if args.debug: DEBUG=True ruleLearningLib.DEBUG=True if args.rbpe: ruleLearningLib.RBPE=True ruleLearningLib.AT_LexicalTagsProcessor.initialize(args.tag_groups_file_name,args.tag_sequences_file_name) RuleApplicationHypothesis.set_target_language(args.target_language) RuleApplicationHypothesis.set_apertium_data_dir(args.apertium_data_dir) #this is useless, I think RuleApplicationHypothesis.set_minimum_covered_words(True) ruleList=RuleList() #load alignment templates if args.alignment_templates: if args.alignment_templates.endswith(".gz"): gfile=gzip.open(args.alignment_templates) else: gfile=open(args.alignment_templates) for line in gfile: line=line.strip().decode('utf-8') at=AlignmentTemplate() at.parse(line) ruleList.add(at) gfile.close() #load sentences sentences=list()
if args.debug: DEBUG = True ruleLearningLib.DEBUG = True if args.rbpe: ruleLearningLib.RBPE = True ruleLearningLib.AT_LexicalTagsProcessor.initialize( args.tag_groups_file_name, args.tag_sequences_file_name) RuleApplicationHypothesis.set_apertium_data_dir(args.apertium_data_dir) RuleApplicationHypothesis.set_target_language(args.target_language) print >> sys.stderr, "Loading ATs ..." ruleList = RuleList() #load alignment templates if args.alignment_templates: if args.alignment_templates.lower().endswith(".gz"): gfile = gzip.open(args.alignment_templates) else: gfile = open(args.alignment_templates) for line in gfile: line = line.strip().decode('utf-8') at = AlignmentTemplate() at.parse(line) ruleList.add(at) gfile.close() print >> sys.stderr, "... done" print >> sys.stderr, "Loading sentences ..."
parser.add_argument('--minimum_covered_words', action='store_true') parser.add_argument('--allow_incompatible_rules', action='store_true') parser.add_argument('--rbpe', action='store_true') parser.add_argument('--tt1_beam', action='store_true') args = parser.parse_args(sys.argv[1:]) if args.debug: DEBUG=True ruleLearningLib.DEBUG=True if args.rbpe: ruleLearningLib.RBPE=True ruleLearningLib.AT_LexicalTagsProcessor.initialize(args.tag_groups_file_name,args.tag_sequences_file_name) ruleList=RuleList() RuleApplicationHypothesis.set_target_language(args.target_language) RuleApplicationHypothesis.set_apertium_data_dir(args.apertium_data_dir) RuleApplicationHypothesis.set_minimum_covered_words(args.minimum_covered_words) #load alignment templates if args.alignment_templates.endswith(".gz"): gfile=gzip.open(args.alignment_templates) else: gfile=open(args.alignment_templates) for line in gfile: line=line.strip().decode('utf-8') at=AlignmentTemplate() at.parse(line) ruleList.add(at)
parser.add_argument('--allow_incompatible_rules', action='store_true') parser.add_argument('--rbpe', action='store_true') parser.add_argument('--tt1_beam', action='store_true') args = parser.parse_args(sys.argv[1:]) if args.debug: DEBUG = True ruleLearningLib.DEBUG = True if args.rbpe: ruleLearningLib.RBPE = True ruleLearningLib.AT_LexicalTagsProcessor.initialize( args.tag_groups_file_name, args.tag_sequences_file_name) ruleList = RuleList() RuleApplicationHypothesis.set_target_language(args.target_language) RuleApplicationHypothesis.set_apertium_data_dir(args.apertium_data_dir) RuleApplicationHypothesis.set_minimum_covered_words( args.minimum_covered_words) #load alignment templates if args.alignment_templates.endswith(".gz"): gfile = gzip.open(args.alignment_templates) else: gfile = open(args.alignment_templates) for line in gfile: line = line.strip().decode('utf-8') at = AlignmentTemplate() at.parse(line)