Secure your code as it's written. Use Snyk Code to scan source code in minutes - no build needed - and fix issues immediately.
tag_list=tag_list,
batchsize=args.batchsize,
constraints=constraints)
else:
doc = [l.strip() for l in fin]
doc = [sentence for sentence in doc if len(sentence) > 0]
tagged_doc = annotate_fun([[word for word in sent.split(' ')] for sent in doc],
tokenize=args.tokenize)
if args.tokenize:
tagged_doc, doc = tagged_doc
res = parser.parse_doc(doc,
probs=probs,
tag_list=tag_list,
batchsize=args.batchsize)
semantic_templates = args.semantic_templates or SEMANTIC_TEMPLATES.get(args.lang)
print_(res, tagged_doc,
format=args.format,
lang=args.lang,
semantic_templates=semantic_templates)
if input_type is None:
sys.stdout.flush()
else:
break
elif format == 'prolog':
if lang == 'en':
print(to_prolog_en(nbest_trees, tagged_doc), end='', file=file)
elif lang == 'ja':
print(to_prolog_ja(nbest_trees, tagged_doc), end='', file=file)
elif format == 'html':
print(to_mathml(nbest_trees), file=file)
elif format == 'jigg_xml_ccg2lambda':
jigg_xml = to_jigg_xml(nbest_trees, tagged_doc)
templates = semantic_templates or SEMANTIC_TEMPLATES.get(lang)
assert templates, f'--semantic-templates must be spcified for language: {lang}'
result_xml_str, _ = ccg2lambda.parse(jigg_xml, str(templates))
print(result_xml_str.decode('utf-8'), file=file)
elif format == 'ccg2lambda':
jigg_xml = to_jigg_xml(nbest_trees, tagged_doc)
templates = semantic_templates or SEMANTIC_TEMPLATES.get(lang)
assert templates, f'--semantic-templates must be spcified for language: {lang}'
_, formulas_list = ccg2lambda.parse(jigg_xml, str(templates))
for i, (parsed, formulas) in enumerate(zip(nbest_trees, formulas_list)):
for (tree, prob), formula in zip(parsed, formulas):
print(f'ID={i} log probability={prob:.4e}\n{formula}', file=file)
elif format == 'conll':
for i, (parsed, tokens) in enumerate(zip(nbest_trees, tagged_doc)):
for tree, prob in parsed:
print(f'# ID={i}\n# log probability={prob:.4e}\n{tree.conll(tokens=tokens)}', file=file)
elif format == 'json':
for i, (parsed, tokens) in enumerate(zip(nbest_trees, tagged_doc), 1):
for tree, prob in parsed:
res = tree.json(tokens=tokens)
res['id'] = i
res['prob'] = prob
print(json.dumps(res), file=file)
if format == 'xml':
print(process_xml(to_xml(nbest_trees, tagged_doc)), file=file)
elif format == 'jigg_xml':
use_symbol = lang == 'ja'
print(process_xml(to_jigg_xml(nbest_trees, tagged_doc, use_symbol=use_symbol)), file=file)
elif format == 'prolog':
if lang == 'en':
print(to_prolog_en(nbest_trees, tagged_doc), end='', file=file)
elif lang == 'ja':
print(to_prolog_ja(nbest_trees, tagged_doc), end='', file=file)
elif format == 'html':
print(to_mathml(nbest_trees), file=file)
elif format == 'jigg_xml_ccg2lambda':
jigg_xml = to_jigg_xml(nbest_trees, tagged_doc)
templates = semantic_templates or SEMANTIC_TEMPLATES.get(lang)
assert templates, f'--semantic-templates must be spcified for language: {lang}'
result_xml_str, _ = ccg2lambda.parse(jigg_xml, str(templates))
print(result_xml_str.decode('utf-8'), file=file)
elif format == 'ccg2lambda':
jigg_xml = to_jigg_xml(nbest_trees, tagged_doc)
templates = semantic_templates or SEMANTIC_TEMPLATES.get(lang)
assert templates, f'--semantic-templates must be spcified for language: {lang}'
_, formulas_list = ccg2lambda.parse(jigg_xml, str(templates))
for i, (parsed, formulas) in enumerate(zip(nbest_trees, formulas_list)):
for (tree, prob), formula in zip(parsed, formulas):
print(f'ID={i} log probability={prob:.4e}\n{formula}', file=file)
elif format == 'conll':
for i, (parsed, tokens) in enumerate(zip(nbest_trees, tagged_doc)):
for tree, prob in parsed:
print(f'# ID={i}\n# log probability={prob:.4e}\n{tree.conll(tokens=tokens)}', file=file)
elif format == 'json':