I am trying to make a string JSON compliant by enclosing the attribute into double quotes within python3 so I can parse the JSON.

The regex seems not to work as intended:

re.sub(r"/(?:[\w] (?=:))/g", "", var)

String for variable var:

{
    pzn: "09900426",
    url: "/url.html",
    packageSizeValue: "28",
    packageSizeUnit: "St",
    sizeValue: "",
    potency: ""
}, {
    pzn: "09900432",
    url: "/url2.html",
    packageSizeValue: "84",
    packageSizeUnit: "St",
    sizeValue: "",
    potency: ""
}

Expected result would be:

{
    "pzn": "09900426",
    "url": "/url.html",
    ...

CodePudding user response：

It look as though you wanted to use

re.sub(r"\w (?=:)", r'"\g<0>"', var)

The \w (?=:) pattern matches any one or more word chars before a :, and the match is replaced with itself enclosed with double quotation marks. See the regex demo.

This is not safe because there can be such matches inside string values. Imagine a potency: "blah:blah:" value. In this case, it would be safer to leverage the indentation and use

re.sub(r"(?m)^(\s*)(\w ):", r'\1"\2":', var)

See the regex demo. Here,

• (?m) - equivalent of re.M / re.MULTILINE` option
• ^ - start of a line
• (\s*) - Group 1 (\1): zero or more whitespaces
• (\w ) - Group 2: one or more word chars
• : - a colon.