def amount_unsure(self):
""" 根据案件概述，初步获得起诉的总金额 """
amount_unsure = None
if self.trial_level == 1:
text = functions.TextProcessor(self.first_basic_text).clean_text
moneys = functions.TextProcessor(text).extract_moneys()
if moneys:
amount_unsure = max(moneys)
return amount_unsure # float, 以万元为单位
def num_of_facts(self):
""" 犯罪事实的数量。根据日期的数量综合判断 """
num_of_facts = None
if self.trial_level == 1:
text = functions.TextProcessor(self.first_fact_text).clean_text
fact_date_ints = []
all_match = settings.pattern_num_of_facts.finditer(text)
for match in all_match:
year_str = match.group(1)
month_str = match.group(2).translate(
str.maketrans({
'春': '3',
'夏': '6',
'秋': '9',
'冬': '12'
}))
fact_date_int = int(year_str) * 100 + int(
month_str) if month_str else int(
year_str) * 100 # 将日期格式化为六位数的int
fact_date_ints.append(fact_date_int)
fact_date_ints = sorted(list(set(fact_date_ints))) # 去重、排序
if len(fact_date_ints) == 1:
num_of_facts = 1
elif len(fact_date_ints) > 1:
num_of_facts = len(fact_date_ints) - 1
return num_of_facts
def is_simple_procedure(self):
""" 是否简易程序 0-否 1-是 默认0 """
is_simple_procedure = 0
text = functions.TextProcessor(self.first_basic_text).clean_text
if '简易程序' in text and '转为普通程序' not in text:
is_simple_procedure = 1
return is_simple_procedure
def job_info(self):
""" 犯罪行为人或犯罪对象的职务信息，包括职务名、单位性质、职务级别。该字段在paper.defendant_info['job']的基础上针对贪污贿赂罪拓展 """
job_info = {'job': None, 'job_type': None, 'job_grade': None}
if self.trial_level == 1:
# 寻找职务名
if self.defendant_info[
'job'] is not None: # 直接引用paper.defendant_info['job']
job_info['job'] = self.defendant_info['job']
else:
text = functions.TextProcessor(self.first_fact_text).clean_text
text = text[:text.find('证据')]
job_match = settings.pattern_job_info['job'].search(text)
if job_match:
job_info['job'] = job_match.group(1)
if job_info['job'] is not None:
# 判断单位性质
for job_type in settings.JOB_TYPES:
if job_info['job_type'] is None:
for job_type_key in settings.JOB_TYPE_DICT[job_type]:
if job_type in (
'X', 'S'): # 对行政机关、事业单位和人民团体特殊处理，只检查最后6个字
_job = job_info['job'][-6:] if len(
job_info['job']) > 6 else job_info['job']
if job_type_key in _job:
job_info['job_type'] = job_type
break
else: # 其他类型的单位普通处理，对全部职务名称检索关键词
if job_type_key in job_info['job']:
job_info['job_type'] = job_type
break
return job_info
def is_plus_investigated(self):
""" 是否有补充侦查 0-否 1-是 默认0 """
is_plus_investigated = 0
if self.trial_level == 1:
text = functions.TextProcessor(self.first_basic_text).clean_text
if '补充侦查' in text:
is_plus_investigated = 1
return is_plus_investigated # int
def amount_sure(self):
""" 根据法院认定情况或已查明的事实，初步获得认定的总金额 """
amount_sure = None
if self.trial_level == 1:
# 首先在法院认定情况中寻找
text = functions.TextProcessor(self.first_opinion_text).clean_text
text = text[:text.find('辩护')] # 截至'辩护'
moneys = functions.TextProcessor(text).extract_moneys()
if moneys:
amount_sure = max(moneys)
# 如果找不到，再在已查明的事实中寻找
else:
text = functions.TextProcessor(self.first_fact_text).clean_text
text = text[:text.find('证据')] # 定位事实部分
moneys = functions.TextProcessor(text).extract_moneys()
if moneys:
amount_sure = max(moneys)
