feature_function=Dwarf -#feature_function=InputIndicator -#feature_function=LexNullJump -#feature_function=NewJump -#feature_function=NgramFeatures -#feature_function=NonLatinCount -#feature_function=OutputIndicator -#feature_function=RuleIdentityFeatures -#feature_function=RuleNgramFeatures -#feature_function=RuleShape -#feature_function=SourceSpanSizeFeatures -#feature_function=SourceWordPenalty -#feature_function=SpanFeatures diff --git a/training/dtrain/hstreaming/dtrain.ini b/training/dtrain/hstreaming/dtrain.ini deleted file mode 100644 index a2c219a1..00000000 --- a/training/dtrain/hstreaming/dtrain.ini +++ /dev/null @@ -1,15 +0,0 @@ -input=- -output=- -decoder_config=cdec.ini -tmp=/var/hadoop/mapred/local/ -epochs=1 -k=100 -N=4 -learning_rate=0.0001 -gamma=0 -scorer=stupid_bleu -sample_from=kbest -filter=uniq -pair_sampling=XYX -pair_threshold=0 -select_weights=last diff --git a/training/dtrain/hstreaming/dtrain.sh b/training/dtrain/hstreaming/dtrain.sh deleted file mode 100755 index 877ff94c..00000000 --- a/training/dtrain/hstreaming/dtrain.sh +++ /dev/null @@ -1,9 +0,0 @@ -#!/bin/bash -# script to run dtrain with a task id - -pushd . &>/dev/null -cd .. -ID=$(basename $(pwd)) # attempt_... -popd &>/dev/null -./dtrain -c dtrain.ini --hstreaming $ID - diff --git a/training/dtrain/hstreaming/hadoop-streaming-job.sh b/training/dtrain/hstreaming/hadoop-streaming-job.sh deleted file mode 100755 index 92419956..00000000 --- a/training/dtrain/hstreaming/hadoop-streaming-job.sh +++ /dev/null @@ -1,30 +0,0 @@ -#!/bin/sh - -EXP=a_simple_test - -# change these vars to fit your hadoop installation -HADOOP_HOME=/usr/lib/hadoop-0.20 -JAR=contrib/streaming/hadoop-streaming-0.20.2-cdh3u1.jar -HSTREAMING="$HADOOP_HOME/bin/hadoop jar $HADOOP_HOME/$JAR" - - IN=input_on_hdfs -OUT=output_weights_on_hdfs - -# you can -reducer to NONE if you want to -# do feature selection/averaging locally (e.g. to -# keep weights of all epochs) -$HSTREAMING \ - -mapper "dtrain.sh" \ - -reducer "ruby lplp.rb l2 select_k 100000" \ - -input $IN \ - -output $OUT \ - -file dtrain.sh \ - -file lplp.rb \ - -file ../dtrain \ - -file dtrain.ini \ - -file cdec.ini \ - -file ../test/example/nc-wmt11.en.srilm.gz \ - -jobconf mapred.reduce.tasks=30 \ - -jobconf mapred.max.map.failures.percent=0 \ - -jobconf mapred.job.name="dtrain $EXP" - diff --git a/training/dtrain/hstreaming/red-test b/training/dtrain/hstreaming/red-test deleted file mode 100644 index 2623d697..00000000 --- a/training/dtrain/hstreaming/red-test +++ /dev/null @@ -1,9 +0,0 @@ -a 1 -b 2 -c 3.5 -a 1 -b 2 -c 3.5 -d 1 -e 2 -__SHARD_COUNT__ 2 diff --git a/training/dtrain/hstreaming/lplp.rb b/training/dtrain/lplp.rb index f0cd58c5..86e835e8 100755 --- a/training/dtrain/hstreaming/lplp.rb +++ b/training/dtrain/lplp.rb @@ -84,34 +84,28 @@ def _test() end #_test() -# actually do something + def usage() - puts "lplp.rb <l0,l1,l2,linfty,mean,median> <cut|select_k> <k|threshold> [n] < <input>" + puts "lplp.rb <l0,l1,l2,linfty,mean,median> <cut|select_k> <k|threshold> <#shards> < <input>" puts " l0...: norms for selection" puts "select_k: only output top k (according to the norm of their column vector) features" puts " cut: output features with weight >= threshold" puts " n: if we do not have a shard count use this number for averaging" - exit + exit 1 end -if ARGV.size < 3 then usage end +if ARGV.size < 4 then usage end norm_fun = method(ARGV[0].to_sym) type = ARGV[1] x = ARGV[2].to_f - -shard_count_key = "__SHARD_COUNT__" +shard_count = ARGV[3].to_f STDIN.set_encoding 'utf-8' STDOUT.set_encoding 'utf-8' w = {} -shard_count = 0 while line = STDIN.gets key, val = line.split /\s+/ - if key == shard_count_key - shard_count += 1 - next - end if w.has_key? key w[key].push val.to_f else @@ -119,8 +113,6 @@ while line = STDIN.gets end end -if ARGV.size == 4 then shard_count = ARGV[3].to_f end - if type == 'cut' cut(w, norm_fun, shard_count, x) elsif type == 'select_k' diff --git a/training/dtrain/pairsampling.h b/training/dtrain/pairsampling.h index 84be1efb..3f67e209 100644 --- a/training/dtrain/pairsampling.h +++ b/training/dtrain/pairsampling.h @@ -19,7 +19,7 @@ cmp_hyp_by_score_d(ScoredHyp a, ScoredHyp b) } inline void -all_pairs(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, score_t threshold, unsigned max, float _unused=1) +all_pairs(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, score_t threshold, unsigned max, bool misranked_only, float _unused=1) { sort(s->begin(), s->end(), cmp_hyp_by_score_d); unsigned sz = s->size(); @@ -27,6 +27,7 @@ all_pairs(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, sc unsigned count = 0; for (unsigned i = 0; i < sz-1; i++) { for (unsigned j = i+1; j < sz; j++) { + if (misranked_only && !((*s)[i].model <= (*s)[j].model)) continue; if (threshold > 0) { if (accept_pair((*s)[i].score, (*s)[j].score, threshold)) training.push_back(make_pair((*s)[i], (*s)[j])); @@ -51,7 +52,7 @@ all_pairs(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, sc */ inline void -partXYX(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, score_t threshold, unsigned max, float hi_lo) +partXYX(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, score_t threshold, unsigned max, bool misranked_only, float hi_lo) { unsigned sz = s->size(); if (sz < 2) return; @@ -64,9 +65,7 @@ partXYX(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, scor unsigned count = 0; for (unsigned i = 0; i < sep_hi; i++) { for (unsigned j = sep_hi; j < sz; j++) { -#ifdef DTRAIN_FASTER_PERCEPTRON - if ((*s)[i].model <= (*s)[j].model) { -#endif + if (misranked_only && !((*s)[i].model <= (*s)[j].model)) continue; if (threshold > 0) { if (accept_pair((*s)[i].score, (*s)[j].score, threshold)) training.push_back(make_pair((*s)[i], (*s)[j])); @@ -78,9 +77,6 @@ partXYX(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, scor b = true; break; } -#ifdef DTRAIN_FASTER_PERCEPTRON - } -#endif } if (b) break; } @@ -88,9 +84,7 @@ partXYX(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, scor while (sep_lo > 0 && (*s)[sep_lo-1].score == (*s)[sep_lo].score) --sep_lo; for (unsigned i = sep_hi; i < sz-sep_lo; i++) { for (unsigned j = sz-sep_lo; j < sz; j++) { -#ifdef DTRAIN_FASTER_PERCEPTRON - if ((*s)[i].model <= (*s)[j].model) { -#endif + if (misranked_only && !((*s)[i].model <= (*s)[j].model)) continue; if (threshold > 0) { if (accept_pair((*s)[i].score, (*s)[j].score, threshold)) training.push_back(make_pair((*s)[i], (*s)[j])); @@ -99,9 +93,6 @@ partXYX(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, scor training.push_back(make_pair((*s)[i], (*s)[j])); } if (++count == max) return; -#ifdef DTRAIN_FASTER_PERCEPTRON - } -#endif } } } @@ -119,7 +110,7 @@ _PRO_cmp_pair_by_diff_d(pair<ScoredHyp,ScoredHyp> a, pair<ScoredHyp,ScoredHyp> b return (fabs(a.first.score - a.second.score)) > (fabs(b.first.score - b.second.score)); } inline void -PROsampling(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, score_t threshold, unsigned max, float _unused=1) +PROsampling(vector<ScoredHyp>* s, vector<pair<ScoredHyp,ScoredHyp> >& training, score_t threshold, unsigned max, bool _unused=false, float _also_unused=0) { unsigned max_count = 5000, count = 0, sz = s->size(); bool b = false; diff --git a/training/dtrain/parallelize.rb b/training/dtrain/parallelize.rb index eb4148f5..e661416e 100755 --- a/training/dtrain/parallelize.rb +++ b/training/dtrain/parallelize.rb @@ -1,80 +1,149 @@ #!/usr/bin/env ruby +require 'trollop' -if ARGV.size != 5 +def usage STDERR.write "Usage: " - STDERR.write "ruby parallelize.rb <#shards> <input> <refs> <epochs> <dtrain.ini>\n" - exit + STDERR.write "ruby parallelize.rb -c <dtrain.ini> [-e <epochs=10>] [--randomize/-z] [--reshard/-y] -s <#shards|0> [-p <at once=9999>] -i <input> -r <refs> [--qsub/-q] [--dtrain_binary <path to dtrain binary>] [-l \"l2 select_k 100000\"]\n" + exit 1 end -cdec_dir = '/path/to/cdec_dir' -dtrain_bin = "#{cdec_dir}/training/dtrain/dtrain_local" -ruby = '/usr/bin/ruby' -lplp_rb = "#{cdec_dir}/training/dtrain/hstreaming/lplp.rb" -lplp_args = 'l2 select_k 100000' -gzip = '/bin/gzip' +opts = Trollop::options do + opt :config, "dtrain config file", :type => :string + opt :epochs, "number of epochs", :type => :int, :default => 10 + opt :lplp_args, "arguments for lplp.rb", :type => :string, :default => "l2 select_k 100000" + opt :randomize, "randomize shards before each epoch", :type => :bool, :short => '-z', :default => false + opt :reshard, "reshard after each epoch", :type => :bool, :short => '-y', :default => false + opt :shards, "number of shards", :type => :int + opt :processes_at_once, "have this number (max) running at the same time", :type => :int, :default => 9999 + opt :input, "input", :type => :string + opt :references, "references", :type => :string + opt :qsub, "use qsub", :type => :bool, :default => false + opt :dtrain_binary, "path to dtrain binary", :type => :string +end +usage if not opts[:config]&&opts[:shards]&&opts[:input]&&opts[:references] + -num_shards = ARGV[0].to_i -input = ARGV[1] -refs = ARGV[2] -epochs = ARGV[3].to_i -ini = ARGV[4] +dtrain_dir = File.expand_path File.dirname(__FILE__) +if not opts[:dtrain_binary] + dtrain_bin = "#{dtrain_dir}/dtrain" +else + dtrain_bin = opts[:dtrain_binary] +end +ruby = '/usr/bin/ruby' +lplp_rb = "#{dtrain_dir}/lplp.rb" +lplp_args = opts[:lplp_args] +cat = '/bin/cat' +ini = opts[:config] +epochs = opts[:epochs] +rand = opts[:randomize] +reshard = opts[:reshard] +predefined_shards = false +if opts[:shards] == 0 + predefined_shards = true + num_shards = 0 +else + num_shards = opts[:shards] +end +input = opts[:input] +refs = opts[:references] +use_qsub = opts[:qsub] +shards_at_once = opts[:processes_at_once] `mkdir work` -def make_shards(input, refs, num_shards) +def make_shards(input, refs, num_shards, epoch, rand) lc = `wc -l #{input}`.split.first.to_i + index = (0..lc-1).to_a + index.reverse! + index.shuffle! if rand shard_sz = lc / num_shards leftover = lc % num_shards in_f = File.new input, 'r' + in_lines = in_f.readlines refs_f = File.new refs, 'r' + refs_lines = refs_f.readlines shard_in_files = [] shard_refs_files = [] + in_fns = [] + refs_fns = [] 0.upto(num_shards-1) { |shard| - shard_in = File.new "work/shard.#{shard}.in", 'w+' - shard_refs = File.new "work/shard.#{shard}.refs", 'w+' + in_fn = "work/shard.#{shard}.#{epoch}.in" + shard_in = File.new in_fn, 'w+' + in_fns << in_fn + refs_fn = "work/shard.#{shard}.#{epoch}.refs" + shard_refs = File.new refs_fn, 'w+' + refs_fns << refs_fn 0.upto(shard_sz-1) { |i| - shard_in.write in_f.gets - shard_refs.write refs_f.gets + j = index.pop + shard_in.write in_lines[j] + shard_refs.write refs_lines[j] } shard_in_files << shard_in shard_refs_files << shard_refs } while leftover > 0 - shard_in_files[-1].write in_f.gets - shard_refs_files[-1].write refs_f.gets + j = index.pop + shard_in_files[-1].write in_lines[j] + shard_refs_files[-1].write refs_lines[j] leftover -= 1 end (shard_in_files + shard_refs_files).each do |f| f.close end in_f.close refs_f.close + return [in_fns, refs_fns] end -make_shards input, refs, num_shards +input_files = [] +refs_files = [] +if predefined_shards + input_files = File.new(input).readlines.map {|i| i.strip } + refs_files = File.new(refs).readlines.map {|i| i.strip } + num_shards = input_files.size +else + input_files, refs_files = make_shards input, refs, num_shards, 0, rand +end 0.upto(epochs-1) { |epoch| + puts "epoch #{epoch+1}" pids = [] input_weights = '' if epoch > 0 then input_weights = "--input_weights work/weights.#{epoch-1}" end weights_files = [] - 0.upto(num_shards-1) { |shard| - pids << Kernel.fork { - `#{dtrain_bin} -c #{ini}\ - --input work/shard.#{shard}.in\ - --refs work/shard.#{shard}.refs #{input_weights}\ - --output work/weights.#{shard}.#{epoch}\ - &> work/out.#{shard}.#{epoch}` + shard = 0 + remaining_shards = num_shards + while remaining_shards > 0 + shards_at_once.times { + break if remaining_shards==0 + qsub_str_start = qsub_str_end = '' + local_end = '' + if use_qsub + qsub_str_start = "qsub -cwd -sync y -b y -j y -o work/out.#{shard}.#{epoch} -N dtrain.#{shard}.#{epoch} \"" + qsub_str_end = "\"" + local_end = '' + else + local_end = "&>work/out.#{shard}.#{epoch}" + end + pids << Kernel.fork { + `#{qsub_str_start}#{dtrain_bin} -c #{ini}\ + --input #{input_files[shard]}\ + --refs #{refs_files[shard]} #{input_weights}\ + --output work/weights.#{shard}.#{epoch}#{qsub_str_end} #{local_end}` + } + weights_files << "work/weights.#{shard}.#{epoch}" + shard += 1 + remaining_shards -= 1 } - weights_files << "work/weights.#{shard}.