enable programmatic setting of grammar objects, cleanup of grammar handling

4 files changed, 35 insertions, 48 deletions

diff --git a/decoder/decoder.cc b/decoder/decoder.cc
index 333f0fb6..ad4e9e07 100644
--- a/decoder/decoder.cc
+++ b/decoder/decoder.cc
@@ -743,16 +743,14 @@ bool Decoder::Decode(const string& input, DecoderObserver* o) {
 }
 vector<weight_t>& Decoder::CurrentWeightVector() { return pimpl_->CurrentWeightVector(); }
 const vector<weight_t>& Decoder::CurrentWeightVector() const { return pimpl_->CurrentWeightVector(); }
-void Decoder::SetSupplementalGrammar(const std::string& grammar_string) {
-  assert(pimpl_->translator->GetDecoderType() == "SCFG");
-  static_cast<SCFGTranslator&>(*pimpl_->translator).SetSupplementalGrammar(grammar_string);
+void Decoder::AddSupplementalGrammar(GrammarPtr gp) {
+  static_cast<SCFGTranslator&>(*pimpl_->translator).AddSupplementalGrammar(gp);
 }
-void Decoder::SetSentenceGrammarFromString(const std::string& grammar_str) {
+void Decoder::AddSupplementalGrammarFromString(const std::string& grammar_string) {
   assert(pimpl_->translator->GetDecoderType() == "SCFG");
-  static_cast<SCFGTranslator&>(*pimpl_->translator).SetSentenceGrammarFromString(grammar_str);
+  static_cast<SCFGTranslator&>(*pimpl_->translator).AddSupplementalGrammarFromString(grammar_string);
 }
 
-
 bool DecoderImpl::Decode(const string& input, DecoderObserver* o) {
   string buf = input;
   NgramCache::Clear();   // clear ngram cache for remote LM (if used)
diff --git a/decoder/decoder.h b/decoder/decoder.h
index 6b2f7b16..bef2ff5e 100644
--- a/decoder/decoder.h
+++ b/decoder/decoder.h
@@ -37,6 +37,8 @@ struct DecoderObserver {
   virtual void NotifyDecodingComplete(const SentenceMetadata& smeta);
 };
 
+struct Grammar;  // TODO once the decoder interface is cleaned up,
+                 // this should be somewhere else
 struct Decoder {
   Decoder(int argc, char** argv);
   Decoder(std::istream* config_file);
@@ -54,8 +56,8 @@ struct Decoder {
   // add grammar rules (currently only supported by SCFG decoders)
   // that will be used on subsequent calls to Decode. rules should be in standard
   // text format. This function does NOT read from a file.
-  void SetSupplementalGrammar(const std::string& grammar);
-  void SetSentenceGrammarFromString(const std::string& grammar_str);
+  void AddSupplementalGrammar(boost::shared_ptr<Grammar> gp);
+  void AddSupplementalGrammarFromString(const std::string& grammar_string);
  private:
   boost::program_options::variables_map conf;
   boost::shared_ptr<DecoderImpl> pimpl_;
diff --git a/decoder/scfg_translator.cc b/decoder/scfg_translator.cc
index 185f979a..aaa6c40b 100644
--- a/decoder/scfg_translator.cc
+++ b/decoder/scfg_translator.cc
@@ -20,7 +20,6 @@
 #define reverse_foreach BOOST_REVERSE_FOREACH
 
 using namespace std;
-static bool usingSentenceGrammar = false;
 static bool printGrammarsUsed = false;
 
 struct SCFGTranslatorImpl {
@@ -91,31 +90,31 @@ struct SCFGTranslatorImpl {
   bool show_tree_structure_;
   unsigned int ctf_iterations_;
   vector<GrammarPtr> grammars;
-  GrammarPtr sup_grammar_;
+  set<GrammarPtr> sup_grammars_;
 
-  struct Equals { Equals(const GrammarPtr& v) : v_(v) {}
-                  bool operator()(const GrammarPtr& x) const { return x == v_; } const GrammarPtr& v_; };
+  struct ContainedIn {
+    ContainedIn(const set<GrammarPtr>& gs) : gs_(gs) {}
+    bool operator()(const GrammarPtr& x) const { return gs_.find(x) != gs_.end(); }
+    const set<GrammarPtr>& gs_;
+  };
 
