Regex for matching grey colors in hexadecimal notation

I've an XML file containing (among a lot of other stuff) hexadecimal color codes. I want to inspect all such codes that result in shades of grey. The pattern is

• a hash tag (#) then
• [a-fA-F0-9] repeated exactly 3 or 6 times
• [^a-fA-F0-9]

So it should match #eee, #EEEEEE, #333333 but not #00006d, #123456 and so on. Regarding the regex flavor, it should ideally work in Notepad++, if that's no option then Python 2.7 is an alternative.

I tried using the backreference operator I found here. So far, my best attempt is

#([0-9a-fA-F])\1{3}[^0-9a-fA-F]

but I'm having some troubles:

• It seems I should replace {3} by {2} for matching exactly three repetitions but I don't see why
• I'm clueless as to how to match 6 repetitions as well. I think that {3,6} should match 3, 4, 5 or 6 repetitions but how can I exclude 4 and 5 repetitions? I thought about #([0-9a-fA-F])\1{3}[^0-9a-fA-F]|#([0-9a-fA-F])\1{6}[^0-9a-fA-F] but there must be a less ugly syntax for that, right?

You can use thie regex:

#([0-9A-Fa-f])([0-9A-Fa-f])((?=\2)\1|(?:\1\2){2})\b

RegEx Demo

This should work on PCRE or on Python/Java/Javascript as well.

RegEx Details:

• #: Match a #
• ([0-9A-Fa-f]): capture group #1 to capture any one hex character
• ([0-9A-Fa-f]): capture group #2 to capture any one hex character
• (: Start capture group #3
  • (?=\2): Make sure next character is same as group #2
  • \1: Match same character as in group #1
  • |: OR
  • (?:\1\2){2}: Match 2 characters as in group #1 then group #2. Repeat this group twice.
• ): End of capture group #3
• \b: Word boundary