def queryDocs(query):
"""
TD: Scheme for repeating words in the query
returns: [(score, doc),...]
"""
query = re.sub('[,.!?\-]', ' ', query)
query = toUnicode(query)
query = filterWord(query)
queryWords = query.split()
qtmp = []
intent = []
for w in queryWords:
if w in intentWords:
intent.append(w)
else:
qtmp.append(w)
queryWords = qtmp
res1 = selectAndExpand(queryWords, graph, globalDict)
try:
res2 = selectAndExpand(queryWords, titleGraph, globalDictTitle)
for k, v in res2.items():
res1[k] += v*1.5
except Exception as e:
pass
for k, v in res1.items():
title = getTitle(k)
for inw in intent:
for w in title:
if editDistance(w, inw) >= 0.7:
res1[k] += 0.2
break
res = res1.items()
res.sort(key=lambda x: -x[1])
for i in range(len(res)):
# print res[i][0]
res[i] = (idxToFile[res[i][0]], res[i][1])
return res