regex and internationalisation

OwenM

I'm validating user input on the browser to exclude special chars like !"£$%^&* using a regex something like [a-zA-Z0-9 ]

The regex is doing what I want, my question is about internationalisation, (which is being done later, (I know))

Will I have to adjust my regex per locale and hence should it itself be an externalisable string so that it is easier to do?

Tks,
Owen.

Colonel Panic

Best to plan ahead and have it possible to localise! Especially if it's being done later.

smcelhinney

In this instance what I would do is externalise in a properties file or locale bundle (Java speak).

Eg.

regex_en.properties might contain
valid_string_regex=/[A-Za-z0-9 ]+/

and regex_fr.properties might contain
valid_string_regex=/[A-Za-z0-9éáí ]+/ (..etc)

HTH
Stephen

Skrynesaver

Many languages' regex library is based on Perl's (PCRE - Perl compatible regular expressions). Perl has locale dependent character class defined what you seem to want is \w (a word character). As it locale dependent i18n is automatic.

Check the docs for your languages regex implementation.