new_tiku_structure_v3_sci/ruku_test.py

import re
import time

import configs

wordid = "6195e6d034b3123bf9d1fbf7"
htmlt = """
一、选择题
1.随着$4G$网络和智能手机的普及,各种导航软件层出不穷,给人们的生活带来了极大的便利。
国庆放假期间,小明驾车从家里到济南大明湖游玩,某导航软件推荐的最佳驾车路线里程为
$220$公里,时间为3小时2分,下列说法正确的是
A.“3小时2分”指的是时间间隔
B."$220$公里”指的是位移大小
C.按照推荐的最佳路线走,位移一定最小
D.按照推荐的最佳路线的里程和时间推算,该车的平均速度为$20.1m/s$
"""
items_list = [{
    "stem":"随着$4G$网络和智能手机的普及,各种导航软件层出不穷,给人们的生活带来了极大的便利。国庆放假期间,小明驾车从家里到济南大明湖游玩,某导航软件推荐的最佳驾车路线里程为$220$公里,时间为$3$小时$2$分,下列说法正确的是",
    "key":"见解析",
    "parse": "略",
    "options": ['“$3$小时$2$分”指的是时间间隔', '"$220$公里”指的是位移大小', "按照推荐的最佳路线走,位移一定最小", "按照推荐的最佳路线的里程和时间推算,该车的平均速度为$20.1m/s$"],
    "options_rank": 1
}]

put_key_mjmath = ["/zyk/uploadfiles/wording/6195e6d034b3123bf9d1fbf7/MJMATH-1637661579842495.png"]  # 桶中key
local_mjmath = ["F:/zwj/Text_Structure/img_folder/6195e6d034b3123bf9d1fbf7/svg_mjmath/MJMATH-1637661579842495.png"]  # 本地图片存储位置

ltx2url = {"4G": '<img src="http://zxhx-1302712961.cos.ap-shanghai.myqcloud.com/zyk/uploadfiles/wording/6195e6d034b3123bf9d1fbf7/MJMATH-1637661579842495.png'
                                    + '" width="18.5" height="12.5" data-latex="$4G$" />'}
#
put_key_list = []  # 桶中key
localnewpic_list = []  # 本地图片存储位置
new_imgs = re.findall(r'<img src=".*?/ser_static/.*?/(new_image[^"]*?|eq_img_\d+\.png)"', str(htmlt))
new_imgs2 = re.findall(r'<img src=".*?/ser_static/.*?/(new_image[^"]*?|eq_img_\d+\.png)"', str(items_list))
new_imgs.extend(new_imgs2)
# print(new_imgs)
new_imgs = list(set(new_imgs))
if new_imgs:
    put_key_list.extend(["/zyk/uploadfiles/wording/" + str(wordid) + "/{}".format(img_name)
                    for img_name in new_imgs])
    file_path = "F:/zwj/Text_Structure/img_folder/" + str(wordid)
    localnewpic_list.extend([file_path + "/field_eq/" + img_name if "eq_img_" in img_name
                             else file_path + "/" + img_name for img_name in new_imgs])

old_imgs = re.findall(r'<img src=".*?/ser_static[/\\](.*?/word/media)/(image[^"]*?)"', str(htmlt))
if old_imgs:  # 针对最开始入库的情况，后期都直接在wordbin时就上传腾讯云了
    # 以前新添加的图片在左预览页仍以base64保存，若再解析，则会按新的方法保存到本地
    # 只是最开始解析的老图片是保存在服务器本地的，若入库时，则一起拿出来再上传腾讯云
    file_path = "F:/zwj/word_folder/" + old_imgs[0][0]
    put_key_list.extend(["/zyk/uploadfiles/wording/" + str(wordid) + "/{}".format(img[1])
                         for img in old_imgs])
    localnewpic_list.extend([file_path + "/" + img[1]  for img in old_imgs])
if True:
    put_key_list.extend(put_key_mjmath)
    localnewpic_list.extend(local_mjmath)
# ------------------------------------------------------------------------------------------

items_res_to_xbk = []
if localnewpic_list:  # 本地有新图片时
    new_img_online = '<img src="http://' + configs.public_bucket_addr + "/zyk/uploadfiles/wording/" + str(wordid)
    new_img_local = '<img src="' + configs.new_img_ip + '/' + str(wordid)  # 结构化中产生的图片
    old_img_local = '<img src="' + configs.old_img_ip + '/' + old_imgs[0][0] if old_imgs else ""  # wordbin最初产生的图片

