Pyspark正则表达式替换嵌套数据帧中的方括号

Question

我有一个如下所示的数据帧：

    [{"Group":"g1","Members":
    [{"Question":"qqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqqqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqqqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqq","Answer":"aaa"}]},
    {"Group":"g2","Members":
    [{"Question":"qqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqq","Answer":"aaa"},
    {"Question":"qqqqqqq","Answer":"aaa"}]}]

我需要在以下输出中对此进行分解：

我尝试使用模式regexp\u replace（“DTB”、“\[\124;\]”、“”）删除方括号，以便获得一个字典来使用拆分和分解函数。但是我的正则表达式没有按照我的要求工作

这方面的任何帮助都将是巨大的