#{epoch}" - } - pids.each { |pid| Process.wait(pid) } - cat = File.new('work/weights_cat', 'w+') - weights_files.each { |f| cat.write File.new(f, 'r').read } - cat.close - `#{ruby} #{lplp_rb} #{lplp_args} #{num_shards} < work/weights_cat &> work/weights.#{epoch}` + pids.each { |pid| Process.wait(pid) } + pids.clear + end + `#{cat} work/weights.*.#{epoch} > work/weights_cat` + `#{ruby} #{lplp_rb} #{lplp_args} #{num_shards} < work/weights_cat > work/weights.#{epoch}` + if rand and reshard and epoch+1!=epochs + input_files, refs_files = make_shards input, refs, num_shards, epoch+1, rand + end } `rm work/weights_cat` -`#{gzip} work/*` diff --git a/training/dtrain/score.cc b/training/dtrain/score.cc index 34fc86a9..96d6e10a 100644 --- a/training/dtrain/score.cc +++ b/training/dtrain/score.cc @@ -49,7 +49,7 @@ BleuScorer::Score(vector<WordID>& hyp, vector<WordID>& ref, * for Machine Translation" * (Lin & Och '04) * - * NOTE: 0 iff no 1gram match + * NOTE: 0 iff no 1gram match ('grounded') */ score_t StupidBleuScorer::Score(vector<WordID>& hyp, vector<WordID>& ref, @@ -74,6 +74,35 @@ StupidBleuScorer::Score(vector<WordID>& hyp, vector<WordID>& ref, } /* + * fixed 'stupid' bleu + * + * as in "Optimizing for Sentence-Level BLEU+1 + * Yields Short Translations" + * (Nakov et al. '12) + */ +score_t +FixedStupidBleuScorer::Score(vector<WordID>& hyp, vector<WordID>& ref, + const unsigned /*rank*/, const unsigned /*src_len*/) +{ + unsigned hyp_len = hyp.size(), ref_len = ref.size(); + if (hyp_len == 0 || ref_len == 0) return 0.; + NgramCounts counts = make_ngram_counts(hyp, ref, N_); + unsigned M = N_; + vector<score_t> v = w_; + if (ref_len < N_) { + M = ref_len; + for (unsigned i = 0; i < M; i++) v[i] = 1/((score_t)M); + } + score_t sum = 0, add = 0; + for (unsigned i = 0; i < M; i++) { + if (i == 0 && (counts.sum_[i] == 0 || counts.clipped_[i] == 0)) return 0.; + if (i == 1) add = 1; + sum += v[i] * log(((score_t)counts.clipped_[i] + add)/((counts.sum_[i] + add))); + } + return brevity_penalty(hyp_len, ref_len+1) * exp(sum); // <- fix +} + +/* * smooth bleu * * as in "An End-to-End Discriminative Approach diff --git a/training/dtrain/score.h b/training/dtrain/score.h index f317c903..bddaa071 100644 --- a/training/dtrain/score.h +++ b/training/dtrain/score.h @@ -148,6 +148,11 @@ struct StupidBleuScorer : public LocalScorer score_t Score(vector<WordID>& hyp, vector<WordID>& ref, const unsigned /*rank*/, const unsigned /*src_len*/); }; +struct FixedStupidBleuScorer : public LocalScorer +{ + score_t Score(vector<WordID>& hyp, vector<WordID>& ref, const unsigned /*rank*/, const unsigned /*src_len*/); +}; + struct SmoothBleuScorer : public LocalScorer { score_t Score(vector<WordID>& hyp, vector<WordID>& ref, const unsigned /*rank*/, const unsigned /*src_len*/); diff --git a/training/dtrain/test/example/README b/training/dtrain/test/example/README deleted file mode 100644 index 6937b11b..00000000 --- a/training/dtrain/test/example/README +++ /dev/null @@ -1,8 +0,0 @@ -Small example of input format for distributed training. -Call dtrain from cdec/dtrain/ with ./dtrain -c test/example/dtrain.ini . - -For this to work, undef 'DTRAIN_LOCAL' in dtrain.h -and recompile. - -Data is here: http://simianer.de/#dtrain - diff --git a/training/dtrain/test/example/expected-output b/training/dtrain/test/example/expected-output deleted file mode 100644 index 05326763..00000000 --- a/training/dtrain/test/example/expected-output +++ /dev/null @@ -1,89 +0,0 @@ - cdec cfg 'test/example/cdec.ini' -Loading the LM will be faster if you build a binary file. -Reading test/example/nc-wmt11.en.srilm.gz -----5---10---15---20---25---30---35---40---45---50---55---60---65---70---75---80---85---90---95--100 -**************************************************************************************************** - Example feature: Shape_S00000_T00000 -Seeding random number sequence to 2912000813 - -dtrain -Parameters: - k 100 - N 4 - T 2 - scorer 'stupid_bleu' - sample from 'kbest' - filter 'uniq' - learning rate 1 - gamma 0 - loss margin 0 - pairs 'XYX' - hi lo 0.1 - pair threshold 0 - select weights 'VOID' - l1 reg 0 'none' - max pairs 4294967295 - cdec cfg 'test/example/cdec.ini' - input 'test/example/nc-wmt11.1k.gz' - output '-' - stop_after 10 -(a dot represents 10 inputs) -Iteration #1 of 2. - . 10 -Stopping after 10 input sentences. -WEIGHTS - Glue = -637 - WordPenalty = +1064 - LanguageModel = +1175.3 - LanguageModel_OOV = -1437 - PhraseModel_0 = +1935.6 - PhraseModel_1 = +2499.3 - PhraseModel_2 = +964.96 - PhraseModel_3 = +1410.8 - PhraseModel_4 = -5977.9 - PhraseModel_5 = +522 - PhraseModel_6 = +1089 - PassThrough = -1308 - --- - 1best avg score: 0.16963 (+0.16963) - 1best avg model score: 64485 (+64485) - avg # pairs: 1494.4 - avg # rank err: 702.6 - avg # margin viol: 0 - non0 feature count: 528 - avg list sz: 85.7 - avg f count: 102.75 -(time 0.083 min, 0.5 s/S) - -Iteration #2 of 2. - . 10 -WEIGHTS - Glue = -1196 - WordPenalty = +809.52 - LanguageModel = +3112.1 - LanguageModel_OOV = -1464 - PhraseModel_0 = +3895.5 - PhraseModel_1 = +4683.4 - PhraseModel_2 = +1092.8 - PhraseModel_3 = +1079.6 - PhraseModel_4 = -6827.7 - PhraseModel_5 = -888 - PhraseModel_6 = +142 - PassThrough = -1335 - --- - 1best avg score: 0.277 (+0.10736) - 1best avg model score: -3110.5 (-67595) - avg # pairs: 1144.2 - avg # rank err: 529.1 - avg # margin viol: 0 - non0 feature count: 859 - avg list sz: 74.9 - avg f count: 112.84 -(time 0.067 min, 0.4 s/S) - -Writing weights file to '-' ... -done - ---- -Best iteration: 2 [SCORE 'stupid_bleu'=0.277]. -This took 0.15 min. diff --git a/training/dtrain/test/parallelize/in b/training/dtrain/test/parallelize/in deleted file mode 100644 index a312809f..00000000 --- a/training/dtrain/test/parallelize/in +++ /dev/null @@ -1,10 +0,0 @@ -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.0.gz" id="0">barack obama erhält als vierter us @-@ präsident den frieden nobelpreis</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.1.gz" id="1">der amerikanische präsident barack obama kommt für 26 stunden nach oslo , norwegen , um hier als vierter us @-@ präsident in der geschichte den frieden nobelpreis entgegen zunehmen .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.2.gz" id="2">darüber hinaus erhält er das diplom sowie die medaille und einen scheck über 1,4 mio. dollar für seine außer gewöhnlichen bestrebungen um die intensivierung der welt diplomatie und zusammen arbeit unter den völkern .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.3.gz" id="3">der chef des weißen hauses kommt morgen zusammen mit seiner frau michelle in der nordwegischen metropole an und wird die ganze zeit beschäftigt sein .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.4.gz" id="4">zunächst stattet er dem nobel @-@ institut einen besuch ab , wo er überhaupt zum ersten mal mit den fünf ausschuss mitglieder zusammen trifft , die ihn im oktober aus 172 leuten und 33 organisationen gewählt haben .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.5.gz" id="5">das präsidenten paar hat danach ein treffen mit dem norwegischen könig harald v. und königin sonja eingeplant .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.6.gz" id="6">nachmittags erreicht dann der besuch seinen höhepunkt mit der zeremonie , bei der obama den prestige preis übernimmt .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.7.gz" id="7">diesen erhält er als der vierte us @-@ präsident , aber erst als der dritte , der den preis direkt im amt entgegen nimmt .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.8.gz" id="8">das weiße haus avisierte schon , dass obama bei der übernahme des preises über den afghanistan krieg sprechen wird .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.9.gz" id="9">der präsident will diesem thema nicht ausweichen , weil er weiß , dass er den preis als ein präsident übernimmt , der zur zeit krieg in zwei ländern führt .</seg> diff --git a/training/dtrain/test/parallelize/refs b/training/dtrain/test/parallelize/refs deleted file mode 100644 index 4d3128cb..00000000 --- a/training/dtrain/test/parallelize/refs +++ /dev/null @@ -1,10 +0,0 @@ -barack obama becomes the fourth american president to receive the nobel peace prize -the american president barack obama will fly into oslo , norway for 26 hours to receive the nobel peace prize , the fourth american president in history to do so . -he will receive a diploma , medal and cheque for 1.4 million dollars for his exceptional efforts to improve global diplomacy and encourage international cooperation , amongst other things . -the head of the white house will be flying into the norwegian city in the morning with his wife michelle and will have a busy schedule . -first , he will visit the nobel institute , where he will have his first meeting with the five committee members who selected him from 172 people and 33 organisations . -the presidential couple then has a meeting scheduled with king harald v and queen sonja of norway . -then , in the afternoon , the visit will culminate in a grand ceremony , at which obama will receive the prestigious award . -he will be the fourth american president to be awarded the prize , and only the third to have received it while actually in office . -the white house has stated that , when he accepts the prize , obama will speak about the war in afghanistan . -the president does not want to skirt around this topic , as he realises that he is accepting the prize as a president whose country is currently at war in two countries . diff --git a/training/dtrain/test/parallelize/test/cdec.ini b/training/dtrain/test/parallelize/test/cdec.ini deleted file mode 100644 index 72e99dc5..00000000 --- a/training/dtrain/test/parallelize/test/cdec.ini +++ /dev/null @@ -1,22 +0,0 @@ -formalism=scfg -add_pass_through_rules=true -intersection_strategy=cube_pruning -cubepruning_pop_limit=200 -scfg_max_span_limit=15 -feature_function=WordPenalty -feature_function=KLanguageModel /stor/dat/wmt12/en/news_only/m/wmt12.news.en.3.kenv5 -#feature_function=ArityPenalty -#feature_function=CMR2008ReorderingFeatures -#feature_function=Dwarf -#feature_function=InputIndicator -#feature_function=LexNullJump -#feature_function=NewJump -#feature_function=NgramFeatures -#feature_function=NonLatinCount -#feature_function=OutputIndicator -#feature_function=RuleIdentityFeatures -#feature_function=RuleNgramFeatures -#feature_function=RuleShape -#feature_function=SourceSpanSizeFeatures -#feature_function=SourceWordPenalty -#feature_function=SpanFeatures diff --git a/training/dtrain/test/parallelize/test/dtrain.ini b/training/dtrain/test/parallelize/test/dtrain.ini deleted file mode 100644 index 03f9d240..00000000 --- a/training/dtrain/test/parallelize/test/dtrain.ini +++ /dev/null @@ -1,15 +0,0 @@ -k=100 -N=4 -learning_rate=0.0001 -gamma=0 -loss_margin=0 -epochs=1 -scorer=stupid_bleu -sample_from=kbest -filter=uniq -pair_sampling=XYX -hi_lo=0.1 -select_weights=last -print_weights=Glue WordPenalty LanguageModel LanguageModel_OOV PhraseModel_0 PhraseModel_1 PhraseModel_2 PhraseModel_3 PhraseModel_4 PhraseModel_5 PhraseModel_6 PassThrough -tmp=/tmp -decoder_config=cdec.ini diff --git a/training/dtrain/test/parallelize/test/in b/training/dtrain/test/parallelize/test/in deleted file mode 100644 index a312809f..00000000 --- a/training/dtrain/test/parallelize/test/in +++ /dev/null @@ -1,10 +0,0 @@ -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.0.gz" id="0">barack obama erhält als vierter us @-@ präsident den frieden nobelpreis</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.1.gz" id="1">der amerikanische präsident barack obama kommt für 26 stunden nach oslo , norwegen , um hier als vierter us @-@ präsident in der geschichte den frieden nobelpreis entgegen zunehmen .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.2.gz" id="2">darüber hinaus erhält er das diplom sowie die medaille und einen scheck über 1,4 mio. dollar für seine außer gewöhnlichen bestrebungen um die intensivierung der welt diplomatie und zusammen arbeit unter den völkern .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.3.gz" id="3">der chef des weißen hauses kommt morgen zusammen mit seiner frau michelle in der nordwegischen metropole an und wird die ganze zeit beschäftigt sein .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.4.gz" id="4">zunächst stattet er dem nobel @-@ institut einen besuch ab , wo er überhaupt zum ersten mal mit den fünf ausschuss mitglieder zusammen trifft , die ihn im oktober aus 172 leuten und 33 organisationen gewählt haben .