def knp_tab(sentence):
tmp_list = []
clause_list = []
echo = subprocess.Popen(
['echo', sentence],
stdout=subprocess.PIPE,
)
juman = subprocess.Popen(
['/home/kensuke-mi/bin/juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(
['knp', '-case', '-tab'],
stdin=juman.stdout,
stdout=subprocess.PIPE,
)
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#print line
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list, clause_num, clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list, clause_num, clause)
info_dic, struc_dic = Syori(clause_list, clause_num, clause,
negative_choice)
set_dic = make_case_set(info_dic)
if not negative_choice == "":
set_dic = add_negative(set_dic, negative_choice)
print "--------------------------"
print "About structure information"
print struc_dic
print "--------------------------"
print "About case dictionary information"
print set_dic
print "--------------------------"
make_sentence.sentence_rule(set_dic, struc_dic)
#print set_dic
#print info_dic
return info_dic, struc_dic
def knp_tab(sentence):
tmp_list = []
clause_list = []
echo = subprocess.Popen(['echo',sentence],
stdout=subprocess.PIPE,
)
juman = subprocess.Popen(['juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(['knp','-case','-tab'],
stdin = juman.stdout,
stdout=subprocess.PIPE,
)
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#print line
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list,clause_num,clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list,clause_num,clause)
info_dic,struc_dic = Syori(clause_list,clause_num,clause,negative_choice)
set_dic = make_case_set(info_dic)
if not negative_choice == "":
set_dic = add_negative(set_dic,negative_choice)
print "--------------------------"
print "About structure information"
print struc_dic
print "--------------------------"
print "About case dictionary information"
print set_dic
print "--------------------------"
make_sentence.sentence_rule(set_dic,struc_dic)
#print set_dic
#print info_dic
return info_dic,struc_dic
def knp_tab(sentence):
tmp_list = []
clause_list = []
sentence = conv_encoding(sentence)
sentence = check_kuten(sentence)
#windowsでは以下のsubprocessはコメントオフ
try:
echo = subprocess.Popen(
['echo', sentence],
stdout=subprocess.PIPE,
)
juman = subprocess.Popen(
['juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(
['knp', '-case', '-tab'],
stdin=juman.stdout,
stdout=subprocess.PIPE,
)
except TypeError:
sys.exit(u'KNPの解析エラーが発生しました。解析を終了します。')
"""
#以下、windows用のコード
#windows用のjuman,knpの入力はcp932でないといけない
__xx__
str = sentence.encode('cp932')
juman = subprocess.Popen(["juman"], stdin=subprocess.PIPE, stdout=subprocess.PIPE, shell=True)
knp = subprocess.Popen(["knp", "-tab"], stdin=juman.stdout, stdout=subprocess.PIPE, shell=True)
juman.stdin.write(str)
juman.stdin.close()
juman.stdout.close()
"""
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#if windows, convertion from cp932 to unicode
#if linux, conversion from 'utf-8'
line = unicode(line, 'utf-8')
#check KNP analysis error
error_check = check_knp_error(line)
if error_check == u'error':
print 'error!'
return '', ''
else:
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#--------------------------------------
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list, clause_num, clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list, clause_num, clause)
#文の構文に関する情報。返ってくるのはハッシュマップ
struc_dic = struc_analyze.structure_analyzer(clause_list, clause_num,
clause)
#文の情報を抽出。返ってくるのはリスト
out_list = syori.Syori(clause_list, clause_num, clause, negative_choice,
frag)
#節が複数節なのか？単節なのか？を判断する。返ってくるのは二値。yes or no
clause_check_result = make_clause.clause_check(out_list)
#そもそもif分けする必要はどこにもないので、そのうち修正すること
if clause_check_result == u"yes":
out_list = make_clause.make_clause_set(out_list)
#別に記述せんでもいいが、明文化しておけばわかりやすいじゃん
if clause_check_result == u"no":
out_list = make_clause.make_clause_set(out_list)
if frag == 1:
print u"*" * 40
print u"result of make clause:", out_list
print u"*" * 40
out_list, orig_index_list = parallel.heiretsu(out_list, frag)
if frag == 1:
print u"*" * 40
out_list, orig_index_list = modifier.modi(out_list, orig_index_list, frag)
if frag == 1:
print u"*" * 40
