How to match language specific chars in a regex?

Question

I know that \w match any word character [a-zA-Z0-9_] or [\p{L}\p{N}_] if compiled with (?u).

The french language has 42 characters [a-zàâæçéêëîïôœùûüÿ]. Is it possible to build my regex according to my locale.

How can I match [a-zàâæçéêëîïôœùûüÿ] with \w?

A partial answer would be to use unicode regexes with \p{Latin}.

Please tag your question with the programming language you are using. You will get a better answer this way. — Tim Biegeleisen, Aug 28 '15 at 06:49
possible duplicate of [Regular expression to match non-English characters?](http://stackoverflow.com/questions/150033/regular-expression-to-match-non-english-characters) — KeyNone, Aug 28 '15 at 06:50
It depends on your regular expression engine, but `\w` typically is, or can be made, locale-sensitive. — chepner, Aug 28 '15 at 06:51
regex101 seems to only use the `u` modifier for matching Unicode characters, which might be more general than you want. (For example, `ø` is not in the French alphabet, but `/\w/u` would match it.) — chepner, Aug 28 '15 at 06:57
The question is now tagged 'perl' and 'pcre'. Which of the two are you asking about? — reinierpost, Aug 28 '15 at 08:15
@reinierpost Both actually. regex101 which I am using quite often is `pcre`, on the shell I prefer `perl` over `sed`, but I am also using `ag` which is `pcre`... — nowox, Aug 28 '15 at 20:07

chepner · Accepted Answer · 2015-08-28T16:30:57.943

2

The l modifier makes the match locale-aware:

"foo" ~= m/\w/l;

Instead of using l directly, though, use use locale per mob's link.

edited Aug 28 '15 at 16:30

answered Aug 28 '15 at 06:53

chepner

Is `l` a `Perl` specific flag? – nowox Aug 28 '15 at 06:55
Possibly? There is no standard set of flags, which is why regular-expression questions require a specific language tag. – chepner Aug 28 '15 at 06:57
3

https://metacpan.org/pod/distribution/perl/pod/perlre.pod#Character-set-modifiers -- recommends that you don't use the `/l` modifier directly. Instead, `use locale` and any regex compiled in the scope of your locale will implicitly use the `'/l` modifier. – mob Aug 28 '15 at 13:47

1 Answers1