return amount_sure # float, 以万元为单位
def is_delayed(self):
""" 是否延期 0-否 1-是 默认0 """
is_delayed = 0
if self.trial_level == 1:
text = functions.TextProcessor(self.first_basic_text).clean_text
match = settings.pattern_is_delayed.search(text)
if match:
is_delayed = 1
return is_delayed
def prosecute_number(self):
""" 获取起诉书号 """
prosecute_number = None
if self.trial_level == 1:
text = functions.TextProcessor(self.first_basic_text).clean_text
match = settings.pattern_prosecute_number.search(text)
if match:
prosecute_number = match.group(1)
return prosecute_number # str
def is_bad_effect(self):
""" 是否造成恶劣社会影响/国家和人民利益损失 0否1是 默认0 """
is_bad_effect = 0
if self.trial_level == 1:
text = functions.TextProcessor(self.first_opinion_text).clean_text
text = text[:text.find('辩护')]
match = settings.pattern_bad_effect.search(text)
if match:
is_bad_effect = 1
return is_bad_effect
def is_zishou(self):
""" 是否有自首情节 0-否 1-是 默认0 """
is_zishou = 0
if self.trial_level == 1:
text = functions.TextProcessor(self.first_opinion_text).clean_text
# 消除辩护意见
for sentence in self.defensive_opinion_sentences:
text = text.replace(sentence, '')
if '自首' in text:
is_zishou = 1
return is_zishou
def defensive_opinion_sentences(self):
""" 在法院认定意见中，获取含辩护意见的多个句子元组或空元组 """
defensive_opinion_sentences = []
if self.trial_level == 1:
opinion_sentences = functions.TextProcessor(
self.first_opinion_text).sentences
if opinion_sentences:
if '本院认为' in opinion_sentences[0]:
for s in opinion_sentences:
if '辩护' in s:
defensive_opinion_sentences.append(s)
return tuple(defensive_opinion_sentences) # tuple(str, )
def money_usage(self):
""" 赃款的用途 """
money_usage = None
if self.trial_level == 1:
if self.cause in ('贪污罪', '受贿罪', '挪用公款罪'):
text = functions.TextProcessor(
self.first_opinion_text).clean_text
match = settings.pattern_money_usage.search(text)
if match:
money_usage = match.group(1)
return money_usage
def is_suohui(self):
""" 是否有索贿情节 0否1是 默认0 """
is_suohui = None
if self.trial_level == 1:
if self.cause == '受贿罪':
is_suohui = 0
text = functions.TextProcessor(
self.first_opinion_text).clean_text
text = text[:text.find('辩护')]
if '索贿' in text:
is_suohui = 1
return is_suohui
def is_tuizang(self):
""" 是否退赃 0否1是 默认0 """
is_tuizang = None
if self.trial_level == 1:
if self.cause in ('贪污罪', '受贿罪'):
is_tuizang = 0
text = functions.TextProcessor(
self.first_opinion_text).clean_text
text = text[:text.find('辩护')]
if '退' in text: # 退回 退赃 退缴 退清 退出 退交 退还 退赔 退完
is_tuizang = 1
return is_tuizang
def is_seek_promote(self):
""" 是否谋求他人职务调整 0否1是 默认0 """
is_seek_promote = None
if self.trial_level == 1:
if self.cause == '受贿罪':
is_seek_promote = 0
text = functions.TextProcessor(
self.first_opinion_text).clean_text
text = text[:text.find('辩护')]
if '提拔' in text:
is_seek_promote = 1
return is_seek_promote
def prosecutors(self):
""" 获取公诉人姓名列表 """
prosecutors = []
if self.trial_level == 1:
text = functions.TextProcessor(self.first_basic_text).clean_text
match = settings.pattern_prosecutors.search(text)
if match:
prosecutors = list(
map(
lambda a: settings.pattern_prosecutors_delete_strings.
