new_tiku_structure_2021/ceshi.py

# !/usr/bin/env/python
# -*- coding:utf-8 -*-

from flask import Flask, render_template, request
from datetime import timedelta
import requests

from utils.img2latex import get_ocrlatex_by_url

headers = {'Pragma': 'no-cache', 'Cache-Control': 'no-cache',
           'X-USER-TOKEN': 'eyJhbGciOiJIUzUxMiJ9.eyJzdWIiOiIyMCIsInVzZXJOYW1lIjoiSFgwMDEyMTM4IiwiZXhwIjoxNTY2Mzg3ODA1LCJ1c2VySWQiOiIyMCIsImlhdCI6MTU2NjM3MzQwNX0.9w__2RGncpdMvbiQAmk75ThgPwDaTXF1VyY1xKVs_zFduUEoJ3_6X2q3ZVYsFIpChKeLEWelhwlmMxEP4lU9QA',
           'Origin': 'http://xx.cn',
           'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.157 Safari/537.36',
           'Accept': '*/*',
           # 'Content-Type': 'multipart/form-data; boundary=----WebKitFormBoundarycDnnz7eKalCPv6GJ',
           'Referer': '', 'Accept-Language': 'zh-CN,zh;q=0.9,en;q=0.8'}

url = "http://192.168.1.140:18085/upload"
import json
data = {'callback_url': "http://zsytk.dev.xueping.com/wordapi/index/wid/9668",
         "mydata": json.dumps([{"file_type": "", "file_url": "http://zsytk.dev.xueping.com/data/word/2020/08/5f327c08f3576.doc"}],ensure_ascii=False).encode("utf-8")
         }

# print(requests.Request('POST', url, data=data, json=json).prepare().body.decode(
#     'ascii'))  # 打印字段名和类型
# r = requests.post(url, data=data, headers=headers)
# print(r.text)

#
# { %
# for sub in item[2:5] %}
# { % if "src" in sub %}
# < img
# src = "data:;base64,{{ img_stream }}" >
# { % endif %}
# { % endfor %}
import re

pattern_2 = re.compile(
    r"(([CDE]\s*[.．、､]|\([CDE]\)).+?)(?<![:：])\s\(([1-9]|1[0-9])\)(.+?([是为有]|等于)[(（]\s*[)）]\n)")

w1 = "C.jfg (3)mvm等于()\n"


# p1=re.search(r"每(小?题|空)[\u4e00-\u9fa5]{,5}(\d[\d.]{,2}\d{,2})分", "二、填空题（每空2分，共8分）")
# print(p1)
# p2 = re.findall("([1-9]|[1-4][0-9])[)）]?\s*(?=题)", "33题")
# print(p2)
#
# print(list(range(1,2)))


# print("{0}{1}".format(3,5))
# def fun1(nn, **kwargs):
#     print(kwargs)
#     print(kwargs['index'])
# fun1(1, all_type=1, num=2, index=3)
# print(2==2>3)
# p1 = re.search("(?<![：:.．、､])\s+([1-9]|[1-4][0-9])\s*[.．、､](?!png)", 'yy  49.png')
# print(p1)
# items_str = '\n1.yy22.pn'
#
# for no in re.findall(r'\n+\s*([1-9]|[1-4][0-9])\s*[.．、､]|(?<![：:.．、､])\s*([1-9]|[1-4][0-9])\s*[.．、､](?!png)', items_str):
#     print(no)


#
# con_id_line = []  # 题号的行索引,第几行
# topicno = []  # 题号序列
# topicno_line_dict = dict(zip([], []))  # 题号to行索引字典
# print(len(topicno_line_dict))
# print("$\\frac")
#
# w_info = re.search(' width="(\d+[.\d]*?)\s*([pxt]?)"', ' width="16"')
# print(w_info.group(1),w_info.group(2))
#
# ee = [1,2,3]
# ee=sorted(ee,reverse=True)
# print(ee)
# from PIL import Image
# local_p = r"F:\zwj\Text_Structure\img_folder\5fc64a514994183dda7e74ea\new_image1624343877418770.png"
# w = Image.open(local_p)
# w.close()
from concurrent.futures import ThreadPoolExecutor, ProcessPoolExecutor

kk = open(r"F:\zwj\parse_online58\res_folder\76736f316fe30323c8ddc51cf2b3f392__2021_08_30_10_15_58__828__1870.json",
     'r', encoding='utf-8')
con=json.loads(kk.read())
item_str = str(con['items'][4])
item_str = re.sub('(<img src=".*?".*? data-latex=)"\$.+?\$"',r'\1""', item_str)
print(eval(item_str))
print('------------------------------------')
all_imgs_no_latex = re.findall(r'(<img src="(.*?)".*? data-latex="\$?\s*\$?")', item_str)
if all_imgs_no_latex:
    imgurl_no_latex = [k[1] for k in all_imgs_no_latex]
    imgurl_no_latex = list(set(imgurl_no_latex))
    print(imgurl_no_latex)
    print('--------------------------------')
    # mathpixs = []
    # if len(imgurl_no_latex) <= 20:
    #     mathpixs.extend([get_ocrlatex_by_url(i) for i in imgurl_no_latex])
    # else:
    #     if len(imgurl_no_latex) <= 50:
    #         executor = ThreadPoolExecutor(max_workers=2)  # 开2个线程会稍微快点
    #     else:
    #         executor = ThreadPoolExecutor(max_workers=4)
    #     for data in executor.map(get_ocrlatex_by_url, imgurl_no_latex):
    #         mathpixs.append(data)
    mathpixs = [' 1: 1 ', ' b ', ')', ' 1: 1 ', '', ' a^{2}: b^{2} ', '(', 'R _ { 1 } = \\frac { p h } { a ^ { 2 } }', 'R _ { 2 } = \\frac { p b } { b h } = \\frac { p } { h }', ' a: b ', ' h ', ' R_{1}: R_{2}=b^{2}: a^{2} ', ' R_{2}=\\frac{\\rho h}{b^{2}} ', ' a: b ', ' a ']

    def sub1(ss):
        new_ss = ss.group(1) + ' data-type="math-ocr" ocr-latex="$'+mathpixs[mi]+'$"'+ss.group(2)
        return new_ss.replace(' data-type="math"', "")
    # r'\1 ocr-latex="${}$"'.format(mathpixs[mi])
    for mi, m in enumerate(imgurl_no_latex):
        # item_str = re.sub(r'(<img src="' + m + r'".*?) data-latex="\$?\s*\$?"',
        #                    sub1, item_str)
        if mathpixs[mi]:
            item_str = re.sub(r'(<img src="{}".*?) data-latex="\$?\s*\$?"(.*?(?<=")\s*/?>)'.format(m),
                              sub1, item_str)

    items_list = eval(item_str)
    print(items_list)