def get_sum_list(paper_part):
"""
get sum_list_p from paper_argu, or get sum_list_c from paper_opinion + paper_fact
:return: <list> sum_list_p or sum_list_c
"""
sum_list = []
sum_list_gongji, sum_list_not_gongji = [], []
sum_matchs = constant.sum_pattern.finditer(paper_part)
for match in sum_matchs:
# 转化为浮点数
sum_float = tools.convert_to_arabic(
tools.reserve_only_in(match.group(1), constant.SUM_CHARS))
# 向前看10字符
look_ahead_context = paper_part[match.start(1) - 10:match.start(1)]
# 寻找'共计'
if constant.sum_total_pattern.search(look_ahead_context):
# 总结句中的'共计'
if tools.look_ahead_back(paper_part,
match.span(0),
ahead_dict={
'num': 20,
'char': '收受',
'request_have': True
},
back_dict={
'num': 50,
'char': '受贿罪',
'request_have': True
}):
sum_list.append(sum_float)
return sum_list # 退出函数
# 非总结句中的'共计'
else:
if sum_float not in sum_list_gongji and sum_float < 1000000000: # 大于10亿直接排除
sum_list_gongji.append(sum_float)
# 非'共计'
else:
if sum_float not in sum_list_not_gongji and sum_float < 1000000000: # 大于10亿直接排除
sum_list_not_gongji.append(sum_float)
# 取总和较小、非空的list
if len(sum_list_gongji) > 0 and len(sum_list_not_gongji) > 0:
sum_list = sum_list_gongji if sum(sum_list_gongji) < sum(
sum_list_not_gongji) else sum_list_not_gongji
else:
sum_list = sum_list_gongji if len(
sum_list_gongji) > 0 else sum_list_not_gongji
return sum_list
def decide_zhufan(opinion):
"""
decide zhufan
:param opinion: <str> paper_opinion
:return: <bool> True or False
"""
decide_zhufan_tag = False
zhufan_match = constant.zhufan_pattern.search(opinion)
if zhufan_match:
if tools.look_ahead_back(
opinion,
zhufan_match.span(0),
ahead_dict={'num': 10, 'char': '不', 'request_have': False},
back_dict={'num': 10, 'char': '意见', 'request_have': False}
):
decide_zhufan_tag = True
return decide_zhufan_tag
def get_tag_ligong(opinion):
"""
get tag_ligong from paper_opinion
:param opinion: <str> paper_opinion
:return: <int> 0 for no, 1 for yes
"""
tag_ligong = 0
ligong_match = constant.ligong_pattern.search(opinion)
if ligong_match:
if tools.look_ahead_back(
opinion,
ligong_match.span(0),
ahead_dict={'num': 10, 'char': '不', 'request_have': False},
back_dict={'num': 25, 'char': '不', 'request_have': False}
) and tools.look_back(
opinion,
ligong_match.span(0)[1],
back_dict={'num': 25, 'char': '误', 'request_have': False}
):
tag_ligong = 1
return tag_ligong