summaryrefslogtreecommitdiff
path: root/extractor
diff options
context:
space:
mode:
Diffstat (limited to 'extractor')
-rw-r--r--extractor/CMakeLists.txt141
-rw-r--r--extractor/Makefile.am192
-rw-r--r--extractor/extract.cc14
-rw-r--r--extractor/scorer2_test.cc (renamed from extractor/scorer_test.cc)0
4 files changed, 150 insertions, 197 deletions
diff --git a/extractor/CMakeLists.txt b/extractor/CMakeLists.txt
new file mode 100644
index 00000000..93a524cc
--- /dev/null
+++ b/extractor/CMakeLists.txt
@@ -0,0 +1,141 @@
+INCLUDE_DIRECTORIES(${CMAKE_CURRENT_SOURCE_DIR}/../utils)
+INCLUDE_DIRECTORIES(${CMAKE_CURRENT_SOURCE_DIR}/)
+INCLUDE_DIRECTORIES(${CMAKE_CURRENT_SOURCE_DIR}/features)
+
+find_package(OpenMP)
+if (OPENMP_FOUND)
+ set (CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${OpenMP_C_FLAGS}")
+ set (CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${OpenMP_CXX_FLAGS}")
+endif()
+
+find_package(GTest)
+find_package(GMock)
+if(GTEST_FOUND)
+ if(GMOCK_FOUND)
+ #rule_factory_test.cc
+ set(TEST_SRCS alignment_test.cc
+ data_array_test.cc
+ fast_intersector_test.cc
+ grammar_extractor_test.cc
+ matchings_finder_test.cc
+ matchings_sampler_test.cc
+ phrase_location_sampler_test.cc
+ phrase_test.cc
+ precomputation_test.cc
+ rule_extractor_helper_test.cc
+ rule_extractor_test.cc
+ scorer2_test.cc
+ suffix_array_sampler_test.cc
+ suffix_array_test.cc
+ target_phrase_extractor_test.cc
+ translation_table_test.cc
+ vocabulary_test.cc)
+ INCLUDE_DIRECTORIES(${GTEST_INCLUDE_DIRS})
+ INCLUDE_DIRECTORIES(${GMOCK_INCLUDE_DIRS})
+ foreach(testSrc ${TEST_SRCS})
+ #Extract the filename without an extension (NAME_WE)
+ get_filename_component(testName ${testSrc} NAME_WE)
+
+ #Add compile target
+ add_executable(${testName} ${testSrc})
+
+ #link to Boost libraries AND your targets and dependencies
+ target_link_libraries(${testName} extractor ${GMOCK_BOTH_LIBRARIES} ${GTEST_BOTH_LIBRARIES} ${Boost_LIBRARIES} ${ZLIB_LIBRARIES})
+
+ #I like to move testing binaries into a testBin directory
+ set_target_properties(${testName} PROPERTIES
+ RUNTIME_OUTPUT_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
+
+ #Finally add it to test execution -
+ #Notice the WORKING_DIRECTORY and COMMAND
+ add_test(NAME ${testName} COMMAND ${CMAKE_CURRENT_SOURCE_DIR}/${testName}
+ WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
+ endforeach(testSrc)
+ endif(GMOCK_FOUND)
+endif(GTEST_FOUND)
+
+set(sacompile_SRCS sacompile.cc)
+add_executable(sacompile ${sacompile_SRCS})
+target_link_libraries(sacompile extractor ${Boost_LIBRARIES} z)
+
+set(run_extractor_SRCS run_extractor.cc)
+add_executable(run_extractor ${run_extractor_SRCS})
+target_link_libraries(run_extractor extractor ${Boost_LIBRARIES} z)
+
+
+set(extract_SRCS extract.cc)
+add_executable(extract ${extract_SRCS})
+target_link_libraries(extract extractor utils ${Boost_LIBRARIES} z)
+
+
+set(extractor_STAT_SRCS
+ alignment.cc
+ backoff_sampler.cc
+ data_array.cc
+ fast_intersector.cc
+ features/count_source_target.cc
+ features/feature.cc
+ features/is_source_singleton.cc
+ features/is_source_target_singleton.cc
+ features/max_lex_source_given_target.cc
+ features/max_lex_target_given_source.cc
+ features/sample_source_count.cc
+ features/target_given_source_coherent.cc
+ features/count_source_target.h
+ features/feature.h
+ features/is_source_singleton.h
+ features/is_source_target_singleton.h
+ features/max_lex_source_given_target.h
+ features/max_lex_target_given_source.h
+ features/sample_source_count.h
+ features/target_given_source_coherent.h
+ grammar.cc
+ grammar_extractor.cc
+ matchings_finder.cc
+ matchings_sampler.cc
