def __init__(self, path):
print "Loding"
self.tag2id = WordId()
self.word2id = WordId()
self.path = path
self.parent_tag_finder = ParentTagger()
print "Loading done"
def __init__(self, path):
print "Loding"
self.tag2id = WordId()
self.word2id = WordId()
self.path = path
self.parent_tag_finder = ParentTagger()
print "Loading done"
class TagWord(object):
def __init__(self, path):
print "Loding"
self.tag2id = WordId()
self.word2id = WordId()
self.path = path
self.parent_tag_finder = ParentTagger()
print "Loading done"
def _txt_tag_generator(self):
path = self.path
tag2id = self.tag2id
data_files = glob(join(path, "*.data"))
zhihu_data = [join(path, "zhihu")]
zhihu_data.extend(data_files)
print "Processing..."
g = open(join(path, "topic_dict"))
topic_dict = loads(g.read())
count = 0
for data_src in zhihu_data:
print "Processing...", data_src
with open(data_src) as f:
for line in f:
# if count > 1000:
# break
# count += 1
data = loads(line)
if "tags" in data:
tags = data["tags"]
else:
continue
tags_processed = []
if "zhihu" not in data_src:
for tag in tags:
if tag in topic_dict and tag not in banned_tag_list:
tags_processed.append(tag)
if not tags_processed:
continue
else:
tags = tags_processed
# print tags
# raw_input()
"""
查找上级标签
"""
parent_list = self.parent_tag_finder.get_parent_tag_list_by_list(tags)
tags.extend(parent_list)
id_list = tag2id.id_list_by_word_list(tags)
yield data["txt"], id_list
def txt_tag_generator(self):
word2id = self.word2id
for k, v in self._txt_tag_generator():
words = [i for i in list(seg_txt(str(k).lower())) if not i.isdigit()]
yield word2id.id_list_by_word_list(words), v
def tofile(self):
word_id2tag_id = list(self.txt_tag_generator())
path = DATA_DIR
self.tag2id.tofile(join(path, "tag2id"))
self.word2id.tofile(join(path, "word2id"))
tofile(join(path, "word_id2tag_id"), word_id2tag_id)
class TagWord(object):
def __init__(self, path):
print "Loding"
self.tag2id = WordId()
self.word2id = WordId()
self.path = path
self.parent_tag_finder = ParentTagger()
print "Loading done"
def _txt_tag_generator(self):
path = self.path
tag2id = self.tag2id
data_files = glob(join(path, '*.data'))
zhihu_data = [join(path, 'zhihu')]
zhihu_data.extend(data_files)
print 'Processing...'
g = open(join(path, 'topic_dict'))
topic_dict = loads(g.read())
count = 0
for data_src in zhihu_data:
print 'Processing...', data_src
with open(data_src) as f:
for line in f:
#if count > 1000:
# break
#count += 1
data = loads(line)
if 'tags' in data:
tags = data['tags']
else:
continue
tags_processed = []
if 'zhihu' not in data_src:
for tag in tags:
if tag in topic_dict and tag not in banned_tag_list:
tags_processed.append(tag)
if not tags_processed:
continue
else:
tags = tags_processed
#print tags
#raw_input()
'''
查找上级标签
'''
parent_list = self.parent_tag_finder.get_parent_tag_list_by_list(tags)
tags.extend(parent_list)
id_list = tag2id.id_list_by_word_list(tags)
yield data['txt'], id_list
def txt_tag_generator(self):
word2id = self.word2id
for k, v in self._txt_tag_generator():
words = [i for i in list(seg_txt(str(k).lower())) if not i.isdigit()]
yield word2id.id_list_by_word_list(words) , v
def tofile(self):
word_id2tag_id = list(self.txt_tag_generator())
path = DATA_DIR
self.tag2id.tofile(join(path, 'tag2id'))
self.word2id.tofile(join(path, 'word2id'))
tofile(join(path, 'word_id2tag_id'), word_id2tag_id)