sub('', a),
match.group(1).split('、')))
return prosecutors # list[str, ]
def is_punished_by_party_admin(self):
""" 是否曾因贪污、受贿受过党纪、行政处分 0否1是 默认0 """
is_punished_by_party_admin = None
if self.trial_level == 1:
if self.cause in ('贪污罪', '受贿罪'):
is_punished_by_party_admin = 0
text = functions.TextProcessor(
self.first_opinion_text).clean_text
text = text[:text.find('辩护')]
match = settings.pattern_punished_by_party_admin.search(text)
if match:
is_punished_by_party_admin = 1
return is_punished_by_party_admin
def is_special_money(self):
""" 是否贪污特定款项 0否1是 默认0 """
is_special_money = None
if self.trial_level == 1:
if self.cause in ('贪污罪', '挪用公款罪'):
is_special_money = 0
text = functions.TextProcessor(
self.first_opinion_text).clean_text
text = text[:text.find('辩护')]
match = settings.pattern_special_money.search(text)
if match:
is_special_money = 1
return is_special_money
def is_tanbai(self):
""" 是否有坦白情节 0-否 1-是 默认0 """
""" 包含表述：坦白；认罪；如实供述；交代 """
is_tanbai = 0
if self.is_zishou: # 是自首的一定是坦白
is_tanbai = 1
elif self.trial_level == 1:
text = functions.TextProcessor(self.first_opinion_text).clean_text
# 消除辩护意见
for sentence in self.defensive_opinion_sentences:
text = text.replace(sentence, '')
tanbai_match = settings.pattern_tanbai.search(text)
if tanbai_match:
is_tanbai = 1
return is_tanbai
def is_punished_by_criminal_law(self):
""" 是否曾因故意犯罪受过刑事追究 0否1是 默认0 """
is_punished_by_criminal_law = None
if self.trial_level == 1:
if self.cause in ('贪污罪', '受贿罪'):
if self.is_leifan == 1: # 如果是累犯，该字段值自动为1
is_punished_by_criminal_law = 1
else:
is_punished_by_criminal_law = 0
text = functions.TextProcessor(
self.first_opinion_text).clean_text
text = text[:text.find('辩护')]
match = settings.pattern_punished_by_criminal_law.search(
text)
if match:
is_punished_by_criminal_law = 1
return is_punished_by_criminal_law
def to_html(self, html_path):
""" 输出文书内容到html文件。必须指定文件的绝对路径html_path """
""" 按段落输出，同时输出各段落标记 """
with open(html_path, 'w', encoding='utf-8') as f:
f.write(
settings.html_template_head.replace('{title}',
str(self.paper_id)))
jid, cause, title, case_number, court = functions.ItemDumper( # 格式化输出
self.jid, self.cause, self.title, self.case_number,
self.court).format()
f.write("""<p>{0} {1}</p>\n<p>{2}</p>\n<p>{3}</p>\n<p>{4}</p>\n""".
format(jid, cause, title, case_number, court))
for para in self.all_paragraphs:
f.write("<p>{0}.{1}</p>\n<p>{2}</p>\n".format(
para[0], para[1],
functions.TextProcessor(para[3]).clean_text))
f.write(settings.html_template_tail)
print('to html finished at paper_id: {}'.format(self.paper_id))
return 0
def gongfan(self):
""" 共犯状态 0-不区分主从 1-主犯 2-从犯 默认None """
""" 目前只适用于单人的判决书 """
gongfan = None
if self.trial_level == 1:
text = functions.TextProcessor(self.first_opinion_text).clean_text
# 消除辩护意见
for sentence in self.defensive_opinion_sentences:
text = text.replace(sentence, '')
no_zhucong_match = settings.pattern_gongfan['no_zhucong'].search(
text)
if no_zhucong_match:
gongfan = 0
else:
zhucong_match = settings.pattern_gongfan['zhucong'].search(
text)
if zhucong_match:
if zhucong_match.group(1) == '主':
gongfan = 1
elif zhucong_match.group(1) == '从':
gongfan = 2
return gongfan
def penalty(self):
""" 判决结果 """
penalty = None
if self.trial_level == 1:
text = functions.TextProcessor(
self.first_judge_text).clean_text.split(' ')[0]
penalty = {
'many': None,
'freedom': None,
'property': None,
'right': None,
'delay': None
}
# many 确定罪数
many_strings = settings.pattern_penalty['many'].findall(text)
if many_strings: # 如果提取不到罪数，则认为该句存在问题，放弃继续提取；如有罪数，则改其他项的None为0
penalty = {
'many': len(many_strings),
'freedom': 0,
'property': 0.0,
'right': 0,
'delay': 0
}
text = settings.pattern_penalty['split'].split(text)[
-1] # 定位最终执行语句
# freedom 主刑
for k, v in settings.pattern_penalty['freedom'].items():
freedom_match = v.search(text)
if freedom_match:
if k == 'juyi': # 拘役用负数表示
penalty[
'freedom'] = -functions.TextProcessor.period2num(
freedom_match.group(1))
elif k == 'youqitx': # 有期徒刑用正数表示
penalty[
'freedom'] = functions.TextProcessor.period2num(
freedom_match.group(1))
elif k in ('wuqitx', 'sixing'): # 无期徒刑、死刑直接写入