    def sub1(s):
        if s:
            s = re.sub(r'(<img src="[^"]*?[a-z\d])\\(?!\\)([^"]*?")', r"\1/\2", s)  # 将路径中的\改为/
            s = s.replace(new_img_local, new_img_online)
            if old_img_local:
                return s.replace(old_img_local, new_img_online)
        return s
    def sub2(s):
        if s:
            all_ltx = re.findall(r'\$.*?\$', s)  # 查找试题结构中的公式
            all_ltx.extend(re.findall('\\\\\(.*?\\\\\)', s))
            all_ltx = list(set(all_ltx))
            for ltx in all_ltx:
                new_ltx = ltx.replace("$", "").replace("\\(", "").replace("\\)", "")
                if new_ltx in ltx2url:  # 将latex换为其渲染图片的线上可访问地址
                    s = s.replace(ltx, ltx2url[new_ltx])
        return s

    time6 = time.time()
    # 4>> 结构化题目中图片地址替换,需要区分下学管端还是云题库！！！！！一定会保存一份在资源库
    items_res_to_zyk = items_list.copy()
    for one_items in items_res_to_zyk:
        for k in ["stem", "key", "parse", "options"]:  # "analysis",
            if k in one_items:
                if k == "options":
                    one_items[k] = list(map(sub1, one_items[k]))
                else:
                    one_items[k] = sub1(one_items[k])

    # -----------难度和知识点自动标注------------------------
    t11 = time.time()
    diffs_xbk = [3] * len(items_res_to_zyk)

    # items_res_to_xbk = items_res_to_zyk.copy()
    # if self.callback_url and self.ltx2url:
    for nn, one_items in enumerate(items_res_to_zyk):
        new_one_item = {"difficulty": diffs_xbk[nn],
                        "knowledge": {}}
        # if 1:
        #     new_one_item["topic_type_id"] = one_items["checkType"]["id"]
        # if one_items["checkType"]["name"] == "填空题":
        #     new_one_item["blank_num"] = one_items["blank_num"]
        keys_items = ["stem", "key", "parse", "options", "options_rank"]
        # if one_items['img_status'] == 1 and ("stem_img" in one_items and one_items["stem_img"]):
        #     logger.info("----【word_id:{}】mathjax2svg所取的字段是带img的".format(self.wordid))
        #     keys_items = ["stem_img", "key_img", "parse_img", "options_img"]

        if 1 and ltx2url:
            for k in keys_items:
                if k in one_items:
                    if k == "options":
                        one_items[k] = list(map(sub2, one_items[k]))
                    elif k != "options_rank":
                        one_items[k] = sub2(one_items[k])
                    new_one_item[k] = one_items[k]
                else:
                    print("----【word_id:{}】第{}道题{}字段有问题".format(wordid, one_items["topic_num"], k))
        else:
            for k in keys_items:
                if k in one_items:
                    new_one_item[k] = one_items[k]
        # if self.subject == "高中物理":
        #     temp_items = {"topic_id": one_items["topic_num"]}
        #     temp_items["content"] = new_one_item["stem"]
        #     temp_items["parse"] = str(new_one_item["key"]) + "<br/>" + new_one_item["parse"] + "<br/>" + \
        #                           new_one_item["analysis"]
        #     temp_items["option"] = new_one_item["options"] if "options" in new_one_item else []
        #     auto_kps = self.get_phy_kps_auto(temp_items)
        #     # print("自动标注考点：", auto_kps)
        #     new_one_item["knowledge"] = auto_kps
        items_res_to_xbk.append(new_one_item)


    # 5>> ocr-htmlt中图片地址替换成云上地址
    htmlt = re.sub(r'(<img src="[^"]*?[a-z\d])\\(?!\\)([^"]*?")', r"\1/\2", htmlt)
    htmlt = htmlt.replace(new_img_local, new_img_online)
    if old_img_local:
        htmlt = htmlt.replace(old_img_local, new_img_online)
    print(htmlt)
    print(items_res_to_xbk)
    print("----【word_id:{}】结构化试题中图片地址替换时间：{}".format(wordid, time.time() - time6))