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.5.gz" id="5">das präsidenten paar hat danach ein treffen mit dem norwegischen könig harald v. und königin sonja eingeplant .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.6.gz" id="6">nachmittags erreicht dann der besuch seinen höhepunkt mit der zeremonie , bei der obama den prestige preis übernimmt .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.7.gz" id="7">diesen erhält er als der vierte us @-@ präsident , aber erst als der dritte , der den preis direkt im amt entgegen nimmt .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.8.gz" id="8">das weiße haus avisierte schon , dass obama bei der übernahme des preises über den afghanistan krieg sprechen wird .</seg> -<seg grammar="/stor/dat/wmt12/dev/newstest2010/g/grammar.out.9.gz" id="9">der präsident will diesem thema nicht ausweichen , weil er weiß , dass er den preis als ein präsident übernimmt , der zur zeit krieg in zwei ländern führt .</seg> diff --git a/training/dtrain/test/parallelize/test/refs b/training/dtrain/test/parallelize/test/refs deleted file mode 100644 index 4d3128cb..00000000 --- a/training/dtrain/test/parallelize/test/refs +++ /dev/null @@ -1,10 +0,0 @@ -barack obama becomes the fourth american president to receive the nobel peace prize -the american president barack obama will fly into oslo , norway for 26 hours to receive the nobel peace prize , the fourth american president in history to do so . -he will receive a diploma , medal and cheque for 1.4 million dollars for his exceptional efforts to improve global diplomacy and encourage international cooperation , amongst other things . -the head of the white house will be flying into the norwegian city in the morning with his wife michelle and will have a busy schedule . -first , he will visit the nobel institute , where he will have his first meeting with the five committee members who selected him from 172 people and 33 organisations . -the presidential couple then has a meeting scheduled with king harald v and queen sonja of norway . -then , in the afternoon , the visit will culminate in a grand ceremony , at which obama will receive the prestigious award . -he will be the fourth american president to be awarded the prize , and only the third to have received it while actually in office . -the white house has stated that , when he accepts the prize , obama will speak about the war in afghanistan . -the president does not want to skirt around this topic , as he realises that he is accepting the prize as a president whose country is currently at war in two countries . diff --git a/training/dtrain/test/toy/input b/training/dtrain/test/toy/input deleted file mode 100644 index 4d10a9ea..00000000 --- a/training/dtrain/test/toy/input +++ /dev/null @@ -1,2 +0,0 @@ -0 ich sah ein kleines haus i saw a little house [S] ||| [NP,1] [VP,2] ||| [1] [2] ||| logp=0 [NP] ||| ich ||| i ||| logp=0 [NP] ||| ein [NN,1] ||| a [1] ||| logp=0 [NN] ||| [JJ,1] haus ||| [1] house ||| logp=0 house_rule=1 [NN] ||| [JJ,1] haus ||| [1] shell ||| logp=0 shell_rule=1 [JJ] ||| kleines ||| small ||| logp=0 small_rule=1 [JJ] ||| kleines ||| little ||| logp=0 little_rule=1 [JJ] ||| grosses ||| big ||| logp=0 [JJ] ||| grosses ||| large ||| logp=0 [VP] ||| [V,1] [NP,2] ||| [1] [2] ||| logp=0 [V] ||| sah ||| saw ||| logp=0 [V] ||| fand ||| found ||| logp=0 -1 ich fand ein kleines haus i found a little house [S] ||| [NP,1] [VP,2] ||| [1] [2] ||| logp=0 [NP] ||| ich ||| i ||| logp=0 [NP] ||| ein [NN,1] ||| a [1] ||| logp=0 [NN] ||| [JJ,1] haus ||| [1] house ||| logp=0 house_rule=1 [NN] ||| [JJ,1] haus ||| [1] shell ||| logp=0 shell_rule=1 [JJ] ||| kleines ||| small ||| logp=0 small_rule=1 [JJ] ||| kleines ||| little ||| logp=0 little_rule=1 [JJ] ||| grosses ||| big ||| logp=0 [JJ] ||| grosses ||| large ||| logp=0 [VP] ||| [V,1] [NP,2] ||| [1] [2] ||| logp=0 [V] ||| sah ||| saw ||| logp=0 [V] ||| fand ||| found ||| logp=0 diff --git a/training/latent_svm/Makefile.am b/training/latent_svm/Makefile.am new file mode 100644 index 00000000..65c5e038 --- /dev/null +++ b/training/latent_svm/Makefile.am @@ -0,0 +1,6 @@ +bin_PROGRAMS = latent_svm + +latent_svm_SOURCES = latent_svm.cc +latent_svm_LDADD = ../..//decoder/libcdec.a ../../klm/search/libksearch.a ../../mteval/libmteval.a ../../utils/libutils.a ../../klm/lm/libklm.a ../../klm/util/libklm_util.a ../../klm/util/double-conversion/libklm_util_double.a + +AM_CPPFLAGS = -W -Wall -Wno-sign-compare -I$(top_srcdir)/utils -I$(top_srcdir)/decoder -I$(top_srcdir)/mteval diff --git a/training/latent_svm/latent_svm.cc b/training/latent_svm/latent_svm.cc new file mode 100644 index 00000000..ab9c1d5d --- /dev/null +++ b/training/latent_svm/latent_svm.cc @@ -0,0 +1,412 @@ +/* +Points to note regarding variable names: +total_loss and prev_loss actually refer not to loss, but the metric (usually BLEU) +*/ +#include <sstream> +#include <iostream> +#include <vector> +#include <cassert> +#include <cmath> + +//boost libraries +#include <boost/shared_ptr.hpp> +#include <boost/program_options.hpp> +#include <boost/program_options/variables_map.hpp> + +//cdec libraries +#include "config.h" +#include "hg_sampler.h" +#include "sentence_metadata.h" +#include "scorer.h" +#include "verbose.h" +#include "viterbi.h" +#include "hg.h" +#include "prob.h" +#include "kbest.h" +#include "ff_register.h" +#include "decoder.h" +#include "filelib.h" +#include "fdict.h" +#include "weights.h" +#include "sparse_vector.h" +#include "sampler.h" + +using namespace std; +using boost::shared_ptr; +namespace po = boost::program_options; + +bool invert_score; +boost::shared_ptr<MT19937> rng; //random seed ptr + +void RandomPermutation(int len, vector<int>* p_ids) { + vector<int>& ids = *p_ids; + ids.resize(len); + for (int i = 0; i < len; ++i) ids[i] = i; + for (int i = len; i > 0; --i) { + int j = rng->next() * i; + if (j == i) i--; + swap(ids[i-1], ids[j]); + } +} + +bool InitCommandLine(int argc, char** argv, po::variables_map* conf) { + po::options_description opts("Configuration options"); + opts.add_options() + ("weights,w",po::value<string>(),"[REQD] Input feature weights file") + ("input,i",po::value<string>(),"[REQD] Input source file for development set") + ("passes,p", po::value<int>()->default_value(15), "Number of passes through the training data") + ("weights_write_interval,n", po::value<int>()->default_value(1000), "Number of lines between writing out weights") + ("reference,r",po::value<vector<string> >(), "[REQD] Reference translation(s) (tokenized text file)") + ("mt_metric,m",po::value<string>()->default_value("ibm_bleu"), "Scoring metric (ibm_bleu, nist_bleu, koehn_bleu, ter, combi)") + ("regularizer_strength,C", po::value<double>()->default_value(0.01), "regularization strength") + ("mt_metric_scale,s", po::value<double>()->default_value(1.0), "Cost function is -mt_metric_scale*BLEU") + ("costaug_log_bleu,l", "Flag converts BLEU to log space. Cost function is thus -mt_metric_scale*log(BLEU). Not on by default") + ("average,A", "Average the weights (this is a weighted average due to the scaling factor)") + ("mu,u", po::value<double>()->default_value(0.0), "weight (between 0 and 1) to scale model score by for oracle selection") + ("stepsize_param,a", po::value<double>()->default_value(0.01), "Stepsize parameter, during optimization") + ("stepsize_reduce,t", "Divide step size by sqrt(number of examples seen so far), as per Ratliff et al., 2007") + ("metric_threshold,T", po::value<double>()->default_value(0.0), "Threshold for diff between oracle BLEU and cost-aug BLEU for updating the weights") + ("check_positive,P", "Check that the loss is positive before updating") + ("k_best_size,k", po::value<int>()->default_value(250), "Size of hypothesis list to search for oracles") + ("best_ever,b", "Keep track of the best hypothesis we've ever seen (metric score), and use that as the reference") + ("random_seed,S", po::value<uint32_t>(), "Random seed (if not specified, /dev/random will be used)") + ("decoder_config,c",po::value<string>(),"Decoder configuration file"); + po::options_description clo("Command line options"); + clo.add_options() + ("config", po::value<string>(), "Configuration file") + ("help,h", "Print this help message and exit"); + po::options_description dconfig_options, dcmdline_options; + dconfig_options.add(opts); + dcmdline_options.add(opts).add(clo); + + po::store(parse_command_line(argc, argv, dcmdline_options), *conf); + if (conf->count("config")) { + ifstream config((*conf)["config"].as<string>().c_str()); + po::store(po::parse_config_file(config, dconfig_options), *conf); + } + po::notify(*conf); + + if (conf->count("help") || !conf->count("weights") || !conf->count("input") || !conf->count("decoder_config") || !conf->count("reference")) { + cerr << dcmdline_options << endl; + return false; + } + return true; +} + +double scaling_trick = 1; // see http://blog.smola.org/post/940672544/fast-quadratic-regularization-for-online-learning +/*computes and returns cost augmented score for negative example selection*/ +double cost_augmented_score(const LogVal<double> model_score, const double mt_metric_score, const double mt_metric_scale, const bool logbleu) { + if(logbleu) { + if(mt_metric_score != 0) + // NOTE: log(model_score) is just the model score feature weights * features + return log(model_score) * scaling_trick + (- mt_metric_scale * log(mt_metric_score)); + else + return -1000000; + } + // NOTE: log(model_score) is just the model score feature weights * features + return log(model_score) * scaling_trick + (- mt_metric_scale * mt_metric_score); +} + +/*computes and returns mu score, for oracle selection*/ +double muscore(const vector<weight_t>& feature_weights, const SparseVector<double>& feature_values, const double mt_metric_score, const double mu, const bool logbleu) { + if(logbleu) { + if(mt_metric_score != 0) + return feature_values.dot(feature_weights) * mu + (1 - mu) * log(mt_metric_score); + else + return feature_values.dot(feature_weights) * mu + (1 - mu) * (-1000000); // log(0) is -inf + } + return feature_values.dot(feature_weights) * mu + (1 - mu) * mt_metric_score; +} + +static const double kMINUS_EPSILON = -1e-6; + +struct HypothesisInfo { + SparseVector<double> features; + double mt_metric_score; + // The model score changes when the feature weights change, so it is not stored here + // It must be recomputed every time +}; + +struct GoodOracle { + shared_ptr<HypothesisInfo> good; +}; + +struct TrainingObserver : public DecoderObserver { + TrainingObserver(const int k, + const DocScorer& d, + vector<GoodOracle>* o, + const vector<weight_t>& feat_weights, + const double metric_scale, + const double Mu, + const bool bestever, + const bool LogBleu) : ds(d), feature_weights(feat_weights), oracles(*o), kbest_size(k), mt_metric_scale(metric_scale), mu(Mu), best_ever(bestever), log_bleu(LogBleu) {} + const DocScorer& ds; + const vector<weight_t>& feature_weights; + vector<GoodOracle>& oracles; + shared_ptr<HypothesisInfo> cur_best; + shared_ptr<HypothesisInfo> cur_costaug_best; + shared_ptr<HypothesisInfo> cur_ref; + const int kbest_size; + const double mt_metric_scale; + const double mu; + const bool best_ever; + const bool log_bleu; + + const HypothesisInfo& GetCurrentBestHypothesis() const { + return *cur_best; + } + + const HypothesisInfo& GetCurrentCostAugmentedHypothesis() const { + return *cur_costaug_best; + } + + const HypothesisInfo& GetCurrentReference() const { + return *cur_ref; + } + + virtual void NotifyTranslationForest(const SentenceMetadata& smeta, Hypergraph* hg) { + UpdateOracles(smeta.GetSentenceID(), *hg); + } + + shared_ptr<HypothesisInfo> MakeHypothesisInfo(const SparseVector<double>& feats, const double metric) { + shared_ptr<HypothesisInfo> h(new HypothesisInfo); + h->features = feats; + h->mt_metric_score = metric; + return h; + } + + void UpdateOracles(int sent_id, const Hypergraph& forest) { + //shared_ptr<HypothesisInfo>& cur_ref = oracles[sent_id].good; + cur_ref = oracles[sent_id].good; + if(!best_ever) + cur_ref.reset(); + + KBest::KBestDerivations<vector<WordID>, ESentenceTraversal> kbest(forest, kbest_size); + double costaug_best_score = 0; + + for (int i = 0; i < kbest_size; ++i) { + const KBest::KBestDerivations<vector<WordID>, ESentenceTraversal>::Derivation* d = + kbest.LazyKthBest(forest.nodes_.size() - 1, i); + if (!d) break; + double mt_metric_score = ds[sent_id]->ScoreCandidate(d->yield)->ComputeScore(); //this might need to change!! + const SparseVector<double>& feature_vals = d->feature_values; + double costaugmented_score = cost_augmented_score(d->score, mt_metric_score, mt_metric_scale, log_bleu); //note that d->score, i.e., model score, is