def get_parsed_trees_from_string(tree_strings):
# tree_strings separated by "\n"
parsed_trees = []
for line in tree_strings:
line = line.strip()
#print line
if line != '':
parsed_trees.append(
LexicalizedTree.fromstring(line,
leaf_pattern='(?<=\\s)[^\)\(]+'))
return parsed_trees
def process_single_sentence(self, doc, raw_text):
sentence = Sentence(len(doc.sentences), raw_text, doc)
parse_tree_str, deps_str = self.parse_single_sentence(raw_text)
parse = LexicalizedTree.fromstring(parse_tree_str,
leaf_pattern='(?<=\\s)[^\)\(]+')
sentence.set_unlexicalized_tree(parse)
for (token_id, te) in enumerate(parse.leaves()):
word = te
token = Token(word, token_id + 1, sentence)
sentence.add_token(token)
heads = self.get_heads(sentence, deps_str.split('\n'))
sentence.heads = heads
sentence.set_lexicalized_tree(
prep_utils.create_lexicalized_tree(parse, heads))
doc.add_sentence(sentence)
def process_single_sentence(self, doc, raw_text, end_of_para):
sentence = Sentence(len(doc.sentences),
raw_text + ('<s>' if not end_of_para else '<P>'),
doc)
parse_tree_str, deps_str = self.parse_single_sentence(raw_text)
parse_tree_str = parse_tree_str.decode('utf-8').encode(
'ascii', 'ignore')
parse = LexicalizedTree.fromstring(parse_tree_str,
leaf_pattern='(?<=\\s)[^\)\(]+')
sentence.set_unlexicalized_tree(parse)
for (token_id, te) in enumerate(parse.leaves()):
word = te
token = Token(word, token_id + 1, sentence)
sentence.add_token(token)
heads = self.get_heads(sentence, deps_str.split('\n'))
sentence.heads = heads
sentence.set_lexicalized_tree(
prep_utils.create_lexicalized_tree(parse, heads))
doc.add_sentence(sentence)
def process_single_sentence(self, doc, raw_text, end_of_para):
sentence = Sentence(len(doc.sentences),
raw_text + (b'<s>' if not end_of_para else b'<P>'),
doc)
parse_tree_str, deps_str = self.parse_single_sentence(raw_text)
# self.parse_single_sentence(raw_text) returns different result from
# self.syntax_parser.parse_sentence(raw_text)
parse = LexicalizedTree.fromstring(parse_tree_str,
leaf_pattern='(?<=\\s)[^\)\(]+')
sentence.set_unlexicalized_tree(parse)
for (token_id, te) in enumerate(parse.leaves()):
word = te
token = Token(word, token_id + 1, sentence)
sentence.add_token(token)
heads = self.get_heads(sentence, deps_str.split('\n'))
sentence.heads = heads
sentence.set_lexicalized_tree(
prep_utils.create_lexicalized_tree(parse, heads))
doc.add_sentence(sentence)
