How to replace a part of substring in Python?

You may use

re.sub(r'(?<=\d),(?=\d{1,2}(?!\d))', r'.', s)

See the regex demo.

Details

• (?<=\d) - a digit must be right before the...
• , - comma that should be followed with...
• (?=\d{1,2}(?!\d)) - 1 or 2 digits (\d{1,2}) not followed with another digit ((?!\d)).

Python test:

import re
s="Euro Style = $12,345,67, US Style = $12,345,67, $1,234, $12,124,345,456 $0,89  $12,345,678"
print(re.sub(r'(?<=\d),(?=\d{1,2}(?!\d))', '.', s))
# => Euro Style = $12,345.67, US Style = $12,345.67, $1,234, $12,124,345,456 $0.89  $12,345,678

Note that in case your numbers are never glued to _ or letters, you may also use a little simpler regex with a word boundary:

(?<=\d),(?=\d{1,2}\b)
                  ^^ 

See this regex demo.

Or even - if you do not care if there is a digit before , or not:

re.sub(r',(\d{1,2})\b', r'\1', s)

See this regex demo.

And finally, if you need to also change $12,212,12345678 to $12,212.12345678

re.sub(r',(\d{1,2}|\d{4,})\b', r'.\1', s)

See yet another regex demo.

The (\d{1,2}|\d{4,}) is an alternation group that matches either 1 or 2 digits or more than 4 digits, thus eliminating 3 digit chunks followed with a word boundary.