| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133 |
- #!/usr/bin/env/python
- # -*- coding:utf-8 -*-
- import re
- def get_con(subcon, item_no_type, item_no, **kwargs):
- """
- # ------在下一题【解析】在本题【答案】之间找到下一题【stem】的位置--------
- # 主要统计题号的位置，空行信息作为辅助
- :return:
- """
- errmsg_dict = {}
- # item_no = []
- type_info = kwargs['all_type'][kwargs['num']] if 'all_type' in kwargs else ""
- index = kwargs['index']
- errmsg_bef = type_info+"第{0}道题(在整篇文档中为第{1}题)的题文和上一题的解析之间" if type_info \
- else "整篇文档中第{0}题的题文和上一题的解析之间{1}"
- count = 1 # 题量个数
- for id in range(4, len(subcon), 4):
- if id < len(subcon) - 1: # 最后一个不用管
- count += 1 # 当前在本大题中是第几个
- ssub = subcon[id].strip().split("\n") # 首尾空行先去掉
- blank_line = [i for i, v in enumerate(ssub) if v.strip() == ""] # 空格的索引
- # 索引to题号字典，获取可能的题号的位置
- pattern1 = r"([1-9]|[1-4][0-9])\s*[.．、､]" if item_no_type==1 else r"\(([1-9]|[1-4][0-9])\)\s*[.．、､]?"
- line_topicno_dict = {i: re.match(pattern1, v.strip()).group(1)
- for i, v in enumerate(ssub)
- if re.match(pattern1, v.strip())}
- # print("line_topicno_dict",line_topicno_dict)
- con_id_line = list(line_topicno_dict.keys()) # 题号的行索引,第几行
- topicno = list(line_topicno_dict.values()) # 题号序列
- topicno_line_dict = dict(zip(topicno, con_id_line)) # 题号to行索引字典
- if len(con_id_line) != len(topicno_line_dict): # 肯定len(con_id_line)>1
- # 相同序号不是题目序号时， 出现相同可疑题号
- from collections import Counter
- topicno_set = [i for i, j in dict(Counter(topicno)).items() if j == 1]
- if len(topicno_set) == 1 and 0 <= int(topicno_set[0]) - (item_no[-1] + 1) <= 1:
- ssub.insert(topicno_line_dict[topicno_set[0]], "【stem】")
- item_no.append(int(topicno_set[0]))
- else:
- item_no.append(item_no[-1] + 1)
- errmsg = "【多个相同的题目序号或题目序号有误,可能导致拆分题目错误】"
- errmsg = errmsg_bef.format(str(count), str(index + count)) + errmsg if type_info \
- else errmsg_bef.format(str(index + count), "") + errmsg
- errmsg_dict[id / 4 - 1] = errmsg if (id / 4 - 1) not in errmsg_dict \
- else errmsg_dict[id / 4 - 1] + ";" + errmsg
- errmsg_dict[id / 4] = errmsg if id / 4 not in errmsg_dict \
- else errmsg_dict[id / 4] + ";" + errmsg
- # 应取中间偏下方的序号为准！！！
- if len(blank_line) == 1: # 大概率空格处
- ssub[blank_line[0]] = "【stem】"
- else:
- intervals = [(len("".join(ssub[0:r])), len("".join(ssub[r:]))) for r in
- con_id_line]
- intervals = [k for k, r in enumerate(intervals) if
- r[0] > 100 and 50 <= r[1] <= 200]
- if len(intervals) == 1:
- print("【多个相同的题目序号】切分不严谨")
- ssub.insert(con_id_line[intervals[0]], "【stem】")
- else: # 多个可疑题号分不出来
- ssub.insert(0, "【stem】") # 分不出来，先替换
- ssub.insert(0, "")
- else:
- if len(con_id_line) == 1: # 一个题号
- if len(blank_line) == 1 and con_id_line[0] - blank_line[0] == 1: # 空格在前，题号在后