passed in + if (i == 0) { //i.e., setting up cur_best to be model score highest, and initializing costaug_best + cur_best = MakeHypothesisInfo(feature_vals, mt_metric_score); + cur_costaug_best = cur_best; + costaug_best_score = costaugmented_score; + } + if (costaugmented_score > costaug_best_score) { // kbest_mira's cur_bad, i.e., "fear" derivation + cur_costaug_best = MakeHypothesisInfo(feature_vals, mt_metric_score); + costaug_best_score = costaugmented_score; + } + double cur_muscore = mt_metric_score; + if (!cur_ref) // kbest_mira's cur_good, i.e., "hope" derivation + cur_ref = MakeHypothesisInfo(feature_vals, cur_muscore); + else { + double cur_ref_muscore = cur_ref->mt_metric_score; + if(mu > 0) { //select oracle with mixture of model score and BLEU + cur_ref_muscore = muscore(feature_weights, cur_ref->features, cur_ref->mt_metric_score, mu, log_bleu); + cur_muscore = muscore(feature_weights, d->feature_values, mt_metric_score, mu, log_bleu); + } + if (cur_muscore > cur_ref_muscore) //replace oracle + cur_ref = MakeHypothesisInfo(feature_vals, mt_metric_score); + } + } + } +}; + +void ReadTrainingCorpus(const string& fname, vector<string>* c) { + ReadFile rf(fname); + istream& in = *rf.stream(); + string line; + while(in) { + getline(in, line); + if (!in) break; + c->push_back(line); + } +} + +bool ApproxEqual(double a, double b) { + if (a == b) return true; + return (fabs(a-b)/fabs(b)) < 0.000001; +} + +int main(int argc, char** argv) { + register_feature_functions(); + SetSilent(true); // turn off verbose decoder output + + po::variables_map conf; + if (!InitCommandLine(argc, argv, &conf)) return 1; + + if (conf.count("random_seed")) + rng.reset(new MT19937(conf["random_seed"].as<uint32_t>())); + else + rng.reset(new MT19937); + + const bool best_ever = conf.count("best_ever") > 0; + vector<string> corpus; + ReadTrainingCorpus(conf["input"].as<string>(), &corpus); + + const string metric_name = conf["mt_metric"].as<string>(); //set up scoring; this may need to be changed!! + + ScoreType type = ScoreTypeFromString(metric_name); + if (type == TER) { + invert_score = true; + } else { + invert_score = false; + } + DocScorer ds(type, conf["reference"].as<vector<string> >(), ""); + cerr << "Loaded " << ds.size() << " references for scoring with " << metric_name << endl; + if (ds.size() != corpus.size()) { + cerr << "Mismatched number of references (" << ds.size() << ") and sources (" << corpus.size() << ")\n"; + return 1; + } + + ReadFile ini_rf(conf["decoder_config"].as<string>()); + Decoder decoder(ini_rf.stream()); + + // load initial weights + vector<weight_t>& decoder_weights = decoder.CurrentWeightVector(); //equivalent to "dense_weights" vector in kbest_mira.cc + SparseVector<weight_t> sparse_weights; //equivaelnt to kbest_mira.cc "lambdas" + Weights::InitFromFile(conf["weights"].as<string>(), &decoder_weights); + Weights::InitSparseVector(decoder_weights, &sparse_weights); + + //initializing other algorithm and output parameters + const double c = conf["regularizer_strength"].as<double>(); + const int weights_write_interval = conf["weights_write_interval"].as<int>(); + const double mt_metric_scale = conf["mt_metric_scale"].as<double>(); + const double mu = conf["mu"].as<double>(); + const double metric_threshold = conf["metric_threshold"].as<double>(); + const double stepsize_param = conf["stepsize_param"].as<double>(); //step size in structured SGD optimization step + const bool stepsize_reduce = conf.count("stepsize_reduce") > 0; + const bool costaug_log_bleu = conf.count("costaug_log_bleu") > 0; + const bool average = conf.count("average") > 0; + const bool checkpositive = conf.count("check_positive") > 0; + + assert(corpus.size() > 0); + vector<GoodOracle> oracles(corpus.size()); + TrainingObserver observer(conf["k_best_size"].as<int>(), // kbest size + ds, // doc scorer + &oracles, + decoder_weights, + mt_metric_scale, + mu, + best_ever, + costaug_log_bleu); + int cur_sent = 0; + int line_count = 0; + int normalizer = 0; + double total_loss = 0; + double prev_loss = 0; + int dots = 0; // progess bar + int cur_pass = 0; + SparseVector<double> tot; + tot += sparse_weights; //add initial weights to total + normalizer++; //add 1 to normalizer + int max_iteration = conf["passes"].as<int>(); + string msg = "# LatentSVM tuned weights"; + vector<int> order; + int interval_counter = 0; + RandomPermutation(corpus.size(), &order); //shuffle corpus + while (line_count <= max_iteration * corpus.size()) { //loop over all (passes * num sentences) examples + //if ((interval_counter * 40 / weights_write_interval) > dots) { ++dots; cerr << '.'; } //check this + if ((cur_sent * 40 / corpus.size()) > dots) { ++dots; cerr << '.';} + if (interval_counter == weights_write_interval) { //i.e., we need to write out weights + sparse_weights *= scaling_trick; + tot *= scaling_trick; + scaling_trick = 1; + cerr << " [SENTENCE NUMBER= " << cur_sent << "\n"; + cerr << " [AVG METRIC LAST INTERVAL =" << ((total_loss - prev_loss) / weights_write_interval) << "]\n"; + cerr << " [AVG METRIC THIS PASS THUS FAR =" << (total_loss / cur_sent) << "]\n"; + cerr << " [TOTAL LOSS: =" << total_loss << "\n"; + Weights::ShowLargestFeatures(decoder_weights); + //dots = 0; + interval_counter = 0; + prev_loss = total_loss; + if (average){ + SparseVector<double> x = tot; + x /= normalizer; + ostringstream sa; + sa << "weights.latentsvm-" << line_count/weights_write_interval << "-avg.gz"; + x.init_vector(&decoder_weights); + Weights::WriteToFile(sa.str(), decoder_weights, true, &msg); + } + else { + ostringstream os; + os << "weights.latentsvm-" << line_count/weights_write_interval << ".gz"; + sparse_weights.init_vector(&decoder_weights); + Weights::WriteToFile(os.str(), decoder_weights, true, &msg); + } + } + if (corpus.size() == cur_sent) { //i.e., finished a pass + //cerr << " [AVG METRIC LAST PASS=" << (document_metric_score / corpus.size()) << "]\n"; + cerr << " [AVG METRIC LAST PASS=" << (total_loss / corpus.size()) << "]\n"; + cerr << " TOTAL LOSS: " << total_loss << "\n"; + Weights::ShowLargestFeatures(decoder_weights); + cur_sent = 0; + total_loss = 0; + dots = 0; + if(average) { + SparseVector<double> x = tot; + x /= normalizer; + ostringstream sa; + sa << "weights.latentsvm-pass" << (cur_pass < 10 ? "0" : "") << cur_pass << "-avg.gz"; + x.init_vector(&decoder_weights); + Weights::WriteToFile(sa.str(), decoder_weights, true, &msg); + } + else { + ostringstream os; + os << "weights.latentsvm-pass" << (cur_pass < 10 ? "0" : "") << cur_pass << ".gz"; + Weights::WriteToFile(os.str(), decoder_weights, true, &msg); + } + cur_pass++; + RandomPermutation(corpus.size(), &order); + } + if (cur_sent == 0) { //i.e., starting a new pass + cerr << "PASS " << (line_count / corpus.size() + 1) << endl; + } + sparse_weights.init_vector(&decoder_weights); // copy sparse_weights to the decoder weights + decoder.SetId(order[cur_sent]); //assign current sentence + decoder.Decode(corpus[order[cur_sent]], &observer); // decode/update oracles + + const HypothesisInfo& cur_best = observer.GetCurrentBestHypothesis(); //model score best + const HypothesisInfo& cur_costaug = observer.GetCurrentCostAugmentedHypothesis(); //(model + cost) best; cost = -metric_scale*log(BLEU) or -metric_scale*BLEU + //const HypothesisInfo& cur_ref = *oracles[order[cur_sent]].good; //this oracle-best line only picks based on BLEU + const HypothesisInfo& cur_ref = observer.GetCurrentReference(); //if mu > 0, this mu-mixed oracle will be picked; otherwise, only on BLEU + total_loss += cur_best.mt_metric_score; + + double step_size = stepsize_param; + if (stepsize_reduce){ // w_{t+1} = w_t - stepsize_t * grad(Loss) + step_size /= (sqrt(cur_sent+1.0)); + } + //actual update step - compute gradient, and modify sparse_weights + if(cur_ref.mt_metric_score - cur_costaug.mt_metric_score > metric_threshold) { + const double loss = (cur_costaug.features.dot(decoder_weights) - cur_ref.features.dot(decoder_weights)) * scaling_trick + mt_metric_scale * (cur_ref.mt_metric_score - cur_costaug.mt_metric_score); + if (!checkpositive || loss > 0.0) { //can update either all the time if check positive is off, or only when loss > 0 if it's on + sparse_weights -= cur_costaug.features * step_size / ((1.0-2.0*step_size*c)*scaling_trick); // cost augmented hyp orig - + sparse_weights += cur_ref.features * step_size / ((1.0-2.0*step_size*c)*scaling_trick); // ref orig + + } + } + scaling_trick *= (1.0 - 2.0 * step_size * c); + + tot += sparse_weights; //for averaging purposes + normalizer++; //for averaging purposes + line_count++; + interval_counter++; + cur_sent++; + } + cerr << endl; + if(average) { + tot /= normalizer; + tot.init_vector(decoder_weights); + msg = "# Latent SSVM tuned weights (averaged vector)"; + Weights::WriteToFile("weights.latentsvm-final-avg.gz", decoder_weights, true, &msg); + cerr << "Optimization complete.\n" << "AVERAGED WEIGHTS: weights.latentsvm-final-avg.gz\n"; + } else { + Weights::WriteToFile("weights.latentsvm-final.gz", decoder_weights, true, &msg); + cerr << "Optimization complete.\n"; + } + return 0; +} + diff --git a/training/mira/Makefile.am b/training/mira/Makefile.am index fa4fb22d..8cddc2d7 100644 --- a/training/mira/Makefile.am +++ b/training/mira/Makefile.am @@ -1,6 +1,11 @@ -bin_PROGRAMS = kbest_mira +bin_PROGRAMS = kbest_mira \ + kbest_cut_mira kbest_mira_SOURCES = kbest_mira.cc kbest_mira_LDADD = ../../decoder/libcdec.a ../../klm/search/libksearch.a ../../mteval/libmteval.a ../../utils/libutils.a ../../klm/lm/libklm.a ../../klm/util/libklm_util.a ../../klm/util/double-conversion/libklm_util_double.a + +kbest_cut_mira_SOURCES = kbest_cut_mira.cc +kbest_cut_mira_LDADD = ../../decoder/libcdec.a ../../klm/search/libksearch.a ../../mteval/libmteval.a ../../utils/libutils.a ../../klm/lm/libklm.a ../../klm/util/libklm_util.a ../../klm/util/double-conversion/libklm_util_double.a + AM_CPPFLAGS = -W -Wall -Wno-sign-compare -I$(top_srcdir)/utils -I$(top_srcdir)/decoder -I$(top_srcdir)/mteval diff --git a/training/mira/kbest_cut_mira.cc b/training/mira/kbest_cut_mira.cc new file mode 100644 index 00000000..7df9a18f --- /dev/null +++ b/training/mira/kbest_cut_mira.cc @@ -0,0 +1,954 @@ +#include <sstream> +#include <iostream> +#include <vector> +#include <cassert> +#include <cmath> +#include <algorithm> + +#include "config.h" + + +#include <boost/shared_ptr.hpp> +#include <boost/program_options.hpp> +#include <boost/program_options/variables_map.hpp> + +#include "sentence_metadata.h" +#include "scorer.h" +#include "verbose.h" +#include "viterbi.h" +#include "hg.h" +#include "prob.h" +#include "kbest.h" +#include "ff_register.h" +#include "decoder.h" +#include "filelib.h" +#include "fdict.h" +#include "time.h" +#include "sampler.h" + +#include "weights.h" +#include "sparse_vector.h" + +using namespace std; +using boost::shared_ptr; +namespace po = boost::program_options; + +bool invert_score; +boost::shared_ptr<MT19937> rng; +bool approx_score; +bool no_reweight; +bool no_select; +bool unique_kbest; +int update_list_size; +vector<weight_t> dense_w_local; +double mt_metric_scale; +int optimizer; +int fear_select; +int hope_select; +bool pseudo_doc; +bool sent_approx; +bool checkloss; + +void SanityCheck(const vector<double>& w) { + for (int i = 0; i < w.size(); ++i) { + assert(!isnan(w[i])); + assert(!isinf(w[i])); + } +} + +struct FComp { + const vector<double>& w_; + FComp(const vector<double>& w) : w_(w) {} + bool operator()(int a, int b) const { + return fabs(w_[a]) > fabs(w_[b]); + } +}; + +void ShowLargestFeatures(const vector<double>& w) { + vector<int> fnums(w.size()); + for (int i = 0; i < w.size(); ++i) + fnums[i] = i; + vector<int>::iterator mid = fnums.begin(); + mid += (w.size() > 10 ? 