# -*- coding: utf-8 -*-
#python knock47.py < neko.txt2.cabocha
from knock41 import F41
from collections import OrderedDict
if __name__ == "__main__":
for i, chunks in enumerate(F41()):
for chunk in chunks:
meisi_flag = False
for morph in chunk.morphs:
if meisi_flag:
if morph.surface != u"を":
meisi_flag = False
else:
zyosi_list = list()
kou_list = list()
if chunks[chunk.dst].morphs[0].pos == u"動詞":
dousi = chunks[chunk.dst].morphs[0].base
for src in chunks[chunk.dst].srcs:
kou = ""
zyosi = ""
for morph2 in chunks[src].morphs:
if morph2.pos == u"助詞":
zyosi = morph2.base
kou = chunks[src].chunk2str()
if len(zyosi) != "":
zyosi_list.append(zyosi)
kou_list.append(kou)
zyosi_kou_dic = dict(zip(zyosi_list, kou_list))
#sortした後、辞書の形を保持するOrderedDict
# -*- coding: utf-8 -*-
#python knock45.py < neko.txt2.cabocha> output.txt
from knock41 import F41
if __name__ == "__main__":
for chunks in F41():
for chunk in chunks:
for morph in chunk.morphs:
if morph.pos == u"動詞":
dousi = morph.base
zyosi_list = list()
for src in chunk.srcs:
zyosi = ""
for morph in chunks[src].morphs:
if morph.pos == u"助詞":
zyosi = morph.base
if zyosi != "":
zyosi_list.append(zyosi)
zyosi_list.sort()
#"が"のほうが"と"より先に来る
zyosi = " ".join(zyosi_list)
if len(zyosi) != 0:
print "{}\t{}".format(dousi,zyosi)
#最左の動詞という条件があるので
#これ以上動詞を探すのをやめる
break
"""
cat output.txt | sort | uniq -c | sort -nr | head -20
704 云う と
452 する を
# -*- coding: utf-8 -*-
#cabocha -f1 < neko.txt > neko.txt2.cabocha
#python knock41.py < neko.txt2.cabocha
from knock41 import F41
import sys
if __name__ == "__main__":
sentenses = F41()
for sentense in sentenses:
for chunk in sentense:
for morph in chunk.morphs:
print morph.surface.rstrip("。"),
print "\t",
if chunk.dst > 0:
for morph in sentense[chunk.dst].morphs:
print morph.surface.rstrip("。"),
print "\n",
"""
0, 一
0,
0, 猫 で ある
1, 吾輩 は 猫 で ある
2, 猫 で ある
0, 名前 は 無い
1, まだ 無い
2, 無い
"""