Incorrect Explanation for \W (python)

[byronharvey]

Bug Description

For Python 2.7 when I enter into the Regular expression bar r"\W" I am told in the explanation field that "\W matches any non-word character (equivalent to [^a-zA-Z0-9_])" This is not true because é, for example, does not match "\W", and it does for "[^a-zA-Z0-9_]"

Reproduction steps

Enter "\W" in the expression field

In the test field enter é

Change the expression field to "[^a-zA-Z0-9_]"

Observe that "\W" does not highlight é and "[^a-zA-Z0-9_]" does.

Expected Outcome

The explanation should make note that "\W" is not equivalent to "[^a-zA-Z0-9_]" in all cases, particularly those dealing with accented characters commonly found in other languages.

Browser

Include browser name and version Chrome latest (89.0)

OS

Include OS name and version Big Sur

[Ouims]

You're entirely correct, this is due to python being emulated by PCRE on the website, a better 'test' is to use \w, which matches é by default (read: without the /u modifier), this means that python is always getting emulated with the unicode switch of pcre

For pcre the website gives for \W with /u "\W match any non-word character in any script (equivalent to [^\p{L}\p{N}_])" which is probably correct, which python 2.7 does not seem to support, it would be not ideal to list all of the character/code point which would match \W in unicode though if it can't be done in terms of \p{}, not sure about newer version of python anyway, this can be fixed easily but note that python needs some love overall on the website.