10 : w.size()); + partial_sort(fnums.begin(), mid, fnums.end(), FComp(w)); + cerr << "TOP FEATURES:"; + for (vector<int>::iterator i = fnums.begin(); i != mid; ++i) { + cerr << ' ' << FD::Convert(*i) << '=' << w[*i]; + } + cerr << endl; +} + +bool InitCommandLine(int argc, char** argv, po::variables_map* conf) { + po::options_description opts("Configuration options"); + opts.add_options() + ("input_weights,w",po::value<string>(),"Input feature weights file") + ("source,i",po::value<string>(),"Source file for development set") + ("pass,p", po::value<int>()->default_value(15), "Current pass through the training data") + ("reference,r",po::value<vector<string> >(), "[REQD] Reference translation(s) (tokenized text file)") + ("mt_metric,m",po::value<string>()->default_value("ibm_bleu"), "Scoring metric (ibm_bleu, nist_bleu, koehn_bleu, ter, combi)") + ("optimizer,o",po::value<int>()->default_value(1), "Optimizer (SGD=1, PA MIRA w/Delta=2, Cutting Plane MIRA=3, PA MIRA=4, Triple nbest list MIRA=5)") + ("fear,f",po::value<int>()->default_value(1), "Fear selection (model-cost=1, maxcost=2, maxscore=3)") + ("hope,h",po::value<int>()->default_value(1), "Hope selection (model+cost=1, mincost=2)") + ("max_step_size,C", po::value<double>()->default_value(0.01), "regularization strength (C)") + ("random_seed,S", po::value<uint32_t>(), "Random seed (if not specified, /dev/random will be used)") + ("mt_metric_scale,s", po::value<double>()->default_value(1.0), "Amount to scale MT loss function by") + ("sent_approx,a", "Use smoothed sentence-level BLEU score for approximate scoring") + ("pseudo_doc,e", "Use pseudo-document BLEU score for approximate scoring") + ("no_reweight,d","Do not reweight forest for cutting plane") + ("no_select,n", "Do not use selection heuristic") + ("k_best_size,k", po::value<int>()->default_value(250), "Size of hypothesis list to search for oracles") + ("update_k_best,b", po::value<int>()->default_value(1), "Size of good, bad lists to perform update with") + ("unique_k_best,u", "Unique k-best translation list") + ("weights_output,O",po::value<string>(),"Directory to write weights to") + ("output_dir,D",po::value<string>(),"Directory to place output in") + ("decoder_config,c",po::value<string>(),"Decoder configuration file"); + po::options_description clo("Command line options"); + clo.add_options() + ("config", po::value<string>(), "Configuration file") + ("help,H", "Print this help message and exit"); + po::options_description dconfig_options, dcmdline_options; + dconfig_options.add(opts); + dcmdline_options.add(opts).add(clo); + + po::store(parse_command_line(argc, argv, dcmdline_options), *conf); + if (conf->count("config")) { + ifstream config((*conf)["config"].as<string>().c_str()); + po::store(po::parse_config_file(config, dconfig_options), *conf); + } + po::notify(*conf); + + if (conf->count("help") || !conf->count("input_weights") || !conf->count("decoder_config") || !conf->count("reference")) { + cerr << dcmdline_options << endl; + return false; + } + return true; +} + +//load previous translation, store array of each sentences score, subtract it from current sentence and replace with new translation score + + +static const double kMINUS_EPSILON = -1e-6; +static const double EPSILON = 0.000001; +static const double SMO_EPSILON = 0.0001; +static const double PSEUDO_SCALE = 0.95; +static const int MAX_SMO = 10; +int cur_pass; + +struct HypothesisInfo { + SparseVector<double> features; + vector<WordID> hyp; + double mt_metric; + double hope; + double fear; + double alpha; + double oracle_loss; + SparseVector<double> oracle_feat_diff; + shared_ptr<HypothesisInfo> oracleN; +}; + +bool ApproxEqual(double a, double b) { + if (a == b) return true; + return (fabs(a-b)/fabs(b)) < EPSILON; +} + +typedef shared_ptr<HypothesisInfo> HI; +bool HypothesisCompareB(const HI& h1, const HI& h2 ) +{ + return h1->mt_metric > h2->mt_metric; +}; + + +bool HopeCompareB(const HI& h1, const HI& h2 ) +{ + return h1->hope > h2->hope; +}; + +bool FearCompareB(const HI& h1, const HI& h2 ) +{ + return h1->fear > h2->fear; +}; + +bool FearComparePred(const HI& h1, const HI& h2 ) +{ + return h1->features.dot(dense_w_local) > h2->features.dot(dense_w_local); +}; + +bool HypothesisCompareG(const HI& h1, const HI& h2 ) +{ + return h1->mt_metric < h2->mt_metric; +}; + + +void CuttingPlane(vector<shared_ptr<HypothesisInfo> >* cur_c, bool* again, vector<shared_ptr<HypothesisInfo> >& all_hyp, vector<weight_t> dense_weights) +{ + bool DEBUG_CUT = false; + shared_ptr<HypothesisInfo> max_fear, max_fear_in_set; + vector<shared_ptr<HypothesisInfo> >& cur_constraint = *cur_c; + + if(no_reweight) + { + //find new hope hypothesis + for(int u=0;u!=all_hyp.size();u++) + { + double t_score = all_hyp[u]->features.dot(dense_weights); + all_hyp[u]->hope = 1 * all_hyp[u]->mt_metric + t_score; + } + + //sort hyps by hope score + sort(all_hyp.begin(),all_hyp.end(),HopeCompareB); + + double hope_score = all_hyp[0]->features.dot(dense_weights); + if(DEBUG_CUT) cerr << "New hope derivation score " << hope_score << endl; + + for(int u=0;u!=all_hyp.size();u++) + { + double t_score = all_hyp[u]->features.dot(dense_weights); + all_hyp[u]->fear = -1*all_hyp[u]->mt_metric + 1*all_hyp[0]->mt_metric - hope_score + t_score; //relative loss + } + + sort(all_hyp.begin(),all_hyp.end(),FearCompareB); + + } + //assign maximum fear derivation from all derivations + max_fear = all_hyp[0]; + + if(DEBUG_CUT) cerr <<"Cutting Plane Max Fear "<<max_fear->fear ; + for(int i=0; i < cur_constraint.size();i++) //select maximal violator already in constraint set + { + if (!max_fear_in_set || cur_constraint[i]->fear > max_fear_in_set->fear) + max_fear_in_set = cur_constraint[i]; + } + if(DEBUG_CUT) cerr << "Max Fear in constraint set " << max_fear_in_set->fear << endl; + + if(max_fear->fear > max_fear_in_set->fear + SMO_EPSILON) + { + cur_constraint.push_back(max_fear); + *again = true; + if(DEBUG_CUT) cerr << "Optimize Again " << *again << endl; + } +} + + +double ComputeDelta(vector<shared_ptr<HypothesisInfo> >* cur_p, double max_step_size,vector<weight_t> dense_weights ) +{ + vector<shared_ptr<HypothesisInfo> >& cur_pair = *cur_p; + double loss = cur_pair[0]->oracle_loss - cur_pair[1]->oracle_loss; + + double margin = -(cur_pair[0]->oracleN->features.dot(dense_weights)- cur_pair[0]->features.dot(dense_weights)) + (cur_pair[1]->oracleN->features.dot(dense_weights) - cur_pair[1]->features.dot(dense_weights)); + const double num = margin + loss; + cerr << "LOSS: " << num << " Margin:" << margin << " BLEUL:" << loss << " " << cur_pair[1]->features.dot(dense_weights) << " " << cur_pair[0]->features.dot(dense_weights) <<endl; + + + SparseVector<double> diff = cur_pair[0]->features; + diff -= cur_pair[1]->features; + double diffsqnorm = diff.l2norm_sq(); + double delta; + if (diffsqnorm > 0) + delta = num / (diffsqnorm * max_step_size); + else + delta = 0; + cerr << " D1:" << delta; + //clip delta (enforce margin constraints) + delta = max(-cur_pair[0]->alpha, min(delta, cur_pair[1]->alpha)); + cerr << " D2:" << delta; + return delta; +} + + +vector<shared_ptr<HypothesisInfo> > SelectPair(vector<shared_ptr<HypothesisInfo> >* cur_c) +{ + bool DEBUG_SELECT= false; + vector<shared_ptr<HypothesisInfo> >& cur_constraint = *cur_c; + + vector<shared_ptr<HypothesisInfo> > pair; + + if (no_select || optimizer == 2){ //skip heuristic search and return oracle and fear for pa-mira + + pair.push_back(cur_constraint[0]); + pair.push_back(cur_constraint[1]); + return pair; + + } + + for(int u=0;u != cur_constraint.size();u++) + { + shared_ptr<HypothesisInfo> max_fear; + + if(DEBUG_SELECT) cerr<< "cur alpha " << u << " " << cur_constraint[u]->alpha; + for(int i=0; i < cur_constraint.size();i++) //select maximal violator + { + if(i != u) + if (!max_fear || cur_constraint[i]->fear > max_fear->fear) + max_fear = cur_constraint[i]; + } + if(!max_fear) return pair; // + + + if ((cur_constraint[u]->alpha == 0) && (cur_constraint[u]->fear > max_fear->fear + SMO_EPSILON)) + { + for(int i=0; i < cur_constraint.size();i++) //select maximal violator + { + if(i != u) + if (cur_constraint[i]->alpha > 0) + { + pair.push_back(cur_constraint[u]); + pair.push_back(cur_constraint[i]); + return pair; + } + } + } + if ((cur_constraint[u]->alpha > 0) && (cur_constraint[u]->fear < max_fear->fear - SMO_EPSILON)) + { + for(int i=0; i < cur_constraint.size();i++) //select maximal violator + { + if(i != u) + if (cur_constraint[i]->fear > cur_constraint[u]->fear) + { + pair.push_back(cur_constraint[u]); + pair.push_back(cur_constraint[i]); + return pair; + } + } + } + + } + return pair; //no more constraints to optimize, we're done here + +} + +struct GoodBadOracle { + vector<shared_ptr<HypothesisInfo> > good; + vector<shared_ptr<HypothesisInfo> > bad; +}; + +struct TrainingObserver : public DecoderObserver { + TrainingObserver(const int k, const DocScorer& d, vector<GoodBadOracle>* o, vector<ScoreP>* cbs) : ds(d), oracles(*o), corpus_bleu_sent_stats(*cbs), kbest_size(k) { + + + if(!pseudo_doc && !sent_approx) + if(cur_pass > 0) //calculate corpus bleu score from previous iterations 1-best for BLEU gain + { + ScoreP acc; + for (int ii = 0; ii < corpus_bleu_sent_stats.size(); ii++) { + if (!acc) { acc = corpus_bleu_sent_stats[ii]->GetZero(); } + acc->PlusEquals(*corpus_bleu_sent_stats[ii]); + + } + corpus_bleu_stats = acc; + corpus_bleu_score = acc->ComputeScore(); + } + +} + const DocScorer& ds; + vector<ScoreP>& corpus_bleu_sent_stats; + vector<GoodBadOracle>& oracles; + vector<shared_ptr<HypothesisInfo> > cur_best; + shared_ptr<HypothesisInfo> cur_oracle; + const int kbest_size; + Hypergraph forest; + int cur_sent; + ScoreP corpus_bleu_stats; + float corpus_bleu_score; + + float corpus_src_length; + float curr_src_length; + + const int GetCurrentSent() const { + return cur_sent; + } + + const HypothesisInfo& GetCurrentBestHypothesis() const { + return *cur_best[0]; + } + + const vector<shared_ptr<HypothesisInfo> > GetCurrentBest() const { + return cur_best; + } + + const HypothesisInfo& GetCurrentOracle() const { + return *cur_oracle; + } + + const Hypergraph& GetCurrentForest() const { + return forest; + } + + + virtual void NotifyTranslationForest(const SentenceMetadata& smeta, Hypergraph* hg) { + cur_sent = smeta.GetSentenceID(); + curr_src_length = (float) smeta.GetSourceLength(); + + if(unique_kbest) + UpdateOracles<KBest::FilterUnique>(smeta.GetSentenceID(), *hg); + else + UpdateOracles<KBest::NoFilter<std::vector<WordID> > >(smeta.GetSentenceID(), *hg); + forest = *hg; + + } + + shared_ptr<HypothesisInfo> MakeHypothesisInfo(const SparseVector<double>& feats, const double score, const vector<WordID>& hyp) { + shared_ptr<HypothesisInfo> h(new HypothesisInfo); + h->features = feats; + h->mt_metric = score; + h->hyp = hyp; + return h; + } + + template <class Filter> + void UpdateOracles(int sent_id, const Hypergraph& forest) { + + bool PRINT_LIST= false; + vector<shared_ptr<HypothesisInfo> >& cur_good = oracles[sent_id].good; + vector<shared_ptr<HypothesisInfo> >& cur_bad = oracles[sent_id].bad; + //TODO: look at keeping previous iterations hypothesis lists around + cur_best.clear(); + cur_good.clear(); + cur_bad.clear(); + + vector<shared_ptr<HypothesisInfo> > all_hyp; + + typedef KBest::KBestDerivations<vector<WordID>, ESentenceTraversal,Filter> K; + K kbest(forest,kbest_size); + + for (int i = 0; i < kbest_size; ++i) { + + typename K::Derivation *d = + kbest.LazyKthBest(forest.nodes_.size() - 1, i); + if (!d) break; + + float sentscore; + if(cur_pass > 0 && !pseudo_doc && !sent_approx) + { + ScoreP sent_stats = ds[sent_id]->ScoreCandidate(d->yield); + ScoreP corpus_no_best = corpus_bleu_stats->GetZero(); + + corpus_bleu_stats->Subtract(*corpus_bleu_sent_stats[sent_id], &*corpus_no_best); + sent_stats->PlusEquals(*corpus_no_best, 0.5); + + //compute gain from new sentence in 1-best corpus + sentscore = mt_metric_scale * (sent_stats->ComputeScore() - corpus_no_best->ComputeScore());// - corpus_bleu_score); + } + else if(pseudo_doc) //pseudo-corpus smoothing + { + float src_scale = corpus_src_length + curr_src_length; + ScoreP sent_stats = ds[sent_id]->ScoreCandidate(d->yield); + if(!corpus_bleu_stats){ corpus_bleu_stats = sent_stats->GetZero();} + + sent_stats->PlusEquals(*corpus_bleu_stats); + sentscore = mt_metric_scale * src_scale * sent_stats->ComputeScore(); + + } + else //use sentence-level smoothing ( used when cur_pass=0 if not pseudo_doc) + { + + sentscore = mt_metric_scale * (ds[sent_id]->ScoreCandidate(d->yield)->ComputeScore()); + } + + if (invert_score) sentscore *= -1.0; + + if (i < update_list_size){ + if(PRINT_LIST)cerr << TD::GetString(d->yield) << " ||| " << d->score << " ||| " << sentscore << endl; + cur_best.push_back( MakeHypothesisInfo(d->feature_values, sentscore, d->yield)); + } + + all_hyp.push_back(MakeHypothesisInfo(d->feature_values, sentscore,d->yield)); //store all hyp to extract hope and fear + } + + if(pseudo_doc){ + //update psuedo-doc stats + string details, details2; + corpus_bleu_stats->ScoreDetails(&details2); + ScoreP sent_stats = ds[sent_id]->ScoreCandidate(cur_best[0]->hyp); + corpus_bleu_stats->PlusEquals(*sent_stats); + + sent_stats->ScoreDetails(&details); + sent_stats = corpus_bleu_stats; + corpus_bleu_stats = sent_stats->GetZero(); + corpus_bleu_stats->PlusEquals(*sent_stats, PSEUDO_SCALE); + + corpus_src_length = PSEUDO_SCALE * (corpus_src_length + curr_src_length); + cerr << "ps corpus size: " << corpus_src_length << " " << curr_src_length << "\n" << details << "\n" << details2 << endl; + } + + //figure out how many hyps we can keep maximum + int temp_update_size = update_list_size; + if (all_hyp.size() < update_list_size){ temp_update_size = all_hyp.size();} + + //sort all hyps by sentscore (eg. bleu) + sort(all_hyp.begin(),all_hyp.end(),HypothesisCompareB); + + if(PRINT_LIST){ cerr << "Sorting " << endl; for(int u=0;u!