Question

我想通过运行find_pattern.py来提取解析树，这些是与以下模式的匹配：

SIMPLE_PREDICATE = (ROOT, ((SENTENCE, (NP, VP, PERIOD)),))          
APPOSITION = (SENTENCE, ((NP, (NP, COMMA, NP, COMMA)), VP, PERIOD))

但是输出显示：

Traceback (most recent call last):
 File "C:\Users\Anon\Desktop\ZAHID\Working Model 1\zahid\Practice\New folder\find_pattern.py", line 40, in <module>
appos = sent_extract.find_appositions(parse_trees)
 File "C:\Users\Anon\Desktop\ZAHID\Working Model 1\zahid\Practice\New folder\sent_extract.py", line 30, in find_appositions
s in search_for_matches(parse_tree, APPOSITION)]
 File "C:\Users\Anon\Desktop\ZAHID\Working Model 1\zahid\Practice\New folder\sent_extract.py", line 58, in search_for_matches
if is_match(parse_tree, pattern):
 File "C:\Users\Anon\Desktop\ZAHID\Working Model 1\zahid\Practice\New folder\sent_extract.py", line 45, in is_match
if tree.label() == parent and len(tree) == len(children): 
AttributeError: 'listiterator' object has no attribute 'label'

find_pattern.py

import os
import parse_article
import stanford_parser
import sent_extract

article_filename = 'C:\Users\Anon\Desktop\ZAHID\Working Model 1\zahid\Practice\New folder\____.html'

sentences = parse_article.parse_html(article_filename)

user='' 
parser = stanford_parser.create_parser(user)
parse_trees = parser.raw_parse_sents(sentences)

appos = sent_extract.find_appositions(parse_trees)
print(appos)

sent_extract.py

from nltk.tree import Tree
from tags import *

SIMPLE_PREDICATE = (ROOT, ((SENTENCE, (NP, VP, PERIOD)),))          
APPOSITION = (SENTENCE, ((NP, (NP, COMMA, NP, COMMA)), VP, PERIOD)) 

def find_predicates(parse_trees):
  preds = []
  for parse_tree in parse_trees:
    if is_match(parse_tree, SIMPLE_PREDICATE):         
      preds.append(parse_tree[0])
  return preds

def find_appositions(parse_trees):
  appos = []
  for parse_tree in parse_trees:
    appos += [(s[0,0], s[0,2]) for
        s in search_for_matches(parse_tree, APPOSITION)]
  return appos


def is_match(tree, pattern):
  if not isinstance(pattern, tuple):
    return tree.label() == pattern

  else:
    parent = pattern[0]
    children = pattern[1]
    if tree.label() == parent and len(tree) == len(children): 
      for i in xrange(len(tree)):
        ith_child = tree[i]
        if not is_match(ith_child, children[i]):
          return False
      return True

def search_for_matches(parse_tree, pattern):
  matches = []
  if is_match(parse_tree, pattern):
    matches.append(parse_tree)
  for child in parse_tree:
    if isinstance(child, Tree):
      matches += search_for_matches(child, pattern)
  return matches

tags.py

ADJP = 'ADJP'
ADVP = 'ADVP'
NUMBER = 'CD'
DET = 'DT'
PREP = 'IN'
ADJ = 'JJ'
ADJ_COMP = 'JJR'
ADJ_SUP = 'JJS'
MODAL = 'MD'
NOUN = 'NN'
NOUN_PROPER = 'NNP'
NOUN_PL = 'NNS'
NP = 'NP'
POSS = 'POS'
PP = 'PP'
PRONOUN = 'PRP'
PRONOUN_POSS = 'PRP$'
ADVERB = 'RB'
ROOT = 'ROOT'
SENTENCE = 'S'
SBAR = 'SBAR'
WH_QUESTION = 'SBARQ'
BIN_QUESTION = 'SQ'
TO = 'TO'
VERB_INF = 'VB'
VERB_PAST = 'VBD'
VERB_PLURAL = 'VBP'
VERB_3SG = 'VBZ'
VP = 'VP'
WHNP = 'WHNP'
WHADJP = 'WHADJP'
WHADVP = 'WHADVP'
WDT = 'WDT'
WP_POSS = 'WP$'
COMMA = ','
PERIOD = '.'

如何解决“'listiterator'对象没有属性'label'”

0 个答案: