os.makedirs(output_path)
logging.info("Initializing lexicon.")
lexicon = DictLexicon(output_path)
logging.info("Initializing storage.")
storage = LdbStorage(output_path)
storage.init_db()
storage.open_db()
logging.info("Initializing index.")
index = InvertedIndex(output_path,
field_properties=[
("sentence_id", numpy.int32),
])
index.init_index()
index.open()
logging.info("Initializing sentence stream and its parser.")
sentence_stream = LFSentenceStream(input_path, language=arguments.language)
sentence_parser = LFSentenceParser()
sentence_indexer = LFSentenceIndexer(lexicon)
logging.info("Initializing indexing pipeline.")
indexing_pipeline = IndexingPipeline(lexicon, index, storage)
logging.info("Start indexing file: %s" % input_path)
input_mb_size = float(os.path.getsize(input_path)) / (1024**2)
logging.info("Input size: %.2fMB" % input_mb_size)
indexing_pipeline.index_stream(sentence_stream, sentence_parser,
sentence_indexer)
logging.info("Initializing lexicon.")
lexicon = DictLexicon(output_path)
lexicon.load()
logging.info("Initializing storage.")
storage = LdbStorage(output_path)
storage.init_db()
storage.open_db()
logging.info("Initializing index.")
index = InvertedIndex(output_path, field_properties=[
("document_id", numpy.int32),
])
index.init_index()
index.open()
logging.info("Initializing ruwac stream and its parser.")
if arguments.language == "spa" or arguments.language == "eng":
sentence_stream = GigawordStream(open(input_path, "rb"))
sentence_parser = GigawordParser(language=arguments.language)
elif arguments.language == "rus":
sentence_stream = RuwacStream(open(input_path, "rb"))
sentence_parser = RuwacParser()
else:
raise Exception("Unsupported language: %s" % arguments.language)
sentence_indexer = RuwacIndexer(lexicon)