+ matchings_trie.cc
+ phrase.cc
+ phrase_builder.cc
+ phrase_location.cc
+ phrase_location_sampler.cc
+ precomputation.cc
+ rule.cc
+ rule_extractor.cc
+ rule_extractor_helper.cc
+ rule_factory.cc
+ scorer.cc
+ suffix_array.cc
+ suffix_array_sampler.cc
+ target_phrase_extractor.cc
+ time_util.cc
+ translation_table.cc
+ vocabulary.cc
+ alignment.h
+ backoff_sampler.h
+ data_array.h
+ fast_intersector.h
+ grammar.h
+ grammar_extractor.h
+ matchings_finder.h
+ matchings_sampler.h
+ matchings_trie.h
+ phrase.h
+ phrase_builder.h
+ phrase_location.h
+ phrase_location_sampler.h
+ precomputation.h
+ rule.h
+ rule_extractor.h
+ rule_extractor_helper.h
+ rule_factory.h
+ sampler.h
+ scorer.h
+ suffix_array.h
+ suffix_array_sampler.h
+ target_phrase_extractor.h
+ time_util.h
+ translation_table.h
+ vocabulary.h)
+
+add_library(extractor STATIC ${extractor_STAT_SRCS})
+
diff --git a/extractor/Makefile.am b/extractor/Makefile.am
deleted file mode 100644
index a406d9dc..00000000
--- a/extractor/Makefile.am
+++ /dev/null
@@ -1,192 +0,0 @@
-
-bin_PROGRAMS = sacompile run_extractor extract
-
-EXTRA_PROGRAMS = alignment_test \
- data_array_test \
- fast_intersector_test \
- feature_count_source_target_test \
- feature_is_source_singleton_test \
- feature_is_source_target_singleton_test \
- feature_max_lex_source_given_target_test \
- feature_max_lex_target_given_source_test \
- feature_sample_source_count_test \
- feature_target_given_source_coherent_test \
- grammar_extractor_test \
- matchings_finder_test \
- matchings_sampler_test \
- phrase_location_sampler_test \
- phrase_test \
- precomputation_test \
- rule_extractor_helper_test \
- rule_extractor_test \
- rule_factory_test \
- scorer_test \
- suffix_array_sampler_test \
- suffix_array_test \
- target_phrase_extractor_test \
- translation_table_test \
- vocabulary_test
-
-if HAVE_GTEST
- RUNNABLE_TESTS = alignment_test \
- data_array_test \
- fast_intersector_test \
- feature_count_source_target_test \
- feature_is_source_singleton_test \
- feature_is_source_target_singleton_test \
- feature_max_lex_source_given_target_test \
- feature_max_lex_target_given_source_test \
- feature_sample_source_count_test \
- feature_target_given_source_coherent_test \
- grammar_extractor_test \
- matchings_finder_test \
- matchings_sampler_test \
- phrase_location_sampler_test \
- phrase_test \
- precomputation_test \
- rule_extractor_helper_test \
- rule_extractor_test \
- rule_factory_test \
- scorer_test \
- suffix_array_sampler_test \
- suffix_array_test \
- target_phrase_extractor_test \
- translation_table_test \
- vocabulary_test
-endif
-
-noinst_PROGRAMS = $(RUNNABLE_TESTS)
-
-TESTS = $(RUNNABLE_TESTS)
-
-alignment_test_SOURCES = alignment_test.cc
-alignment_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-data_array_test_SOURCES = data_array_test.cc
-data_array_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-fast_intersector_test_SOURCES = fast_intersector_test.cc
-fast_intersector_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-feature_count_source_target_test_SOURCES = features/count_source_target_test.cc
-feature_count_source_target_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-feature_is_source_singleton_test_SOURCES = features/is_source_singleton_test.cc
-feature_is_source_singleton_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-feature_is_source_target_singleton_test_SOURCES = features/is_source_target_singleton_test.cc
-feature_is_source_target_singleton_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-feature_max_lex_source_given_target_test_SOURCES = features/max_lex_source_given_target_test.cc
