I have a Python regex which validates names given by a user in a chatbot.
(?=.{2,100}$)[^\W\d_]{2,}(?:[-\s][^\W\d_]{2,})*
As you can see in the demo, it works great, but I have still problems for the following inputs:
As you can see, I want my regex match even if the user types his name with a punctuation mark. How would you improve the regex?
To enhance the current pattern you can replace {2,} limiting quantifiers with + to allow single char prefixes like O' or d' and add an optional [.?!]? pattern at the end:
^(?=.{2,100}$)[^\W\d_]+(?:[-' ][^\W\d_]+)*[.?!]?$
See the regex demo. You may add more punctuation chars you want to allow at the end of the name into the character class (remember to escape ^, ], - and \ inside square brackets).
Details:
^ - start of string(?=.{2,100}$) - there must be two to 100 chars in the string[^\W\d_]+ - one or more letters(?:[-' ][^\W\d_]+)* - zero or more occurrences of a space, ' or - followed with one or more letters[.?!]? - an optional ?, !, .$ - end of string.