I need to extract content of unbalanced paren construction. In manual for PCRE i found solution for matching balanced parens.
<\[ ( (?>[^(<\[|\]>)]+) | (?R) )* \]>
For my test
<[<[ab<[cd]>]><[ef]>
It extracts
0.0: <[ab<[cd]>]>
0.1: <[ef]>
But i want to extract same content without outermost parens:
0.0: ab<[cd]>
0.1: ef
Could anybody point a solution?
Well, from the look of your regex, the content inside the outermost enclosure (they're not parentheses in any normal usage of the term) is already being captured in a parenthesis group. I don't know what context you're using the PCRE library in, but the extractions you want should be present in "match #1" (where the entire pattern match is match #0). i.e. your data should already look like:
0.0: <[ab<[cd]>]>
0.1: <[ef]>
1.0: ab<[cd]>
1.1: ef