- ssub.insert(con_id_line[0], "【stem】")
- item_no.append(int(topicno[0]))
- else:
- if 0 <= int(topicno[0]) - (item_no[-1] + 1) <= 1: # 允许题号相差1个
- ssub.insert(con_id_line[0], "【stem】")
- item_no.append(int(topicno[0]))
- else:
- if len(blank_line) == 1:
- ssub[blank_line[0]] = "【stem】" # 该题序号不对时再考虑空行
- item_no.append(item_no[-1] + 1)
- else:
- ssub.insert(con_id_line[0], "【stem】")
- item_no.append(item_no[-1] + 1)
- errmsg = "【题目序号不连续,可能导致拆分题目错误】"
- errmsg = errmsg_bef.format(str(count), str(index + count)) + errmsg if type_info \
- else errmsg_bef.format(str(index + count), "") + errmsg
- errmsg_dict[id / 4 - 1] = errmsg if (id / 4 - 1) not in errmsg_dict \
- else errmsg_dict[id / 4 - 1] + ";" + errmsg
- errmsg_dict[id / 4] = errmsg if id / 4 not in errmsg_dict \
- else errmsg_dict[id / 4] + ";" + errmsg
- elif len(con_id_line) > 1: # 多个题号时
- if str(item_no[-1] + 1) in topicno:
- item_no.append(item_no[-1] + 1)
- ssub.insert(topicno_line_dict[str(item_no[-1])], "【stem】")
- elif str(item_no[-1] + 2) in topicno:
- item_no.append(item_no[-1] + 2)
- ssub.insert(topicno_line_dict[str(item_no[-1])], "【stem】")
- else:
- item_no.append(item_no[-1] + 1)
- ssub.insert(0, "【stem】") # 分不出来，先替换
- ssub.insert(0, "")
- errmsg = "【题目序号不连续,可能导致拆分题目错误】"
- errmsg = errmsg_bef.format(str(count), str(index + count)) + errmsg if type_info \
- else errmsg_bef.format(str(index + count), "") + errmsg
- errmsg_dict[id / 4 - 1] = errmsg if (id / 4 - 1) not in errmsg_dict \
- else errmsg_dict[id / 4 - 1] + ";" + errmsg
- errmsg_dict[id / 4] = errmsg if id / 4 not in errmsg_dict \
- else errmsg_dict[id / 4] + ";" + errmsg
- else: # 无题号
- item_no.append(item_no[-1] + 1)
- errmsg = "【没有题目序号,可能导致拆分题目错误】"
- errmsg = errmsg_bef.format(str(count), str(index + count)) + errmsg if type_info \
- else errmsg_bef.format(str(index + count), "") + errmsg
- errmsg_dict[id / 4 - 1] = errmsg if (id / 4 - 1) not in errmsg_dict \
- else errmsg_dict[id / 4 - 1] + ";" + errmsg
- errmsg_dict[id / 4] = errmsg if id / 4 not in errmsg_dict \
- else errmsg_dict[id / 4] + ";" + errmsg
- if len(blank_line) == 1:
- ssub[blank_line[0]] = "【stem】"
- else:
- if len(ssub) == 2: # 只有一个换行时
- ssub.insert(1, "【stem】")
- elif item_no[-1] > 49:
- topicno_idx = [i for i, v in enumerate(ssub) if re.match("[5-9][0-9]\s*[.．、､]", v.strip())]
- if len(topicno_idx) == 1:
- ssub.insert(topicno_idx[0], "【stem】")
- else:
- ssub.insert(0, "【stem】") # 分不出来，先替换
- ssub.insert(0, "")
- else:
- ssub.insert(0, "【stem】") # 分不出来，先替换
- ssub.insert(0, "")
- subcon[id] = "\n".join(ssub)
- all_item = re.split(r"【stem】", "\n".join(subcon).strip())
- return all_item, item_no, errmsg_dict, count
|