=all_hyp.size();u++) + cerr << all_hyp[u]->mt_metric << " " << all_hyp[u]->features.dot(dense_w_local) << endl; } + + if(hope_select == 1) + { + //find hope hypothesis using model + bleu + if (PRINT_LIST) cerr << "HOPE " << endl; + for(int u=0;u!=all_hyp.size();u++) + { + double t_score = all_hyp[u]->features.dot(dense_w_local); + all_hyp[u]->hope = all_hyp[u]->mt_metric + t_score; + if (PRINT_LIST) cerr << all_hyp[u]->mt_metric << " H:" << all_hyp[u]->hope << " S:" << t_score << endl; + + } + + //sort hyps by hope score + sort(all_hyp.begin(),all_hyp.end(),HopeCompareB); + } + + //assign cur_good the sorted list + cur_good.insert(cur_good.begin(), all_hyp.begin(), all_hyp.begin()+temp_update_size); + if(PRINT_LIST) { cerr << "GOOD" << endl; for(int u=0;u!=cur_good.size();u++) cerr << cur_good[u]->mt_metric << " " << cur_good[u]->hope << endl;} + + //use hope for fear selection + shared_ptr<HypothesisInfo>& oracleN = cur_good[0]; + + if(fear_select == 1){ //compute fear hyps with model - bleu + if (PRINT_LIST) cerr << "FEAR " << endl; + double hope_score = oracleN->features.dot(dense_w_local); + + if (PRINT_LIST) cerr << "hope score " << hope_score << endl; + for(int u=0;u!=all_hyp.size();u++) + { + double t_score = all_hyp[u]->features.dot(dense_w_local); + + all_hyp[u]->fear = -1*all_hyp[u]->mt_metric + 1*oracleN->mt_metric - hope_score + t_score; //relative loss + all_hyp[u]->oracle_loss = -1*all_hyp[u]->mt_metric + 1*oracleN->mt_metric; + all_hyp[u]->oracle_feat_diff = oracleN->features - all_hyp[u]->features; + all_hyp[u]->oracleN=oracleN; + if (PRINT_LIST) cerr << all_hyp[u]->mt_metric << " H:" << all_hyp[u]->hope << " F:" << all_hyp[u]->fear << endl; + + } + + sort(all_hyp.begin(),all_hyp.end(),FearCompareB); + + } + else if(fear_select == 2) //select fear based on cost + { + sort(all_hyp.begin(),all_hyp.end(),HypothesisCompareG); + } + else //max model score, also known as prediction-based + { + sort(all_hyp.begin(),all_hyp.end(),FearComparePred); + } + cur_bad.insert(cur_bad.begin(), all_hyp.begin(), all_hyp.begin()+temp_update_size); + + if(PRINT_LIST){ cerr<< "BAD"<<endl; for(int u=0;u!=cur_bad.size();u++) cerr << cur_bad[u]->mt_metric << " H:" << cur_bad[u]->hope << " F:" << cur_bad[u]->fear << endl;} + + cerr << "GOOD (BEST): " << cur_good[0]->mt_metric << endl; + cerr << " CUR: " << cur_best[0]->mt_metric << endl; + cerr << " BAD (WORST): " << cur_bad[0]->mt_metric << endl; + } +}; + +void ReadTrainingCorpus(const string& fname, vector<string>* c) { + + + ReadFile rf(fname); + istream& in = *rf.stream(); + string line; + while(in) { + getline(in, line); + if (!in) break; + c->push_back(line); + } +} + +void ReadPastTranslationForScore(const int cur_pass, vector<ScoreP>* c, DocScorer& ds, const string& od) +{ + cerr << "Reading BLEU gain file "; + string fname; + if(cur_pass == 0) + { + fname = od + "/run.raw.init"; + } + else + { + int last_pass = cur_pass - 1; + fname = od + "/run.raw." + boost::lexical_cast<std::string>(last_pass) + ".B"; + } + cerr << fname << "\n"; + ReadFile rf(fname); + istream& in = *rf.stream(); + ScoreP acc; + string line; + int lc = 0; + while(in) { + getline(in, line); + if (line.empty() && !in) break; + vector<WordID> sent; + TD::ConvertSentence(line, &sent); + ScoreP sentscore = ds[lc]->ScoreCandidate(sent); + c->push_back(sentscore); + if (!acc) { acc = sentscore->GetZero(); } + acc->PlusEquals(*sentscore); + ++lc; + + } + + assert(lc > 0); + float score = acc->ComputeScore(); + string details; + acc->ScoreDetails(&details); + cerr << "Previous run: " << details << score << endl; + +} + + +int main(int argc, char** argv) { + register_feature_functions(); + SetSilent(true); // turn off verbose decoder output + + po::variables_map conf; + if (!InitCommandLine(argc, argv, &conf)) return 1; + + if (conf.count("random_seed")) + rng.reset(new MT19937(conf["random_seed"].as<uint32_t>())); + else + rng.reset(new MT19937); + + vector<string> corpus; + + const string metric_name = conf["mt_metric"].as<string>(); + optimizer = conf["optimizer"].as<int>(); + fear_select = conf["fear"].as<int>(); + hope_select = conf["hope"].as<int>(); + mt_metric_scale = conf["mt_metric_scale"].as<double>(); + approx_score = conf.count("approx_score"); + no_reweight = conf.count("no_reweight"); + no_select = conf.count("no_select"); + update_list_size = conf["update_k_best"].as<int>(); + unique_kbest = conf.count("unique_k_best"); + pseudo_doc = conf.count("pseudo_doc"); + sent_approx = conf.count("sent_approx"); + cerr << "Using pseudo-doc:" << pseudo_doc << " Sent:" << sent_approx << endl; + if(pseudo_doc) + mt_metric_scale=1; + + const string weights_dir = conf["weights_output"].as<string>(); + const string output_dir = conf["output_dir"].as<string>(); + ScoreType type = ScoreTypeFromString(metric_name); + + //establish metric used for tuning + if (type == TER) { + invert_score = true; + } else { + invert_score = false; + } + + //load references + DocScorer ds(type, conf["reference"].as<vector<string> >(), ""); + cerr << "Loaded " << ds.size() << " references for scoring with " << metric_name << endl; + vector<ScoreP> corpus_bleu_sent_stats; + + //check training pass,if >0, then use previous iterations corpus bleu stats + cur_pass = conf["pass"].as<int>(); + if(cur_pass > 0) + { + ReadPastTranslationForScore(cur_pass, &corpus_bleu_sent_stats, ds, output_dir); + } + + cerr << "Using optimizer:" << optimizer << endl; + + ReadFile ini_rf(conf["decoder_config"].as<string>()); + Decoder decoder(ini_rf.stream()); + + vector<weight_t>& dense_weights = decoder.CurrentWeightVector(); + + SparseVector<weight_t> lambdas; + Weights::InitFromFile(conf["input_weights"].as<string>(), &dense_weights); + Weights::InitSparseVector(dense_weights, &lambdas); + + const string input = decoder.GetConf()["input"].as<string>(); + if (!SILENT) cerr << "Reading input from " << ((input == "-") ? "STDIN" : input.c_str()) << endl; + ReadFile in_read(input); + istream *in = in_read.stream(); + assert(*in); + string buf; + + const double max_step_size = conf["max_step_size"].as<double>(); + + vector<GoodBadOracle> oracles(ds.size()); + + TrainingObserver observer(conf["k_best_size"].as<int>(), ds, &oracles, &corpus_bleu_sent_stats); + + int cur_sent = 0; + int lcount = 0; + double objective=0; + double tot_loss = 0; + int dots = 0; + SparseVector<double> tot; + SparseVector<double> final_tot; + + SparseVector<double> old_lambdas = lambdas; + tot.clear(); + tot += lambdas; + cerr << "PASS " << cur_pass << " " << endl << lambdas << endl; + ScoreP acc, acc_h, acc_f; + + while(*in) { + getline(*in, buf); + if (buf.empty()) continue; + //TODO: allow batch updating + lambdas.init_vector(&dense_weights); + dense_w_local = dense_weights; + decoder.SetId(cur_sent); + decoder.Decode(buf, &observer); // decode the sentence, calling Notify to get the hope,fear, and model best hyps. + + cur_sent = observer.GetCurrentSent(); + cerr << "SENT: " << cur_sent << endl; + const HypothesisInfo& cur_hyp = observer.GetCurrentBestHypothesis(); + const HypothesisInfo& cur_good = *oracles[cur_sent].good[0]; + const HypothesisInfo& cur_bad = *oracles[cur_sent].bad[0]; + + vector<shared_ptr<HypothesisInfo> >& cur_good_v = oracles[cur_sent].good; + vector<shared_ptr<HypothesisInfo> >& cur_bad_v = oracles[cur_sent].bad; + vector<shared_ptr<HypothesisInfo> > cur_best_v = observer.GetCurrentBest(); + + tot_loss += cur_hyp.mt_metric; + + //score hyps to be able to compute corpus level bleu after we finish this iteration through the corpus + ScoreP sentscore = ds[cur_sent]->ScoreCandidate(cur_hyp.hyp); + if (!acc) { acc = sentscore->GetZero(); } + acc->PlusEquals(*sentscore); + + ScoreP hope_sentscore = ds[cur_sent]->ScoreCandidate(cur_good.hyp); + if (!acc_h) { acc_h = hope_sentscore->GetZero(); } + acc_h->PlusEquals(*hope_sentscore); + + ScoreP fear_sentscore = ds[cur_sent]->ScoreCandidate(cur_bad.hyp); + if (!acc_f) { acc_f = fear_sentscore->GetZero(); } + acc_f->PlusEquals(*fear_sentscore); + + if(optimizer == 4) { //passive-aggresive update (single dual coordinate step) + + double margin = cur_bad.features.dot(dense_weights) - cur_good.features.dot(dense_weights); + double mt_loss = (cur_good.mt_metric - cur_bad.mt_metric); + const double loss = margin + mt_loss; + cerr << "LOSS: " << loss << " Margin:" << margin << " BLEUL:" << mt_loss << " " << cur_bad.features.dot(dense_weights) << " " << cur_good.features.dot(dense_weights) <<endl; + if (loss > 0.0 || !checkloss) { + SparseVector<double> diff = cur_good.features; + diff -= cur_bad.features; + + double diffsqnorm = diff.l2norm_sq(); + double delta; + if (diffsqnorm > 0) + delta = loss / (diffsqnorm); + else + delta = 0; + + if (delta > max_step_size) delta = max_step_size; + lambdas += (cur_good.features * delta); + lambdas -= (cur_bad.features * delta); + + } + } + else if(optimizer == 1) //sgd - nonadapted step size + { + + lambdas += (cur_good.features) * max_step_size; + lambdas -= (cur_bad.features) * max_step_size; + } + else if(optimizer == 5) //full mira with n-best list of constraints from hope, fear, model best + { + vector<shared_ptr<HypothesisInfo> > cur_constraint; + cur_constraint.insert(cur_constraint.begin(), cur_bad_v.begin(), cur_bad_v.end()); + cur_constraint.insert(cur_constraint.begin(), cur_best_v.begin(), cur_best_v.end()); + cur_constraint.insert(cur_constraint.begin(), cur_good_v.begin(), cur_good_v.end()); + + bool optimize_again; + vector<shared_ptr<HypothesisInfo> > cur_pair; + //SMO + for(int u=0;u!=cur_constraint.size();u++) + cur_constraint[u]->alpha =0; + + cur_constraint[0]->alpha =1; //set oracle to alpha=1 + + cerr <<"Optimizing with " << cur_constraint.size() << " constraints" << endl; + int smo_iter = MAX_SMO, smo_iter2 = MAX_SMO; + int iter, iter2 =0; + bool DEBUG_SMO = false; + while (iter2 < smo_iter2) + { + iter =0; + while (iter < smo_iter) + { + optimize_again = true; + for (int i = 0; i< cur_constraint.size(); i++) + for (int j = i+1; j< cur_constraint.size(); j++) + { + if(DEBUG_SMO) cerr << "start " << i << " " << j << endl; + cur_pair.clear(); + cur_pair.push_back(cur_constraint[j]); + cur_pair.push_back(cur_constraint[i]); + double delta = ComputeDelta(&cur_pair,max_step_size, dense_weights); + + if (delta == 0) optimize_again = false; + cur_constraint[j]->alpha += delta; + cur_constraint[i]->alpha -= delta; + double step_size = delta * max_step_size; + + lambdas += (cur_constraint[i]->features) * step_size; + lambdas -= (cur_constraint[j]->features) * step_size; + if(DEBUG_SMO) cerr << "SMO opt " << iter << " " << i << " " << j << " " << delta << " " << cur_pair[0]->alpha << " " << cur_pair[1]->alpha << endl; + } + iter++; + + if(!optimize_again) + { + iter = MAX_SMO; + cerr << "Optimization stopped, delta =0" << endl; + } + } + iter2++; + } + } + else if(optimizer == 2 || optimizer == 3) //PA and Cutting Plane MIRA update + { + bool DEBUG_SMO= true; + vector<shared_ptr<HypothesisInfo> > cur_constraint; + cur_constraint.push_back(cur_good_v[0]); //add oracle to constraint set + bool optimize_again = true; + int cut_plane_calls = 0; + while (optimize_again) + { + if(DEBUG_SMO) cerr<< "optimize again: " << optimize_again << endl; + if(optimizer == 2){ //PA + cur_constraint.push_back(cur_bad_v[0]); + + //check if we have a violation + if(!(cur_constraint[1]->fear > cur_constraint[0]->fear + SMO_EPSILON)) + { + optimize_again = false; + cerr << "Constraint not violated" << endl; + } + } + else + { //cutting plane to add constraints + if(DEBUG_SMO) cerr<< "Cutting Plane " << cut_plane_calls << " with " << lambdas << endl; + optimize_again = false; + cut_plane_calls++; + CuttingPlane(&cur_constraint, &optimize_again, oracles[cur_sent].bad, dense_weights); + if (cut_plane_calls >= MAX_SMO) optimize_again = false; + } + + if(optimize_again) + { + //SMO + for(int u=0;u!