What RegEx string will find the last (rightmost) group of digits in a string?

Question

Looking for a regex string that will let me find the rightmost (if any) group of digits embedded in a string. We only care about contiguous digits. We don't care about sign, commas, decimals, etc. Those, if found should simply be treated as non-digits just like a letter.

This is for replacement/incrementing purposes so we also need to grab everything before and after the detected number so we can reconstruct the string after incrementing the value so we need a tokenized regex.

Here's examples of what we are looking for:

• "abc123def456ghi" should identify the'456'
• "abc123def456ghi789jkl" should identify the'789'
• "abc123def" should identify the'123'
• "123ghi" should identify the'123'
• "abc123,456ghi" should identify the'456'
• "abc-654def" should identify the'654'
• "abcdef" shouldn't return any match

As an example of what we want, it would be something like starting with the name 'Item 4-1a', extracting out the '1' with everything before being the prefix and everything after being the suffix. Then using that, we can generate the values 'Item 4-2a', 'Item 4-3a' and 'Item 4-4a' in a code loop.

Now If I were looking for the first set, this would be easy. I'd just find the first contiguous block of 0 or more non-digits for the prefix, then the block of 1 or more contiguous digits for the number, then everything else to the end would be the suffix.

The issue I'm having is how to define the prefix as including all (if any) numbers except the last set. Everything I try for the prefix keeps swallowing that last set, even when I've tried anchoring it to the end by basically reversing the above.

Answer 1

How about:

^(.*?)(\d+)(\D*)$

then increment the second group and concat all 3.

Explanation:

^         : Begining of string
  (       : start of 1st capture group
    .*?   : any number of any char not greedy
  )       : end group
  (       : start of 2nd capture group
    \d+   : one or more digits
  )       : end group
  (       : start of 3rd capture group
    \D*   : any number of non digit char
  )       : end group
$         : end of string

The first capture group will match all characters until the first digit of last group of digits before the end of the string.

or if you can use named group

^(?<prefix>.*?)(?<number>\d+)(?<suffix>\D*)$