Java Regex Identifying Non-Consecutive Pairs of Characters/Numbers

Question

I'm looking to pick out pairs of word characters in a string, that may not necessarily be beside each other in the string. For example, "ahna32g" should match, due to the pair of 'a' characters. What I currently have is "\\w*(\\w)\\1+\\w*" which is successful if the matching characters are consecutive. I'm quite new to regular expressions, so if a detailed explanation could be given I'd really appreciate it.

Answer 1

You need to insert \\w* between (\\w) and \\1 to let the regex engine match any 0+ word chars in between the repeating chars:

\w*(\w)\w*\1+\w*
       ^^^

See the regex demo .

So, the regex will match

• \\w* - 0+ word chars
• (\\w) - capture a word char into Group 1
• \\w* - will match 0+ word chars
• \\1+ - one or more occurrences of the value inside Group 1
• \\w* - 0+ word chars.

Answer 2

My approach captures the (earliest) repeated characters in group 1 and 2. It's also less steps than the already posted answer.

\\w*?(\\w)(?=\\w*?(\\1))\\w*

\w*? // 0 or more word chars, lazily matched
(\w) // a word char (as group 1)
(?=  // look ahead and assert a match of:
\w*? // 0 or more word chars, lazily matched
(\1) // group 1 (as group 2)
)    // end of assertion
\w*  // 0 or more word chars

• Flags: g

Demo