penalty['freedom'] = freedom_match.group(0)
break
# property 财产刑
if '全部' in text: # 先搜索没收个人全部财产，如有，直接写入字符串
penalty['property'] = '全部'
else:
fajin_match = settings.pattern_penalty['property'][
'fajin'].search(text)
moshou_match = settings.pattern_penalty['property'][
'moshou'].search(text)
fajin_money = functions.TextProcessor(fajin_match.group(
1)).extract_moneys() if fajin_match else None
moshou_money = functions.TextProcessor(
moshou_match.group(
1)).extract_moneys() if moshou_match else None
if fajin_money and moshou_money: # 同时有罚金和没收，合并数额，在前面冠以±号
penalty['property'] = '±{0:.2f}'.format(
fajin_money[0] + moshou_money[0])
elif fajin_money: # 只有罚金，用正值表示
penalty['property'] = fajin_money[0]
elif moshou_money: # 只有没收，用负值表示
penalty['property'] = -moshou_money[0]
# right 资格刑
if '政治权利终身' in text: # 先搜索剥夺政治权利终身，如有，直接写入字符串
penalty['right'] = '终身'
else: # 搜索剥夺政治权利的具体时长
right_match = settings.pattern_penalty['right'].search(
text)
if right_match:
penalty['right'] = functions.TextProcessor.period2num(
right_match.group(1))
# delay 缓刑
delay_match = settings.pattern_penalty['delay'].search(text)
if delay_match:
penalty['delay'] = functions.TextProcessor.period2num(
delay_match.group(1))
# free 检查是否免予处罚、无罪
free_match = settings.pattern_penalty['free'].search(text)
if free_match: # 重置为0
penalty = {
'many': penalty['many'],
'freedom': 0,
'property': 0.0,
'right': 0,
'delay': 0
}
return penalty
def is_designated(self):
""" 是否指定管辖 0-否 1-是 默认0"""
is_designated = 0
if '管辖' in functions.TextProcessor(self.first_basic_text).clean_text:
is_designated = 1
return is_designated
def defendant_info(self):
""" 获取被告人信息字典 """
text = functions.TextProcessor(self.litigant_info_text).clean_text
defendant_info = {
'name': None,
'is_name_covered': None,
'sex': None,
'birth': None,
'age': None,
'tribe': '汉族',
'is_minor': 0,
'educated': None,
'job': None
}
if text:
if self.trial_level == 1:
# 获取更准确的含被告人信息的句子
text_split = text.split(' ')
if len(text_split) < 2:
return defendant_info # 无法正确获得含被告人信息的句子
defendant_text = text_split[1][:text_split[1].find('。')] + '。'
# name, is_name_covered
if self.litigants: # 先引用litigants中的名字
defendant_info['name'] = '+'.join(self.litigants)
else: # 如果没有，再自己查找
name_match = settings.pattern_defendant['name'].search(
defendant_text)
if name_match:
name = name_match.group(1)
defendant_info['name'] = name if len(
name) < 10 else None
if defendant_info['name']:
if '某' in defendant_info['name'] or functions.TextProcessor(
defendant_info['name']).check_exist(
r'[^\u4e00-\u9fff]'):
defendant_info['is_name_covered'] = 1
else:
defendant_info['is_name_covered'] = 0
else:
return defendant_info # 如果找不到姓名，则视为句子有缺陷，不再继续查找其他被告人信息，直接返回默认字典
# sex
if '，男' in defendant_text:
defendant_info['sex'] = 1
elif '，女' in defendant_text:
defendant_info['sex'] = 0
# birth, age
birth_match = settings.pattern_defendant['birth'].search(
defendant_text)
if birth_match:
dates = functions.TextProcessor(
birth_match.group(0)).extract_dates()
if dates:
defendant_info['birth'] = dates[0]
defendant_info[
'age'] = self.judge_date.year - defendant_info[
'birth'].year if self.judge_date else None
if not defendant_info['age']: # 有些判决书直接写了年龄
age_match = settings.pattern_defendant['age'].search(
defendant_text)
if age_match:
defendant_info['age'] = int(age_match.group(1))
# tribe, is_minor
tribe_match = settings.pattern_defendant['tribe'].search(
defendant_text)
if tribe_match:
defendant_info['tribe'] = tribe_match.group(1)
if defendant_info['tribe'] != '汉族':
defendant_info['is_minor'] = 1
# educated 1-小学 2-初中 3-高中、中专 4-大专、专科 5-大学、本科 6-研究生
educated_name = None
for pattern_educated in settings.pattern_defendant['educated']:
educated_match = pattern_educated.search(defendant_text)
if educated_match:
educated_name = educated_match.group(1)
break
if educated_name:
for e_key in settings.EDUCATED_DICT.keys():
if e_key in educated_name:
defendant_info[
'educated'] = settings.EDUCATED_DICT[e_key]
break
# job
job_match = settings.pattern_defendant['job'].search(
defendant_text) # 先在defendant_text中找
if job_match:
defendant_info['job'] = job_match.group(1)
return defendant_info # dict