-feature_max_lex_source_given_target_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-feature_max_lex_target_given_source_test_SOURCES = features/max_lex_target_given_source_test.cc
-feature_max_lex_target_given_source_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-feature_sample_source_count_test_SOURCES = features/sample_source_count_test.cc
-feature_sample_source_count_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-feature_target_given_source_coherent_test_SOURCES = features/target_given_source_coherent_test.cc
-feature_target_given_source_coherent_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-grammar_extractor_test_SOURCES = grammar_extractor_test.cc
-grammar_extractor_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-matchings_finder_test_SOURCES = matchings_finder_test.cc
-matchings_finder_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-matchings_sampler_test_SOURCES = matchings_sampler_test.cc
-matchings_sampler_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-phrase_location_sampler_test_SOURCES = phrase_location_sampler_test.cc
-phrase_location_sampler_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-phrase_test_SOURCES = phrase_test.cc
-phrase_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-precomputation_test_SOURCES = precomputation_test.cc
-precomputation_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-rule_extractor_helper_test_SOURCES = rule_extractor_helper_test.cc
-rule_extractor_helper_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-rule_extractor_test_SOURCES = rule_extractor_test.cc
-rule_extractor_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-rule_factory_test_SOURCES = rule_factory_test.cc
-rule_factory_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-scorer_test_SOURCES = scorer_test.cc
-scorer_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-suffix_array_sampler_test_SOURCES = suffix_array_sampler_test.cc
-suffix_array_sampler_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-suffix_array_test_SOURCES = suffix_array_test.cc
-suffix_array_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-target_phrase_extractor_test_SOURCES = target_phrase_extractor_test.cc
-target_phrase_extractor_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-translation_table_test_SOURCES = translation_table_test.cc
-translation_table_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) $(GMOCK_LDFLAGS) $(GMOCK_LIBS) libextractor.a
-vocabulary_test_SOURCES = vocabulary_test.cc
-vocabulary_test_LDADD = $(GTEST_LDFLAGS) $(GTEST_LIBS) libextractor.a
-
-noinst_LIBRARIES = libextractor.a
-
-sacompile_SOURCES = sacompile.cc
-sacompile_LDADD = libextractor.a
-run_extractor_SOURCES = run_extractor.cc
-run_extractor_LDADD = libextractor.a
-extract_SOURCES = extract.cc
-extract_LDADD = libextractor.a
-
-libextractor_a_SOURCES = \
- alignment.cc \
- backoff_sampler.cc \
- data_array.cc \
- fast_intersector.cc \
- features/count_source_target.cc \
- features/feature.cc \
- features/is_source_singleton.cc \
- features/is_source_target_singleton.cc \
- features/max_lex_source_given_target.cc \
- features/max_lex_target_given_source.cc \
- features/sample_source_count.cc \
- features/target_given_source_coherent.cc \
- features/count_source_target.h \
- features/feature.h \
- features/is_source_singleton.h \
- features/is_source_target_singleton.h \
- features/max_lex_source_given_target.h \
- features/max_lex_target_given_source.h \
- features/sample_source_count.h \
- features/target_given_source_coherent.h \
- grammar.cc \
- grammar_extractor.cc \
- matchings_finder.cc \
