I have an array that looks like this

{
    ...
    {
        OwnValues=
        {
            key1,
            key2,
            key3
        }, 
        ReferencedValues=
        {
            key10, 
            key11, 
            key12,
            key17
        }
    },
    {
        OwnValues=
        {
            key10,
            key20,
            key30
        }, 
        ReferencedValues=
        {
            key2, 
            key55, 
            key17
        }
    },
    {
        OwnValues=
        {
            key15,
            key17,
            key99
        }, 
        ReferencedValues=
        {
            key33, 
            key55, 
            key52
        }
   },
   ...
}

I want to match key17 as a direct child of OwnValues.

I tried this naive look behind (?<=OwnValues=)([\s\S]*key17)

But this does not work, it matches a block between the first occurence of OwnValues= and the last occurence of key17. I would like to exclude ReferencedValues inbetween. What is the proper way to regexpr this?

CodePudding user response：

You could match { and then use a negated character to match key17 without crossing matching curly's.

If you want the capture group, then you can use:

\bOwnValues=\s*{[^{}]*\b(key17)\b

• \bOwnValues= Match the word OwnValues followed by =
• \s*{ Match optional whitespace chars and then match {
• [^{}]* Match 0 times any char except { and }
• \b(key17)\b Capture the word key17 between word boundaries

See a regex demo.

Or if supported a quantifier inside a lookbehind assertion to get a match only:

(?<=\bOwnValues=\s*{[^{}]*)\bkey17\b

See another regex demo.

CodePudding user response：

This regex matches key17 in the context you've described:

OwnValues=[^=]*?(key17)

It merely matches everything that's before the next = character, to make sure the value is within the same section.