In the following string: <div class="x" bla bla bla <div class="x" bla bla bla <div class="x" bla bla bla EXAMPLE STRING  I want to match: <div class="x" bla bla bla EXAMPLE STRING NOT <div class="x" bla bla bla <div class="x" bla bla bla <div ...

Sorry for committing thread necromancy, but I had some time on my hands and this post had no answer . My idea of a regex for this problem looks like this: (?is)<meta\s[^>]*\bcontent\s*=\s*(['"])?((?(1)(?!\1).*?|[^\s>]*))(?(1)\1) In PHP you would do something like: $pattern = ...