- matchings_sampler.cc \
- matchings_trie.cc \
- phrase.cc \
- phrase_builder.cc \
- phrase_location.cc \
- phrase_location_sampler.cc \
- precomputation.cc \
- rule.cc \
- rule_extractor.cc \
- rule_extractor_helper.cc \
- rule_factory.cc \
- scorer.cc \
- suffix_array.cc \
- suffix_array_sampler.cc \
- target_phrase_extractor.cc \
- time_util.cc \
- translation_table.cc \
- vocabulary.cc \
- alignment.h \
- backoff_sampler.h \
- data_array.h \
- fast_intersector.h \
- grammar.h \
- grammar_extractor.h \
- matchings_finder.h \
- matchings_sampler.h \
- matchings_trie.h \
- phrase.h \
- phrase_builder.h \
- phrase_location.h \
- phrase_location_sampler.h \
- precomputation.h \
- rule.h \
- rule_extractor.h \
- rule_extractor_helper.h \
- rule_factory.h \
- sampler.h \
- scorer.h \
- suffix_array.h \
- suffix_array_sampler.h \
- target_phrase_extractor.h \
- time_util.h \
- translation_table.h \
- vocabulary.h
-
-AM_CPPFLAGS = -W -Wall -Wno-sign-compare $(OPENMP_CXXFLAGS) $(GTEST_CPPFLAGS) $(GMOCK_CPPFLAGS)
-AM_LDFLAGS = $(OPENMP_CXXFLAGS)
diff --git a/extractor/extract.cc b/extractor/extract.cc
index e5b6f6ff..08f209cc 100644
--- a/extractor/extract.cc
+++ b/extractor/extract.cc
@@ -14,6 +14,7 @@
const unsigned omp_get_num_threads() { return 1; }
#endif
+#include "filelib.h"
#include "alignment.h"
#include "data_array.h"
#include "features/count_source_target.h"
@@ -42,8 +43,8 @@ using namespace features;
using namespace std;
// Returns the file path in which a given grammar should be written.
-fs::path GetGrammarFilePath(const fs::path& grammar_path, int file_number) {
- string file_name = "grammar." + to_string(file_number);
+fs::path GetGrammarFilePath(const fs::path& grammar_path, int file_number, bool use_zip) {
+ string file_name = "grammar." + to_string(file_number) + (use_zip ? ".gz" : "");
return grammar_path / file_name;
}
@@ -58,6 +59,7 @@ int main(int argc, char** argv) {
("threads,t", po::value<int>()->required()->default_value(1),
threads_option.c_str())
("grammars,g", po::value<string>()->required(), "Grammars output path")
+ ("gzip,z", "Gzip grammars")
("max_rule_span", po::value<int>()->default_value(15),
"Maximum rule span")
("max_rule_symbols", po::value<int>()->default_value(5),
@@ -205,12 +207,14 @@ int main(int argc, char** argv) {
vm["max_rule_symbols"].as<int>(),
vm["max_samples"].as<int>(),
vm["tight_phrases"].as<bool>());
+ const bool use_zip = vm.count("gzip");
// Creates the grammars directory if it doesn't exist.
fs::path grammar_path = vm["grammars"].as<string>();
if (!fs::is_directory(grammar_path)) {
fs::create_directory(grammar_path);
}
+ grammar_path = fs::canonical(grammar_path);
// Reads all sentences for which we extract grammar rules (the paralellization
// is simplified if we read all sentences upfront).
@@ -239,12 +243,12 @@ int main(int argc, char** argv) {
}
Grammar grammar = extractor.GetGrammar(
sentences[i], blacklisted_sentence_ids);
- ofstream output(GetGrammarFilePath(grammar_path, i).c_str());
- output << grammar;
+ WriteFile wf(GetGrammarFilePath(grammar_path, i, use_zip).c_str());
+ *wf.stream() << grammar;
}
for (size_t i = 0; i < sentences.size(); ++i) {
- cout << "<seg grammar=" << GetGrammarFilePath(grammar_path, i) << " id=\""
+ cout << "<seg grammar=" << GetGrammarFilePath(grammar_path, i, use_zip) << " id=\""
<< i << "\"> " << sentences[i] << " </seg> " << suffixes[i] << endl;
}
diff --git a/extractor/scorer_test.cc b/extractor/scorer2_test.cc
index bf77f7ef..bf77f7ef 100644
--- a/extractor/scorer_test.cc
+++ b/extractor/scorer2_test.cc
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-26 05:06:10 +0000