Using a regex to find all lines that contain 3 or more words with consecutive identical letters

Question

An example of a matching line is: Gutter, Rubbing, Nomad, Among, Nit, Turret, Tug

This line contains 3 words that all have 2 consecutive identical letters.

I'm currently using the following regex expression with the grep -P command: (?:.*(?:(\w)\1).*){3}.

This regex works but when I have the following line for example: Gutter, Rutting, Nomad, Among, Nit, Turret, Tugekekekekekekekekekekeke I get a catastrophic backtracking error on regex101.com (https://regex101.com/r/3YFCAj/1).

How can I search for this regex without that backtracking problem?

Answer 1

If you want to use Perl style regex, that should work:

(\b\w*(\w)\2\w*\b.*){3}

Explanation:

• (\w)\2 will match two identical consecutive characters. The \2 will match the second matching group, which is the (\w) that precedes it.
• \w*(\w)\2\w* those characters should be inside a word with other word characters before and/or after.
• \b\w*(\w)\2\w*\b - and those words should be surrounded by word boundaries.
• .* - after each word we can have any other characters.
• (\b\w*(\w)\2\w*\b.*){3} - and finally this part should repeat 3 times to find at least 3 of those words.