=cur_constraint.size();u++) + { + cur_constraint[u]->alpha =0; + } + cur_constraint[0]->alpha = 1; + cerr <<" Optimizing with " << cur_constraint.size() << " constraints" << endl; + int smo_iter = MAX_SMO; + int iter =0; + while (iter < smo_iter) + { + //select pair to optimize from constraint set + vector<shared_ptr<HypothesisInfo> > cur_pair = SelectPair(&cur_constraint); + + if(cur_pair.empty()){ + iter=MAX_SMO; + cerr << "Undefined pair " << endl; + continue; + } //pair is undefined so we are done with this smo + + double delta = ComputeDelta(&cur_pair,max_step_size, dense_weights); + + cur_pair[0]->alpha += delta; + cur_pair[1]->alpha -= delta; + double step_size = delta * max_step_size; + cerr << "step " << step_size << endl; + + lambdas += (cur_pair[1]->features) * step_size; + lambdas -= (cur_pair[0]->features) * step_size; + cerr << " Lambdas " << lambdas << endl; + //reload weights based on update + + dense_weights.clear(); + lambdas.init_vector(&dense_weights); + dense_w_local = dense_weights; + iter++; + + if(DEBUG_SMO) cerr << "SMO opt " << iter << " " << delta << " " << cur_pair[0]->alpha << " " << cur_pair[1]->alpha << endl; + if(no_select) //don't use selection heuristic to determine when to stop SMO, rather just when delta =0 + if (delta == 0) iter = MAX_SMO; + + //only perform one dual coordinate ascent step + if(optimizer == 2) + { + optimize_again = false; + iter = MAX_SMO; + } + } + if(optimizer == 3) + { + if(!no_reweight) //reweight the forest and select a new k-best + { + if(DEBUG_SMO) cerr<< "Decoding with new weights -- now orac are " << oracles[cur_sent].good.size() << endl; + Hypergraph hg = observer.GetCurrentForest(); + hg.Reweight(dense_weights); + if(unique_kbest) + observer.UpdateOracles<KBest::FilterUnique>(cur_sent, hg); + else + observer.UpdateOracles<KBest::NoFilter<std::vector<WordID> > >(cur_sent, hg); + } + } + } + + } + + //print objective after this sentence + double lambda_change = (lambdas - old_lambdas).l2norm_sq(); + double max_fear = cur_constraint[cur_constraint.size()-1]->fear; + double temp_objective = 0.5 * lambda_change;// + max_step_size * max_fear; + + for(int u=0;u!=cur_constraint.size();u++) + { + cerr << cur_constraint[u]->alpha << " " << cur_constraint[u]->hope << " " << cur_constraint[u]->fear << endl; + temp_objective += cur_constraint[u]->alpha * cur_constraint[u]->fear; + } + objective += temp_objective; + + cerr << "SENT OBJ: " << temp_objective << " NEW OBJ: " << objective << endl; + } + + + if ((cur_sent * 40 / ds.size()) > dots) { ++dots; cerr << '.'; } + tot += lambdas; + ++lcount; + cur_sent++; + + cout << TD::GetString(cur_good_v[0]->hyp) << " ||| " << TD::GetString(cur_best_v[0]->hyp) << " ||| " << TD::GetString(cur_bad_v[0]->hyp) << endl; + + } + + cerr << "FINAL OBJECTIVE: "<< objective << endl; + final_tot += tot; + cerr << "Translated " << lcount << " sentences " << endl; + cerr << " [AVG METRIC LAST PASS=" << (tot_loss / lcount) << "]\n"; + tot_loss = 0; + + int node_id = rng->next() * 100000; + cerr << " Writing weights to " << node_id << endl; + Weights::ShowLargestFeatures(dense_weights); + dots = 0; + ostringstream os; + os << weights_dir << "/weights.mira-pass" << (cur_pass < 10 ? "0" : "") << cur_pass << "." << node_id << ".gz"; + string msg = "# MIRA tuned weights ||| " + boost::lexical_cast<std::string>(node_id) + " ||| " + boost::lexical_cast<std::string>(lcount); + lambdas.init_vector(&dense_weights); + Weights::WriteToFile(os.str(), dense_weights, true, &msg); + + SparseVector<double> x = tot; + x /= lcount+1; + ostringstream sa; + string msga = "# MIRA tuned weights AVERAGED ||| " + boost::lexical_cast<std::string>(node_id) + " ||| " + boost::lexical_cast<std::string>(lcount); + sa << weights_dir << "/weights.mira-pass" << (cur_pass < 10 ? "0" : "") << cur_pass << "." << node_id << "-avg.gz"; + x.init_vector(&dense_weights); + Weights::WriteToFile(sa.str(), dense_weights, true, &msga); + + cerr << "Optimization complete.\n"; + return 0; +} + diff --git a/training/mira/kbest_mira.cc b/training/mira/kbest_mira.cc index 8b7993dd..d59b4224 100644 --- a/training/mira/kbest_mira.cc +++ b/training/mira/kbest_mira.cc @@ -8,9 +8,11 @@ #include <boost/program_options.hpp> #include <boost/program_options/variables_map.hpp> +#include "stringlib.h" #include "hg_sampler.h" #include "sentence_metadata.h" -#include "scorer.h" +#include "ns.h" +#include "ns_docscorer.h" #include "verbose.h" #include "viterbi.h" #include "hg.h" @@ -91,8 +93,9 @@ struct GoodBadOracle { }; struct TrainingObserver : public DecoderObserver { - TrainingObserver(const int k, const DocScorer& d, bool sf, vector<GoodBadOracle>* o) : ds(d), oracles(*o), kbest_size(k), sample_forest(sf) {} - const DocScorer& ds; + TrainingObserver(const int k, const DocumentScorer& d, const EvaluationMetric& m, bool sf, vector<GoodBadOracle>* o) : ds(d), metric(m), oracles(*o), kbest_size(k), sample_forest(sf) {} + const DocumentScorer& ds; + const EvaluationMetric& metric; vector<GoodBadOracle>& oracles; std::tr1::shared_ptr<HypothesisInfo> cur_best; const int kbest_size; @@ -121,13 +124,16 @@ struct TrainingObserver : public DecoderObserver { if (sample_forest) { vector<WordID> cur_prediction; ViterbiESentence(forest, &cur_prediction); - float sentscore = ds[sent_id]->ScoreCandidate(cur_prediction)->ComputeScore(); + SufficientStats sstats; + ds[sent_id]->Evaluate(cur_prediction, &sstats); + float sentscore = metric.ComputeScore(sstats); cur_best = MakeHypothesisInfo(ViterbiFeatures(forest), sentscore); vector<HypergraphSampler::Hypothesis> samples; HypergraphSampler::sample_hypotheses(forest, kbest_size, &*rng, &samples); for (unsigned i = 0; i < samples.size(); ++i) { - sentscore = ds[sent_id]->ScoreCandidate(samples[i].words)->ComputeScore(); + ds[sent_id]->Evaluate(samples[i].words, &sstats); + float sentscore = metric.ComputeScore(sstats); if (invert_score) sentscore *= -1.0; if (!cur_good || sentscore > cur_good->mt_metric) cur_good = MakeHypothesisInfo(samples[i].fmap, sentscore); @@ -136,11 +142,13 @@ struct TrainingObserver : public DecoderObserver { } } else { KBest::KBestDerivations<vector<WordID>, ESentenceTraversal> kbest(forest, kbest_size); + SufficientStats sstats; for (int i = 0; i < kbest_size; ++i) { const KBest::KBestDerivations<vector<WordID>, ESentenceTraversal>::Derivation* d = kbest.LazyKthBest(forest.nodes_.size() - 1, i); if (!d) break; - float sentscore = ds[sent_id]->ScoreCandidate(d->yield)->ComputeScore(); + ds[sent_id]->Evaluate(d->yield, &sstats); + float sentscore = metric.ComputeScore(sstats); if (invert_score) sentscore *= -1.0; // cerr << TD::GetString(d->yield) << " ||| " << d->score << " ||| " << sentscore << endl; if (i == 0) @@ -192,15 +200,20 @@ int main(int argc, char** argv) { } vector<string> corpus; ReadTrainingCorpus(conf["source"].as<string>(), &corpus); - const string metric_name = conf["mt_metric"].as<string>(); - ScoreType type = ScoreTypeFromString(metric_name); - if (type == TER) { - invert_score = true; - } else { - invert_score = false; + + string metric_name = UppercaseString(conf["mt_metric"].as<string>()); + if (metric_name == "COMBI") { + cerr << "WARNING: 'combi' metric is no longer supported, switching to 'COMB:TER=-0.5;IBM_BLEU=0.5'\n"; + metric_name = "COMB:TER=-0.5;IBM_BLEU=0.5"; + } else if (metric_name == "BLEU") { + cerr << "WARNING: 'BLEU' is ambiguous, assuming 'IBM_BLEU'\n"; + metric_name = "IBM_BLEU"; } - DocScorer ds(type, conf["reference"].as<vector<string> >(), ""); + EvaluationMetric* metric = EvaluationMetric::Instance(metric_name); + DocumentScorer ds(metric, conf["reference"].as<vector<string> >()); cerr << "Loaded " << ds.size() << " references for scoring with " << metric_name << endl; + invert_score = metric->IsErrorMetric(); + if (ds.size() != corpus.size()) { cerr << "Mismatched number of references (" << ds.size() << ") and sources (" << corpus.size() << ")\n"; return 1; @@ -221,7 +234,7 @@ int main(int argc, char** argv) { assert(corpus.size() > 0); vector<GoodBadOracle> oracles(corpus.size()); - TrainingObserver observer(conf["k_best_size"].as<int>(), ds, sample_forest, &oracles); + TrainingObserver observer(conf["k_best_size"].as<int>(), ds, *metric, sample_forest, &oracles); int cur_sent = 0; int lcount = 0; int normalizer = 0; diff --git a/training/mira/run_mira.pl b/training/mira/run_mira.pl new file mode 100755 index 00000000..d71590ba --- /dev/null +++ b/training/mira/run_mira.pl @@ -0,0 +1,630 @@ +#!/usr/bin/env perl +use strict; +my @ORIG_ARGV=@ARGV; +use Cwd qw(getcwd); +my $SCRIPT_DIR; BEGIN { use Cwd qw/ abs_path /; use File::Basename; $SCRIPT_DIR = dirname(abs_path($0)); +push @INC, $SCRIPT_DIR, "$SCRIPT_DIR/../../environment"; } + +# Skip local config (used for distributing jobs) if we're running in local-only mode +use LocalConfig; +use Getopt::Long; +use IPC::Open2; +use POSIX ":sys_wait_h"; +my $QSUB_CMD = qsub_args(mert_memory()); +my $default_jobs = env_default_jobs(); + +my $srcFile; +my $refFiles; +my $bin_dir = $SCRIPT_DIR; +die "Bin directory $bin_dir missing/inaccessible" unless -d $bin_dir; +my $FAST_SCORE="$bin_dir/../../mteval/fast_score"; +die "Can't execute $FAST_SCORE" unless -x $FAST_SCORE; + +my $iteration = 0.0; +my $max_iterations = 10; +my $metric = "ibm_bleu"; +my $iniFile; +my $weights; +my $initialWeights; +my $jobs = $default_jobs; # number of decode nodes +my $pmem = "1g"; +my $dir; + +my $SCORER = $FAST_SCORE; + +my $UTILS_DIR="$SCRIPT_DIR/../utils"; +require "$UTILS_DIR/libcall.pl"; + +my $parallelize = "$UTILS_DIR/parallelize.pl"; +my $libcall = "$UTILS_DIR/libcall.pl"; +my $sentserver = "$UTILS_DIR/sentserver"; +my $sentclient = "$UTILS_DIR/sentclient"; + +my $run_local = 0; +my $pass_suffix = ''; + +my $cdec ="$bin_dir/kbest_cut_mira"; + +die "Can't find decoder in $cdec" unless -x $cdec; +my $decoder = $cdec; +my $decoderOpt; +my $update_size; +my $approx_score; +my $kbest_size=250; +my $metric_scale=1; +my $optimizer=2; +my $disable_clean = 0; +my $use_make=0; +my $density_prune; +my $cpbin=1; +my $help = 0; +my $epsilon = 0.0001; +my $step_size = 0.01; +my $gpref; +my $unique_kbest; +my $freeze; +my $hopes=1; +my $fears=1; +my $sent_approx=0; +my $pseudo_doc=0; + +my $range = 35000; +my $minimum = 15000; +my $portn = int(rand($range)) + $minimum; + + +# Process command-line options +Getopt::Long::Configure("no_auto_abbrev"); +if (GetOptions( + "decoder=s" => \$decoderOpt, + "jobs=i" => \$jobs, + "density-prune=f" => \$density_prune, + "dont-clean" => \$disable_clean, + "pass-suffix=s" => \$pass_suffix, + "epsilon=s" => \$epsilon, + "help" => \$help, + "local" => \$run_local, + "use-make=i" => \$use_make, + "max-iterations=i" => \$max_iterations, + "pmem=s" => \$pmem, + "cpbin!" => \$cpbin, + "ref-files=s" => \$refFiles, + "metric=s" => \$metric, + "source-file=s" => \$srcFile, + "weights=s" => \$initialWeights, + "optimizer=i" => \$optimizer, + "metric-scale=i" => \$metric_scale, + "kbest-size=i" => \$kbest_size, + "update-size=i" => \$update_size, + "step-size=f" => \$step_size, + "hope-select=i" => \$hopes, + "fear-select=i" => \$fears, + "sent-approx" => \$sent_approx, + "pseudo-doc" => \$pseudo_doc, + "unique-kbest" => \$unique_kbest, + "grammar-prefix=s" => \$gpref, + "freeze" => \$freeze, + "workdir=s" => \$dir, + ) == 0 || @ARGV!=1 || $help) { + print_help(); + exit; +} + +($iniFile) = @ARGV; + + +sub write_config; +sub enseg; +sub print_help; + +my $nodelist; +my $host =check_output("hostname"); chomp $host; +my $bleu; +my $interval_count = 0; +my $logfile; +my $projected_score; + + +#my $refs_comma_sep = get_comma_sep_refs($refFiles); +my $refs_comma_sep = get_comma_sep_refs('r',$refFiles); + +#my $refs_comma_sep_4cdec = get_comma_sep_refs_4cdec($refFiles); + +unless ($dir){ + $dir = "mira"; +} +unless ($dir =~ /^\//){ # convert relative path to absolute path + my $basedir = check_output("pwd"); + chomp $basedir; + $dir = "$basedir/$dir"; +} + +if ($decoderOpt){ $decoder = $decoderOpt; } + +# Initializations and helper functions +srand; + +my @childpids = (); +my @cleanupcmds = (); + +sub cleanup { + print STDERR "Cleanup...\n"; + for my $pid (@childpids){ unchecked_call("kill $pid"); } + for my $cmd (@cleanupcmds){ unchecked_call("$cmd"); } + exit 1; +}; + +# Always call cleanup, no matter how we exit +*CORE::GLOBAL::exit = + sub{ cleanup(); }; +$SIG{INT} = "cleanup"; +$SIG{TERM} = "cleanup"; +$SIG{HUP} = "cleanup"; + + +my $decoderBase = check_output("basename $decoder"); chomp $decoderBase; +my $newIniFile = "$dir/$decoderBase.ini"; +my $inputFileName = "$dir/input"; +my $user = $ENV{"USER"}; + + +# process ini file +-e $iniFile || die "Error: could not open $iniFile for reading\n"; +open(INI, $iniFile); + +use File::Basename qw(basename); +#pass bindir, refs to vars holding bin +sub modbin { + local $_; + my $bindir=shift; + check_call("mkdir -p $bindir"); + -d $bindir || die "couldn't make bindir $bindir"; + for (@_) { + my $src=$$_; + $$_="$bindir/".basename($src); + check_call("cp -p $src $$_"); + } +} +sub dirsize { + opendir ISEMPTY,$_[0]; + return scalar(readdir(ISEMPTY))-1; +} + + + + +if (-e $dir && dirsize($dir)>1 && -e "$dir/weights" ){ # allow preexisting logfile, binaries, but not dist-vest.pl outputs + die "ERROR: working dir $dir already exists\n\n"; +} else { + -e $dir || mkdir $dir; + mkdir "$dir/scripts"; + my $cmdfile="$dir/rerun-mira.sh"; + open CMD,'>',$cmdfile; + print CMD "cd ",&getcwd,"\n"; + my $cline=&cmdline."