Python中解析特殊JSON字符串为字典数组的正确方法

先看你给出的原始字符串：

raw_str = '[\'{"cx":3575,"cy":1657,"r":98}\', \'{"cx":3284,"cy":1706,"r":81}\', \'{"cx":4260,"cy":1611,"r":97}\', \'{"cx":3350,"cy":1632,"r":60}\', \'{"cx":3475,"cy":1586,"r":77}\', \'{"cx":4243,"cy":1474,"r":87}\']'

这个字符串存在几个格式问题：外层由单引号包裹，内部双引号被转成了HTML实体"，每个字典元素还被转义单引号\'包裹。下面提供两种可行的处理方案：

方案一：字符串预处理 + JSON解析 #

手动修正字符串格式后，用JSON模块完成解析：

import json

raw_str = '[\'{"cx":3575,"cy":1657,"r":98}\', \'{"cx":3284,"cy":1706,"r":81}\', \'{"cx":4260,"cy":1611,"r":97}\', \'{"cx":3350,"cy":1632,"r":60}\', \'{"cx":3475,"cy":1586,"r":77}\', \'{"cx":4243,"cy":1474,"r":87}\']'

# 将HTML实体的双引号替换为真实双引号
processed_str = raw_str.replace('"', '"')
# 去掉外层单引号，同时移除内部的转义单引号
processed_str = processed_str[1:-1].replace("'", "")
# 重新拼接为合法的JSON数组格式
valid_json = f"[{processed_str}]"
# 解析为字典数组
result = json.loads(valid_json)

# 验证结果
print(result)
print(type(result))       # 输出 <class 'list'>
print(type(result[0]))    # 输出 <class 'dict'>

方案二：用ast模块解析（更健壮） #

ast模块能正确识别转义字符，处理逻辑更省心：

import ast
import json

raw_str = '[\'{"cx":3575,"cy":1657,"r":98}\', \'{"cx":3284,"cy":1706,"r":81}\', \'{"cx":4260,"cy":1611,"r":97}\', \'{"cx":3350,"cy":1632,"r":60}\', \'{"cx":3475,"cy":1586,"r":77}\', \'{"cx":4243,"cy":1474,"r":87}\']'

# 替换HTML实体
processed_str = raw_str.replace('"', '"')
# 用ast解析成字符串列表
str_list = ast.literal_eval(processed_str)
# 逐个将字符串解析为字典
result = [json.loads(item) for item in str_list]

# 验证结果
print(result)

两种方案对比：

• 方案一流程直观，适合简单场景，但如果字符串结构有变动，可能需要调整预处理逻辑
• 方案二更健壮，ast能处理各种转义情况，适配复杂格式的字符串

内容的提问来源于stack exchange，提问作者Dmitry Sokolov