Regular Expressions Cookbook, 2nd Edition

8.13. Extracting the Query from a URL

Problem

You want to extract the query from a string that holds a URL. For example, you want to extract param=value from http://www.regexcookbook.com?param=value or from /index.html?param=value.

Solution

^[^?#]+\?([^#]+)

Regex options: Case insensitive

Regex flavors: .NET, Java, JavaScript, PCRE, Perl, Python, Ruby

Discussion

Extracting the query from a URL is trivial if you know that your subject text is a valid URL. The query is delimited from the part of the URL before it with a question mark. That is the first question mark allowed anywhere in URLs. Thus, we can easily skip ahead to the first question mark with ‹^[^?#]+\?›. The question mark is a metacharacter only outside character classes, but not inside, so we escape the literal question mark outside the character class. The first ‹^› is an anchor (Recipe 2.5), whereas the second ‹^› negates the character class (Recipe 2.3).

Question marks can appear in URLs as part of the (optional) fragment after the query. So we do need to use ‹^[^?#]+\?›, rather than just ‹\?›, to make sure we have the first question mark in the URL, and make sure that it isn’t part of the fragment in a URL without a query.

The query runs until the start of the fragment, or the end of the URL if there is no fragment. The fragment is delimited from the rest of the URL with a hash sign. Since hash signs are not permitted anywhere except in the fragment, ‹[^#]+› is all we need to match the query. The negated character class matches ...