You're right (except for the "problem understanding" part :p). I just
noticed that I was careless enough not to run my URL-s through any
filters, so if I have a page whose URL part contains a slash, everything
breaks.
If, however, I apply a complete filter, then it messes up all slashes,
which is also not good (i.e. I use urls like
/articles/chapter-title/section-title -- each of this 3 parts defines
one page in the DB).
Indeed the Right Solution is to run a filter for each URL part.. but
this will go so deep in my Perl code that I'll rather use URI::Encode.
So... I still think the uri filter should keep its old behavior.
-M.
Randal L. Schwartz wrote:
"Mihai" == Mihai Bazon <[EMAIL PROTECTED]> writes:
If we want a uri escape that is that aggressive, can it be put under a
different name, and the existing uri be modded back to allow : and /
characters through untouched...
Mihai> $votes++
You clearly don't understand the problem then, or what uri-escaping is about.
This isn't about "voting". This is about *doing the right thing*.
You *cannot* uri-escape a string that already has a path to it.
You can only uri-escape the path steps.
_______________________________________________
templates mailing list
[email protected]
http://lists.template-toolkit.org/mailman/listinfo/templates