-  void SetSupplementalGrammar(const std::string& grammar_string) {
-    grammars.erase(remove_if(grammars.begin(), grammars.end(), Equals(sup_grammar_)), grammars.end());
+  void AddSupplementalGrammarFromString(const std::string& grammar_string) {
+    grammars.erase(remove_if(grammars.begin(), grammars.end(), ContainedIn(sup_grammars_)), grammars.end());
     istringstream in(grammar_string);
-    sup_grammar_.reset(new TextGrammar(&in));
-    grammars.push_back(sup_grammar_);
+    TextGrammar* sent_grammar = new TextGrammar(&in);
+    sent_grammar->SetMaxSpan(max_span_limit);
+    sent_grammar->SetGrammarName("SupFromString");
+    AddSupplementalGrammar(GrammarPtr(sent_grammar));
   }
 
-  struct NameEquals { NameEquals(const string name) : name_(name) {}
-                      bool operator()(const GrammarPtr& x) const { return x->GetGrammarName() == name_; } const string name_; };
+  void AddSupplementalGrammar(GrammarPtr gp) {
+    sup_grammars_.insert(gp);
+    grammars.push_back(gp);
+  }
 
-  void SetSentenceGrammarFromString(const std::string& grammar_str) {
-    assert(grammar_str != "");
-    if (!SILENT) cerr << "Setting sentence grammar" << endl;
-    usingSentenceGrammar = true;
-    istringstream in(grammar_str);
-    TextGrammar* sent_grammar = new TextGrammar(&in);
-    sent_grammar->SetMaxSpan(max_span_limit);
-    sent_grammar->SetGrammarName("__psg");
-    grammars.erase(remove_if(grammars.begin(), grammars.end(), NameEquals("__psg")), grammars.end());
-    grammars.push_back(GrammarPtr(sent_grammar));
+  void RemoveSupplementalGrammars() {
+    grammars.erase(remove_if(grammars.begin(), grammars.end(), ContainedIn(sup_grammars_)), grammars.end());
+    sup_grammars_.clear();
   }
 
   bool Translate(const string& input,
@@ -301,34 +300,22 @@ Check for grammar pointer in the sentence markup, for use with sentence specific
 void SCFGTranslator::ProcessMarkupHintsImpl(const map<string, string>& kv) {
   map<string,string>::const_iterator it = kv.find("grammar");
 
-
-  if (it == kv.end()) {
-    usingSentenceGrammar= false;
-    return;
-  }
-  //Create sentence specific grammar from specified file name and load grammar into list of grammars
-  usingSentenceGrammar = true;
   TextGrammar* sentGrammar = new TextGrammar(it->second);
   sentGrammar->SetMaxSpan(pimpl_->max_span_limit);
   sentGrammar->SetGrammarName(it->second);
-  pimpl_->grammars.push_back(GrammarPtr(sentGrammar));
-
+  pimpl_->AddSupplementalGrammar(GrammarPtr(sentGrammar));
 }
 
-void SCFGTranslator::SetSupplementalGrammar(const std::string& grammar) {
-  pimpl_->SetSupplementalGrammar(grammar);
+void SCFGTranslator::AddSupplementalGrammarFromString(const std::string& grammar) {
+  pimpl_->AddSupplementalGrammarFromString(grammar);
 }
 
-void SCFGTranslator::SetSentenceGrammarFromString(const std::string& grammar_str) {
-  pimpl_->SetSentenceGrammarFromString(grammar_str);
+void SCFGTranslator::AddSupplementalGrammar(GrammarPtr grammar) {
+  pimpl_->AddSupplementalGrammar(grammar);
 }
 
 void SCFGTranslator::SentenceCompleteImpl() {
-
-  if(usingSentenceGrammar)      // Drop the last sentence grammar from the list of grammars
-    {
-      pimpl_->grammars.pop_back();
-    }
+  pimpl_->RemoveSupplementalGrammars();
 }
 
 std::string SCFGTranslator::GetDecoderType() const {
diff --git a/decoder/translator.h b/decoder/translator.h
index cfd3b08a..fc2bb760 100644
--- a/decoder/translator.h
+++ b/decoder/translator.h
@@ -58,8 +58,8 @@ class SCFGTranslatorImpl;
 class SCFGTranslator : public Translator {
  public:
   SCFGTranslator(const boost::program_options::variables_map& conf);
-  void SetSupplementalGrammar(const std::string& grammar);
-  void SetSentenceGrammarFromString(const std::string& grammar);
+  void AddSupplementalGrammar(GrammarPtr gp);
+  void AddSupplementalGrammarFromString(const std::string& grammar);
   virtual std::string GetDecoderType() const;
  protected:
   bool TranslateImpl(const std::string& src,