\n"; + print CMD $cline; + close CMD; + print STDERR $cline; + chmod(0755,$cmdfile); + unless (-e $initialWeights) { + print STDERR "Please specify an initial weights file with --initial-weights\n"; + print_help(); + exit; + } + check_call("cp $initialWeights $dir/weights.0"); + die "Can't find weights.0" unless (-e "$dir/weights.0"); +} +write_config(*STDERR); + +# Generate initial files and values +check_call("cp $iniFile $newIniFile"); +$iniFile = $newIniFile; + +my $newsrc = "$dir/dev.input"; +enseg($srcFile, $newsrc, $gpref); + +$srcFile = $newsrc; +my $devSize = 0; +open F, "<$srcFile" or die "Can't read $srcFile: $!"; +while(<F>) { $devSize++; } +close F; + +my $lastPScore = 0; +my $lastWeightsFile; +my $bestScoreIter=-1; +my $bestScore=-1; +unless ($update_size){$update_size = $kbest_size;} +# main optimization loop +#while (1){ +for (my $opt_iter=0; $opt_iter<$max_iterations; $opt_iter++) { + + print STDERR "\n\nITERATION $opt_iter\n==========\n"; + print STDERR "Using port $portn\n"; + + # iteration-specific files + my $runFile="$dir/run.raw.$opt_iter"; + my $onebestFile="$dir/1best.$opt_iter"; + my $logdir="$dir/logs.$opt_iter"; + my $decoderLog="$logdir/decoder.sentserver.log.$opt_iter"; + my $scorerLog="$logdir/scorer.log.$opt_iter"; + my $weightdir="$dir/weights.pass$opt_iter/"; + check_call("mkdir -p $logdir"); + check_call("mkdir -p $weightdir"); + + #decode + print STDERR "RUNNING DECODER AT "; + print STDERR unchecked_output("date"); +# my $im1 = $opt_iter - 1; + my $weightsFile="$dir/weights.$opt_iter"; + print "ITER $iteration " ; + my $cur_pass = "-p 0$opt_iter"; + my $decoder_cmd = "$decoder -c $iniFile -w $weightsFile $refs_comma_sep -m $metric -s $metric_scale -b $update_size -k $kbest_size -o $optimizer $cur_pass -O $weightdir -D $dir -h $hopes -f $fears -C $step_size"; + if($unique_kbest){ + $decoder_cmd .= " -u"; + } + if($sent_approx){ + $decoder_cmd .= " -a"; + } + if($pseudo_doc){ + $decoder_cmd .= " -e"; + } + if ($density_prune) { + $decoder_cmd .= " --density_prune $density_prune"; + } + my $pcmd; + if ($run_local) { + $pcmd = "cat $srcFile |"; + } elsif ($use_make) { + # TODO: Throw error when jobs is speong with use_make + $pcmd = "cat $srcFile | $parallelize --use-fork -p $pmem -e $logdir -j $use_make --"; + } + else { + $pcmd = "cat $srcFile | $parallelize -p $pmem -e $logdir -j $jobs --baseport $portn --"; + } + my $cmd = "$pcmd $decoder_cmd 2> $decoderLog 1> $runFile"; + print STDERR "COMMAND:\n$cmd\n"; + check_bash_call($cmd); + + my $retries = 0; + my $num_topbest; + while($retries < 6) { + $num_topbest = check_output("wc -l < $runFile"); + print STDERR "NUMBER OF TOP-BEST HYPs: $num_topbest\n"; + if($devSize == $num_topbest) { + last; + } else { + print STDERR "Incorrect number of topbest. Waiting for distributed filesystem and retrying...\n"; + sleep(10); + } + $retries++; + } + die "Dev set contains $devSize sentences, but we don't have topbest for all these! Decoder failure? Check $decoderLog\n" if ($devSize != $num_topbest); + + + #score the output from this iteration + open RUN, "<$runFile" or die "Can't read $runFile: $!"; + open H, ">$runFile.H" or die; + open F, ">$runFile.F" or die; + open B, ">$runFile.B" or die; + while(<RUN>) { + chomp(); + (my $hope,my $best,my $fear) = split(/ \|\|\| /); + print H "$hope \n"; + print B "$best \n"; + print F "$fear \n"; + } + close RUN; + close F; close B; close H; + + my $dec_score = check_output("cat $runFile.B | $SCORER $refs_comma_sep -m $metric"); + my $dec_score_h = check_output("cat $runFile.H | $SCORER $refs_comma_sep -m $metric"); + my $dec_score_f = check_output("cat $runFile.F | $SCORER $refs_comma_sep -m $metric"); + chomp $dec_score; chomp $dec_score_h; chomp $dec_score_f; + print STDERR "DECODER SCORE: $dec_score HOPE: $dec_score_h FEAR: $dec_score_f\n"; + if ($dec_score> $bestScore){ + $bestScoreIter=$opt_iter; + $bestScore=$dec_score; + } + # save space + check_call("gzip -f $runFile"); + check_call("gzip -f $decoderLog"); + my $iter_filler=""; + if($opt_iter < 10) + {$iter_filler="0";} + + my $nextIter = $opt_iter + 1; + my $newWeightsFile = "$dir/weights.$nextIter"; + $lastWeightsFile = "$dir/weights.$opt_iter"; + + average_weights("$weightdir/weights.mira-pass*.*[0-9].gz", $newWeightsFile, $logdir); + system("gzip -f $logdir/kbes*"); + print STDERR "\n==========\n"; + $iteration++; +} +print STDERR "\nBEST ITER: $bestScoreIter :: $bestScore\n\n\n"; + +print STDOUT "$lastWeightsFile\n"; + +sub get_lines { + my $fn = shift @_; + open FL, "<$fn" or die "Couldn't read $fn: $!"; + my $lc = 0; + while(<FL>) { $lc++; } + return $lc; +} + +sub get_comma_sep_refs { + my ($r,$p) = @_; + my $o = check_output("echo $p"); + chomp $o; + my @files = split /\s+/, $o; + return "-$r " . join(" -$r ", @files); +} + + +sub read_weights_file { + my ($file) = @_; + open F, "<$file" or die "Couldn't read $file: $!"; + my @r = (); + my $pm = -1; + while(<F>) { + next if /^#/; + next if /^\s*$/; + chomp; + if (/^(.+)\s+(.+)$/) { + my $m = $1; + my $w = $2; + die "Weights out of order: $m <= $pm" unless $m > $pm; + push @r, $w; + } else { + warn "Unexpected feature name in weight file: $_"; + } + } + close F; + return join ' ', @r; +} + +sub write_config { + my $fh = shift; + my $cleanup = "yes"; + if ($disable_clean) {$cleanup = "no";} + + print $fh "\n"; + print $fh "DECODER: $decoder\n"; + print $fh "INI FILE: $iniFile\n"; + print $fh "WORKING DIR: $dir\n"; + print $fh "SOURCE (DEV): $srcFile\n"; + print $fh "REFS (DEV): $refFiles\n"; + print $fh "EVAL METRIC: $metric\n"; + print $fh "START ITERATION: $iteration\n"; + print $fh "MAX ITERATIONS: $max_iterations\n"; + print $fh "DECODE NODES: $jobs\n"; + print $fh "HEAD NODE: $host\n"; + print $fh "PMEM (DECODING): $pmem\n"; + print $fh "CLEANUP: $cleanup\n"; + print $fh "INITIAL WEIGHTS: $initialWeights\n"; + print $fh "GRAMMAR PREFIX: $gpref\n"; +} + +sub update_weights_file { + my ($neww, $rfn, $rpts) = @_; + my @feats = @$rfn; + my @pts = @$rpts; + my $num_feats = scalar @feats; + my $num_pts = scalar @pts; + die "$num_feats (num_feats) != $num_pts (num_pts)" unless $num_feats == $num_pts; + open G, ">$neww" or die; + for (my $i = 0; $i < $num_feats; $i++) { + my $f = $feats[$i]; + my $lambda = $pts[$i]; + print G "$f $lambda\n"; + } + close G; +} + +sub enseg { + my $src = shift; + my $newsrc = shift; + my $grammarpref = shift; + + open(SRC, $src); + open(NEWSRC, ">$newsrc"); + my $i=0; + while (my $line=<SRC>){ + chomp $line; + if ($line =~ /^\s*<seg/i) { + if($line =~ /id="[0-9]+"/) { + print NEWSRC "$line\n"; + } else { + die "When using segments with pre-generated <seg> tags, you must include a zero-based id attribute"; + } + } + elsif (defined $grammarpref) { + print NEWSRC "<seg id=\"$i\" grammar=\"$grammarpref.$i.gz\">$line</seg>\n";} + else { + print NEWSRC "<seg id=\"$i\">$line</seg>\n"; + } + $i++; + } + close SRC; + close NEWSRC; +} + +sub print_help { + my $executable = check_output("basename $0"); chomp $executable; + print << "Help"; + +Usage: $executable [options] <ini file> + Runs a complete MIRA optimization using the ini file specified. + Example invocation: + run_mira.pl \ + --pmem 3g \ + --max-iterations 20 \ + --optimizer 2 \ + --unique-kbest \ + --jobs 15 \ + --kbest-size 500 \ + --hope-select 1 \ + --fear-select 1 \ + --ref-files "ref.0.soseos ref.1.soseos" \ + --source-file src.soseos \ + --weights weights.init \ + --workdir workdir \ + --grammar-prefix grammars/grammar \ + --step-size 0.01 \ + --metric-scale 10000 \ + +Required: + + --ref-files <files> + Dev set ref files. This option takes only a single string argument. + To use multiple files (including file globbing), this argument should + be quoted. + --source-file <file> + Dev set source file. + --weights <file> + Initial weights file + +General options: + + --help + Print this message and exit. + + --max-iterations <M> + Maximum number of iterations to run. If not specified, defaults + to $max_iterations. + + --metric <method> + Metric to optimize. + Example values: IBM_BLEU, NIST_BLEU, Koehn_BLEU, TER, Combi + + --workdir <dir> + Directory for intermediate and output files. If not specified, the + name is derived from the ini filename. Assuming that the ini + filename begins with the decoder name and ends with ini, the default + name of the working directory is inferred from the middle part of + the filename. E.g. an ini file named decoder.foo.ini would have + a default working directory name foo. + --optimizer <I> + Learning method to use for weight update. Choice are 1) SGD, 2) PA MIRA with Selection from Cutting Plane, 3) Cutting Plane MIRA, 4) PA MIRA,5) nbest MIRA with hope, fear, and model constraints + --metric-scale <I> + Scale MT loss by this amount when computing hope/fear candidates + --kbest-size <I> + Size of k-best list to extract from forest + --update-size <I> + Size of k-best list to use for update (applies to optimizer 5) + --step-size <F> + Controls aggresiveness of update (C) + --hope-select<I> + How to select hope candidate. Choices are 1) model score - cost, 2) min cost + --fear-select <I> + How to select fear candodate. Choices are 1) model score + cost, 2) max cost, 3) max score + --sent-approx + Use smoothed sentence-level MT metric + --pseudo-doc + Use pseudo document to approximate MT metric + --unique-kbest + Extract unique k-best from forest + --grammar-prefix <path> + Path to sentence-specific grammar files + +Job control options: + + --jobs <I> + Number of decoder processes to run in parallel. [default=$default_jobs] + + --pmem <N> + Amount of physical memory requested for parallel decoding jobs + (used with qsub requests only) + + --local + Run single learner + --use-make <I> + Run parallel learners on a single machine through fork. + + +Help +} + + +sub cmdline { + return join ' ',($0,@ORIG_ARGV); +} + +#buggy: last arg gets quoted sometimes? +my $is_shell_special=qr{[ \t\n\\><|&;"'`~*?{}$!()]}; +my $shell_escape_in_quote=qr{[\\"\$`!]}; + +sub escape_shell { + my ($arg)=@_; + return undef unless defined $arg; + if ($arg =~ /$is_shell_special/) { + $arg =~ s/($shell_escape_in_quote)/\\$1/g; + return "\"$arg\""; + } + return $arg; +} + +sub escaped_shell_args { + return map {local $_=$_;chomp;escape_shell($_)} @_; +} + +sub escaped_shell_args_str { + return join ' ',&escaped_shell_args(@_); +} + +sub escaped_cmdline { + return "$0 ".&escaped_shell_args_str(@ORIG_ARGV); +} + +sub average_weights { + + my $path = shift; + my $out = shift; + my $logpath = shift; + print "AVERAGE $path $out\n"; + my %feature_weights= (); + my $total =0; + my $total_mult =0; + sleep(10); + foreach my $file (glob "$path") + { + $file =~ /\/([^\/]+).gz$/; + my $fname = $1; + my $cmd = "gzip -d $file"; + $file =~ s/\.gz//; + check_bash_call($cmd); + my $mult = 0; + print "FILE $file \n"; + open SCORE, "< $file" or next; + $total++; + while( <SCORE> ) { + my $line = $_; + if ($line !~ m/^\#/) + { + my @s = split(" ",$line); + $feature_weights{$s[0]}+= $mult * $s[1]; + } + else + { + (my $msg,my $ran,$mult) = split(/ \|\|\| /); + print "Processing $ran $mult\n"; + } + } + $total_mult += $mult; + + close SCORE; + $cmd = "gzip $file"; check_bash_call($cmd); + } + +#print out new averaged weights + open OUT, "> $out" or next; + for my $f ( keys %feature_weights ) { + print "$f $feature_weights{$f} $total_mult\n"; + my $ave = $feature_weights{$f} / $total_mult; + + print "Printing $f $ave ||| "; + print OUT "$f $ave\n"; + } + +} diff --git a/word-aligner/ortho-norm/ru.pl b/word-aligner/ortho-norm/ru.pl new file mode 100755 index 00000000..34452d06 --- /dev/null +++ b/word-aligner/ortho-norm/ru.pl @@ -0,0 +1,44 @@ +#!/usr/bin/perl -w +use strict; +use utf8; +binmode(STDIN,":utf8"); +binmode(STDOUT,":utf8"); +while(<STDIN>) { + $_ = uc $_; + s/А/a/g; + s/І/i/g; + s/Б/b/g; + s/В/v/g; + s/Г/g/g; + s/Д/d/g; + s/Е/e/g; + s/Ж/zh/g; + s/З/z/g; + s/И/i/g; + s/Й/i/g; + s/К/k/g; + s/Л/l/g; + s/М/m/g; + s/Н/n/g; + s/О/o/g; + s/П/p/g; + s/Р/r/g; + s/С/s/g; + s/Т/t/g; + s/У/u/g; + s/Ф/f/g; + s/Х/kh/g; + s/Ц/c/g; + s/Ч/ch/g; + s/Ш/sh/g; + s/Щ/shch/g; + s/Ъ//g; + s/Ы//g; + s/Ь//g; + s/Э/e/g; + s/Ю/yo/g; + s/Я/ya/g; + $_ = lc $_; + print; +} + |