[Koha-devel] Double encoding of query_cgi

[David Cook]

Hey all:

It looks like there's a problem with the double encoding of query_cgi (at
least in the OPAC).

It looks like Bug 9239 (commit a8f23264) was responsible for adding
uri_escaping to query_cgi into C4::Search.

Unfortunately, we also run query_cgi through "html" and "url" filters on
almost every .tt template page it features upon. 

This means that we always have double-encoded query strings, which. isn't
really a problem most of the time. Chrome and Firefox seem to be able to
handle double encoded query strings just fine. There might be certain
conditions where problems arise due to what's put in the query string. but I
haven't tried a million test cases.

I suppose the thing I'm wondering is. do we remove the filters from the
templates or stop encoding in the module? 

 

David Cook

Systems Librarian

Prosentient Systems

72/330 Wattle St, Ultimo, NSW 2007

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.koha-community.org/pipermail/koha-devel/attachments/20141126/1d3575f6/attachment.html>