def extractOneSent(sent):
"""
extract one sent, find out the
|pid|time|company|department|position|
if they exist in this sentence
"""
global g_retTupleList
#clean the cache
g_retTupleList = []
#combine wors with the same tag into a word
sent = combineSameTag(sent)
#combine the sent into string devided by space
sentStr = " ".join([ x[1] for x in sent ])
#debug
print sentStr
##print sentStr
#sentStr = " ".join([ "%s_%s" % (x[1], x[0]) for x in sent ])
#print sentStr
ret = P.parse(sentStr.encode('utf-8'))
t = nltk.Tree.parse(ret[1])
#t.draw()
# traverse the tree and the sent which will modify the
# global variable g_retTupleList
traverse(t, sent)
# combine the g_retTupleList so that we have the real tuples without \
# redundants
retTuples = []
#debug
for each in g_retTupleList:
#print 'each' , each
#print 'combined', combineTuple(each)
retTuples += combineTuple(each)
#print 'ret', retTuples
return retTuples
def extractOneSent(sent):
"""
extract one sent, find out the
|pid|time|company|department|position|
if they exist in this sentence
"""
global g_retTupleList
#clean the cache
g_retTupleList = []
#combine wors with the same tag into a word
sent = combineSameTag(sent)
#combine the sent into string devided by space
sentStr = " ".join([x[1] for x in sent])
#debug
print sentStr
##print sentStr
#sentStr = " ".join([ "%s_%s" % (x[1], x[0]) for x in sent ])
#print sentStr
ret = P.parse(sentStr.encode('utf-8'))
t = nltk.Tree.parse(ret[1])
#t.draw()
# traverse the tree and the sent which will modify the
# global variable g_retTupleList
traverse(t, sent)
# combine the g_retTupleList so that we have the real tuples without \
# redundants
retTuples = []
#debug
for each in g_retTupleList:
#print 'each' , each
#print 'combined', combineTuple(each)
retTuples += combineTuple(each)
#print 'ret', retTuples
return retTuples
# by zhangzhi @2013-11-08 19:16:28
# Copyright 2013 NONE rights reserved.
import stanfordParserPipeCn as P
import nltk
def traverse(t):
try:
t.node
except AttributeError:
"one leaf is here"
print "Error"
print t,
else:
# Now we know that t.node is defined
print '(', t.node,
print "========="
for child in t:
print "child"
traverse(child)
print ')',
if __name__ == '__main__':
#ret = P.parse("张智 于 2010 年 于 哈尔滨 工业 大学 毕业 ， 毕业 后 在 百度 工作 至 2013 年 。")
ret = P.parse("The dog find the god")
#print ret[1]
t = nltk.Tree.parse(ret[1])
t.draw()
traverse(t)
# Copyright 2013 NONE rights reserved.
import stanfordParserPipeCn as P
import nltk
def traverse(t):
try:
t.node
except AttributeError:
"one leaf is here"
print "Error"
print t,
else:
# Now we know that t.node is defined
print '(', t.node,
print "========="
for child in t:
print "child"
traverse(child)
print ')',
if __name__ == '__main__':
#ret = P.parse("张智 于 2010 年 于 哈尔滨 工业 大学 毕业 ， 毕业 后 在 百度 工作 至 2013 年 。")
ret = P.parse("The dog find the god")
#print ret[1]
t = nltk.Tree.parse(ret[1])
t.draw()
traverse(t)