print u"after make modifier list:", out_list
print u"*" * 40
print u"structure information is:", struc_dic
print u"*" * 40
#__xx__
#demo.demo_test(out_list)
p_a_dic = predicate_dic.make_p_a_dic(out_list, orig_index_list, frag)
make_sentence.sentence(struc_dic, p_a_dic, out_list, frag)
#__xx__
#out_list = parallel.c_heiretsu(out_list,frag)
if frag == 1:
print "*" * 40
print u"Is clause multi or not?:", clause_check_result
return out_list, struc_dic
def knp_tab(sentence):
tmp_list = []
clause_list = []
echo = subprocess.Popen(
['echo', sentence],
stdout=subprocess.PIPE,
)
juman = subprocess.Popen(
['juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(
['knp', '-case', '-tab'],
stdin=juman.stdout,
stdout=subprocess.PIPE,
)
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#print line
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#--------------------------------------
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list, clause_num, clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list, clause_num, clause)
#文の構文に関する情報。返ってくるのはハッシュマップ
struc_dic = struc_analyze.structure_analyzer(clause_list, clause_num,
clause)
#文の情報を抽出。返ってくるのはリスト
out_list = syori.Syori(clause_list, clause_num, clause, negative_choice)
#節が複数節なのか？単節なのか？を判断する。返ってくるのは二値。yes or no
clause_check_result = clause_check(out_list)
print "-----------------------------------"
print "Is clause multi or not?:", clause_check_result
'''
set_dic = make_case_set(info_dic)
if not negative_choice == "":
set_dic = add_negative(set_dic,negative_choice)
print "--------------------------"
print "About structure information"
print struc_dic
print "--------------------------"
print "About case dictionary information"
print set_dic
print "--------------------------"
make_sentence.sentence_rule(set_dic,struc_dic)
#print set_dic
#print info_dic
'''
return out_list, struc_dic
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#--------------------------------------
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list,clause_num,clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list,clause_num,clause)
#文の構文に関する情報。返ってくるのはハッシュマップ
struc_dic = struc_analyze.structure_analyzer(clause_list,clause_num,clause)
#文の情報を抽出。返ってくるのはリスト
out_list = syori.Syori(clause_list,clause_num,clause,negative_choice,frag)
#節が複数節なのか？単節なのか？を判断する。返ってくるのは二値。yes or no
clause_check_result = make_clause.clause_check(out_list)
#そもそもif分けする必要はどこにもないので、そのうち修正すること
if clause_check_result == u"yes":
out_list = make_clause.make_clause_set(out_list)
#別に記述せんでもいいが、明文化しておけばわかりやすいじゃん
if clause_check_result == u"no":
out_list = make_clause.make_clause_set(out_list)
if frag == 1:
def knp_tab(sentence):
tmp_list = []
clause_list = []
#windowsでの使用のために以下のsubprocessはコメントオフ
"""
echo = subprocess.Popen(['echo',sentence],
stdout=subprocess.PIPE,
shell=true
)
juman = subprocess.Popen(['juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(['knp','-case','-tab'],
stdin = juman.stdout,
stdout=subprocess.PIPE,
)
"""
#windows用のjuman,knpの入力はcp931でないといけないらしい
# __xx__
str = sentence.encode('cp932')
juman = subprocess.Popen(["juman"], stdin=subprocess.PIPE, stdout=subprocess.PIPE, shell=True)
knp = subprocess.Popen(["knp", "-tab"], stdin=juman.stdout, stdout=subprocess.PIPE, shell=True)
juman.stdin.write(str)
juman.stdin.close()
juman.stdout.close()
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#convertion from cp932 to unicode
line = unicode(line,'cp932')
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#--------------------------------------
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list,clause_num,clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list,clause_num,clause)
#文の構文に関する情報。返ってくるのはハッシュマップ
struc_dic = struc_analyze.structure_analyzer(clause_list,clause_num,clause)
#文の情報を抽出。返ってくるのはリスト
out_list = syori.Syori(clause_list,clause_num,clause,negative_choice,frag)
#節が複数節なのか？単節なのか？を判断する。返ってくるのは二値。yes or no
clause_check_result = make_clause.clause_check(out_list)
#そもそもif分けする必要はどこにもないので、そのうち修正すること
if clause_check_result == u"yes":
out_list = make_clause.make_clause_set(out_list)
#別に記述せんでもいいが、明文化しておけばわかりやすいじゃん
if clause_check_result == u"no":
out_list = make_clause.make_clause_set(out_list)
if frag == 1:
print u"======================================="
print u"result of make clause:",out_list
print u"======================================="
out_list,orig_index_list = parallel.heiretsu(out_list,frag)
if frag == 1:
print u"======================================="
out_list,orig_index_list = modifier.modi(out_list,orig_index_list,frag)
if frag == 1:
print u"======================================="
print u"after make modifier list:",out_list
print u"======================================="
print u"structure information is:",struc_dic
print u"======================================="
print u"demo out is:",demo.demo_test(out_list)
print u"======================================="
#__xx__
print demo.demo_test(out_list)
out_list = parallel.c_heiretsu(out_list,frag)
if frag == 1:
print "-----------------------------------"
print u"Is clause multi or not?:",clause_check_result
'''
set_dic = make_case_set(info_dic)
if not negative_choice == "":
set_dic = add_negative(set_dic,negative_choice)
print "--------------------------"
print "About structure information"
print struc_dic
print "--------------------------"
print "About case dictionary information"
print set_dic
print "--------------------------"
make_sentence.sentence_rule(set_dic,struc_dic)
#print set_dic
#print info_dic
'''
return out_list,struc_dic
def knp_tab(sentence):
tmp_list = []
clause_list = []
#windowsでの使用のために以下のsubprocessはコメントオフ
"""
echo = subprocess.Popen(['echo',sentence],
stdout=subprocess.PIPE,
shell=true
)
juman = subprocess.Popen(['juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(['knp','-case','-tab'],
stdin = juman.stdout,
stdout=subprocess.PIPE,
)
"""
#windows用のjuman,knpの入力はcp931でないといけないらしい
# __xx__
str = sentence.encode('cp932')
juman = subprocess.Popen(["juman"],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE,
shell=True)
knp = subprocess.Popen(["knp", "-tab"],
stdin=juman.stdout,
stdout=subprocess.PIPE,
shell=True)
juman.stdin.write(str)
juman.stdin.close()
juman.stdout.close()
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#convertion from cp932 to unicode
line = unicode(line, 'cp932')
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#--------------------------------------
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list, clause_num, clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list, clause_num, clause)
#文の構文に関する情報。返ってくるのはハッシュマップ
struc_dic = struc_analyze.structure_analyzer(clause_list, clause_num,
clause)
#文の情報を抽出。返ってくるのはリスト
out_list = syori.Syori(clause_list, clause_num, clause, negative_choice,
frag)
#節が複数節なのか？単節なのか？を判断する。返ってくるのは二値。yes or no
clause_check_result = make_clause.clause_check(out_list)
#そもそもif分けする必要はどこにもないので、そのうち修正すること
if clause_check_result == u"yes":
out_list = make_clause.make_clause_set(out_list)
#別に記述せんでもいいが、明文化しておけばわかりやすいじゃん
if clause_check_result == u"no":
out_list = make_clause.make_clause_set(out_list)
if frag == 1:
print u"======================================="
print u"result of make clause:", out_list
print u"======================================="
out_list, orig_index_list = parallel.heiretsu(out_list, frag)
if frag == 1:
print u"======================================="
out_list, orig_index_list = modifier.modi(out_list, orig_index_list, frag)
if frag == 1:
print u"======================================="
print u"after make modifier list:", out_list
print u"======================================="
print u"structure information is:", struc_dic
print u"======================================="
print u"demo out is:", demo.demo_test(out_list)
print u"======================================="
#__xx__
print demo.demo_test(out_list)
out_list = parallel.c_heiretsu(out_list, frag)
if frag == 1:
print "-----------------------------------"
print u"Is clause multi or not?:", clause_check_result
'''
set_dic = make_case_set(info_dic)
if not negative_choice == "":
set_dic = add_negative(set_dic,negative_choice)
print "--------------------------"
print "About structure information"
print struc_dic
print "--------------------------"
print "About case dictionary information"
print set_dic
print "--------------------------"
make_sentence.sentence_rule(set_dic,struc_dic)
#print set_dic
#print info_dic
