字符串转化问题

比较喜欢玩梦幻西游，最近准备买号，自己看梦幻藏宝阁呢又没空，就想着用 scrapy 抓点符合要求的数据回来，但是，当抓到用户关键信息的时候，网易返回给我的是类似下面的一段字符串，我想着给他转成一个字典或者 python 内部能识别的数据格式，但是我发现我真的不知道怎么转化。我目前的方案之后一个，那就是正则匹配，但是我感觉正则匹配是不是蠢了点。有没有大神帮我解决下这个问题的。

这儿是返回的字符串：(我已经做好了换行，记住是一个字符串，不是 python 的 list、tuple、dict)

'([
        "farm_level":"3",
        "usernum":10836654,
        "iMaxExpt3":20,
        "iBeastSki4":17,
        "rent":88332,
        "iCGTotalAmount":30,
        "xianyu":12,
        "iHp_Max":2711,
        "more_attr":(["attrs":({(["lv":0,"idx":1,]),(["lv":0,"idx":3,]),(["lv":0,"idx":9,]),(["lv":0,"idx":12,]),(["lv":0,"idx":8,]),(["lv":0,"idx":6,]),(["lv":0,"idx":11,]),(["lv":0,"idx":7,]),(["lv":0,"idx":5,]),(["lv":0,"idx":4,]),(["lv":0,"idx":2,]),(["lv":0,"idx":10,]),(["lv":129,"idx":13,]),(["lv":444,"idx":14,]),}),]),

            "AllRider":([7:(["mattrib":"","ExtraGrow":0,"exgrow":12904,"iGrade":125,"iType":508,"all_skills":([]),]),5:(["mattrib":"","ExtraGrow":0,"exgrow":13106,"iGrade":125,"iType":503,"all_skills":([]),]),4:(["mattrib":"体质","ExtraGrow":0,"exgrow":13112,"iGrade":125,"iType":502,"all_skills":([611:1,]),]),6:(["mattrib":"魔力","ExtraGrow":100,"exgrow":23075,"iGrade":125,"iType":508,"all_skills":([600:2,601:3,]),]),2:(["mattrib":"力量","ExtraGrow":90,"exgrow":22297,"iGrade":125,"iType":503,"all_skills":([600:2,601:3,]),]),]),

        "iLearnCash":8213515,
        "iBadness":0,
        "ExAvt":([11:(["order":7,"cName":"锦绣幻梦","iType":12823,]),9:(["order":11,"cName":"藤蔓蔓延","iType":14694,]),6:(["order":4,"cName":"落星织","iType":12750,]),2:(["order":5,"cName":"蓝色妖姬","iType":14404,]),1:(["order":10,"cName":"星光熠熠","iType":21513,]),3:(["order":9,"cName":"锦绣幻梦","iType":12817,]),10:(["order":8,"cName":"两心相悦","iType":14654,]),8:(["order":6,"cName":"星光熠熠","iType":21514,]),7:(["order":1,"cName":"萌草发卡","iType":13696,]),39:(["order":2,"cName":"狼牙棒","iType":11169,]),40:(["order":12,"cName":"怀旧巨魔王炫卡","iType":19287,]),4:(["order":3,"cName":"落星织","iType":40013,]),]),

        "iDef_All":410,
        "iSewski":472,
        "iSumAmount":9,
        "iSpe_All":119,
        "iMaxExpt4":20,
        "sword_score":1000,
        "AchPointTotal":4520,
        "iGoodness":7230,
        "addPoint":28,
        "iSumAmountEx":1,
        "cName":"水儋澹兮生烟",
        "iSaving":0,
        "rent_level":5,
        "iBeastSki3":17,

        ......
)]'

itype

idx

cname

order

12 条回复 • 2018-10-19 12:25:10 +08:00

rabbbit

2018-10-18 17:38:05 +08:00

import json
json.loads('["foo", {"bar":["baz", null, 1.0, 2]}]')

gz911122

2018-10-18 17:52:42 +08:00

这不就是这个正常的 json 吗

KomeijiSatori

2018-10-18 17:56:20 +08:00

https://www.json.cn/wiki.html

cassidyhere

2018-10-18 18:15:47 +08:00

[]换成{}, {}换成[]，删除()，就是 dict 了

frostming

2018-10-18 18:34:00 +08:00

()删掉
{}和[]互换
然后 json.loads

zongren

2018-10-19 08:36:03 +08:00

此贴终结

moxiaowei

2018-10-19 08:54:56 +08:00

@gz911122 这是这正常的 json ？

moxiaowei

2018-10-19 08:55:14 +08:00

@rabbbit 这样能用我又何必来这儿发帖求助

moxiaowei

2018-10-19 08:55:34 +08:00

@cassidyhere 关键就是如何替换

aniua

2018-10-19 09:49:12 +08:00

@moxiaowei 替换不就 replace 么

chenstack

2018-10-19 11:23:14 +08:00

楼上说的差不多了，只是替换后仍不是标准 json，有的键为数字，且有逗号结尾，不能直接 json.loads。还有一个是你给的字符串开头是([，结尾是)]，并不匹配，如果是成对的])，可以这样:
eval(text.replace('([', '{').replace('])', '}').replace('({', '[').replace('})', ']'))
如果想把()替换删掉，{}和[]互换，那替换时得加个临时的替换符号

moxiaowei

2018-10-19 12:25:10 +08:00

哎果然只有死方法，用正则先替换，谢谢大家解析出来了