I'm not a unicode expert but I thought the 'printable characters' in
unicode are sparse -- that there is no range to do what you want.
I wonder why the right thing isn't to fix the server?
On Oct 24, 2007, at 18:16, Elliot Winard <[EMAIL PROTECTED]>
wrote:
I meant that I'd used "[ -Ɛ]" as the pattern.
-e
On Wed, Oct 24, 2007 at 5:59 PM, Elliot Winard wrote:
I'm working on a bug where non-printable characters (specifically
the BEL character) sent by our Laszlo app are causing the server to
choke.
I'd like to restrict the character input to printable characters.
Right now I'm using the pattern "[&$32;-Ɛ]*" on inputtext
fields but I think this will disallow some non-western text.
Is there a better range to include in the pattern that will work
for all printable UTF-8 characters?
I'm just interested how to catch this in the LZX world. I know
that the behavior of the app will depend on character ranges the
back-end supports.
Thanks!
-e