'''
return out_list, struc_dic
def knp_tab(sentence):
tmp_list = []
clause_list = []
echo = subprocess.Popen(['echo',sentence],
stdout=subprocess.PIPE,
)
juman = subprocess.Popen(['juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(['knp','-case','-tab'],
stdin = juman.stdout,
stdout=subprocess.PIPE,
)
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#print line
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#--------------------------------------
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list,clause_num,clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list,clause_num,clause)
#文の構文に関する情報。返ってくるのはハッシュマップ
struc_dic = struc_analyze.structure_analyzer(clause_list,clause_num,clause)
#文の情報を抽出。返ってくるのはリスト
out_list = syori.Syori(clause_list,clause_num,clause,negative_choice)
#節が複数節なのか？単節なのか？を判断する。返ってくるのは二値。yes or no
clause_check_result = clause_check(out_list)
print "-----------------------------------"
print "Is clause multi or not?:",clause_check_result
'''
set_dic = make_case_set(info_dic)
if not negative_choice == "":
set_dic = add_negative(set_dic,negative_choice)
print "--------------------------"
print "About structure information"
print struc_dic
print "--------------------------"
print "About case dictionary information"
print set_dic
print "--------------------------"
make_sentence.sentence_rule(set_dic,struc_dic)
#print set_dic
#print info_dic
'''
return out_list,struc_dic
def knp_tab(sentence):
tmp_list = []
clause_list = []
sentence = conv_encoding(sentence)
sentence = check_kuten(sentence)
#windowsでは以下のsubprocessはコメントオフ
try:
echo = subprocess.Popen(['echo',sentence],
stdout=subprocess.PIPE,
)
juman = subprocess.Popen(['juman'],
stdin=echo.stdout,
stdout=subprocess.PIPE,
)
knp = subprocess.Popen(['knp','-case','-tab'],
stdin = juman.stdout,
stdout=subprocess.PIPE,
)
except TypeError:
sys.exit(u'KNPの解析エラーが発生しました。解析を終了します。')
"""
#以下、windows用のコード
#windows用のjuman,knpの入力はcp932でないといけない
__xx__
str = sentence.encode('cp932')
juman = subprocess.Popen(["juman"], stdin=subprocess.PIPE, stdout=subprocess.PIPE, shell=True)
knp = subprocess.Popen(["knp", "-tab"], stdin=juman.stdout, stdout=subprocess.PIPE, shell=True)
juman.stdin.write(str)
juman.stdin.close()
juman.stdout.close()
"""
end_of_pipe_tab = knp.stdout
for line in end_of_pipe_tab:
#if windows, convertion from cp932 to unicode
#if linux, conversion from 'utf-8'
line = unicode(line,'utf-8')
#check KNP analysis error
error_check = check_knp_error(line)
if error_check == u'error':
print 'error!'
return '',''
else:
line_split = line.split(" ")
tmp_list.append(line_split[0])
clause_list.append(line)
#--------------------------------------
#ここで各処理関数に情報を投げる
clause_num, clause = clause_count(tmp_list)
#returns 0 if not negation, returns 1 if negation
negative_value = negative.find_negation(clause_list,clause_num,clause)
if negative_value == 0:
negative_choice = ""
else:
negative_choice = negative.negation(clause_list,clause_num,clause)
#文の構文に関する情報。返ってくるのはハッシュマップ
struc_dic = struc_analyze.structure_analyzer(clause_list,clause_num,clause)
#文の情報を抽出。返ってくるのはリスト
out_list = syori.Syori(clause_list,clause_num,clause,negative_choice,frag)
#節が複数節なのか？単節なのか？を判断する。返ってくるのは二値。yes or no
clause_check_result = make_clause.clause_check(out_list)
#そもそもif分けする必要はどこにもないので、そのうち修正すること
if clause_check_result == u"yes":
out_list = make_clause.make_clause_set(out_list)
#別に記述せんでもいいが、明文化しておけばわかりやすいじゃん
if clause_check_result == u"no":
out_list = make_clause.make_clause_set(out_list)
if frag == 1:
print u"*"*40
print u"result of make clause:",out_list
print u"*"*40
out_list,orig_index_list = parallel.heiretsu(out_list,frag)
if frag == 1:
print u"*"*40
out_list,orig_index_list = modifier.modi(out_list,orig_index_list,frag)
if frag == 1:
print u"*"*40
print u"after make modifier list:",out_list
print u"*"*40
print u"structure information is:",struc_dic
print u"*"*40
#__xx__
#demo.demo_test(out_list)
p_a_dic = predicate_dic.make_p_a_dic(out_list,orig_index_list,frag)
make_sentence.sentence(struc_dic,p_a_dic,out_list,frag)
#__xx__
#out_list = parallel.c_heiretsu(out_list,frag)
if frag == 1:
print "*"*40
print u"Is clause multi or not?:",clause_check_result
return out_list,struc_dic
