def tag_id_rank_list_by_txt(txt):
txt = txt.lower()
tag_id_list_rank = defaultdict(int)
for word, rank in tf_idf_seg_txt(txt):
#print word
ars = db_tag_bayes.get(word)
if ars:
ar = array('I')
ar.fromstring(ars)
#print len(ar)
#print db_tag_bayes[word]
#print word, ar
for tag_id, bayes in chunkiter(ar, 2):
tag_id_list_rank[tag_id] += (bayes * rank)
result = []
for tag_id, rank in sorted(tag_id_list_rank.iteritems(),
key=itemgetter(1),
reverse=True):
has_tag = False
if tag_id not in ID2NAME:
continue
for i in ID2NAME[tag_id]:
if has_tag:
break
tag_list = list(sp_txt(i))
if tag_list:
for j in tag_list:
#print j, str(j) in txt
if str(j) in txt:
has_tag = True
break
elif i in txt:
has_tag = True
break
if has_tag:
result.append((tag_id, rank))
return result
def tag_id_rank_list_by_txt(txt):
txt = txt.lower()
tag_id_list_rank = defaultdict(int)
for word, rank in tf_idf_seg_txt(txt):
# print word
ars = db_tag_bayes.get(word)
if ars:
ar = array("I")
ar.fromstring(ars)
# print len(ar)
# print db_tag_bayes[word]
# print word, ar
for tag_id, bayes in chunkiter(ar, 2):
tag_id_list_rank[tag_id] += bayes * rank
result = []
for tag_id, rank in sorted(tag_id_list_rank.iteritems(), key=itemgetter(1), reverse=True):
has_tag = False
if tag_id not in ID2NAME:
continue
for i in ID2NAME[tag_id]:
if has_tag:
break
tag_list = list(sp_txt(i))
if tag_list:
for j in tag_list:
# print j, str(j) in txt
if str(j) in txt:
has_tag = True
break
elif i in txt:
has_tag = True
break
if has_tag:
result.append((tag_id, rank))
return result
def loads_id_score(id_score):
r = array('I')
r.fromstring(id_score)
return list(chunkiter(r, 2))
