def test_should_merge_from_cv_tag_scope(self):
structured_document = SimpleStructuredDocument(lines=[SimpleLine([
SimpleToken(TEXT_1, tag_scope=None, tag=TAG_1)
])])
cv_structured_document = SimpleStructuredDocument(lines=[SimpleLine([
SimpleToken(TEXT_1, tag_scope=CV_TAG_SCOPE, tag=TAG_2)
])])
structured_document = merge_with_cv_structured_document(
structured_document, cv_structured_document,
cv_source_tag_scope=CV_TAG_SCOPE
)
assert get_all_token_tags(structured_document) == [TAG_1]
assert get_all_token_tags(structured_document, scope=CV_TAG_SCOPE) == [TAG_2]
def load_and_convert_to_token_props(filename,
cv_filename,
cv_source_tag_scope,
page_range=None):
try:
structured_document = load_structured_document(filename,
page_range=page_range)
if cv_filename:
cv_structured_document = load_structured_document(
cv_filename, page_range=page_range)
structured_document = merge_with_cv_structured_document(
structured_document,
cv_structured_document,
cv_source_tag_scope=cv_source_tag_scope)
return list(structured_document_to_token_props(structured_document))
except StandardError as e:
raise_from(
RuntimeError('failed to process %s (due to %s: %s)' %
(filename, type(e), e)), e)
def main(argv=None):
args = parse_args(argv)
if args.debug:
logging.getLogger().setLevel('DEBUG')
structured_document = load_lxml_structured_document(args.lxml_path)
if args.cv_lxml_path:
cv_structured_document = load_lxml_structured_document(
args.cv_lxml_path)
structured_document = merge_with_cv_structured_document(
structured_document, cv_structured_document)
model = load_crf_model(args.crf_model)
predict_and_annotate_structured_document(structured_document,
model,
tag_scope=args.tag_scope)
get_logger().info('writing result to: %s', args.output_path)
save_structured_document(args.output_path, structured_document)