On Wed, Nov 26, 2003 at 11:40:19AM +0000, Jean-Michel Hiver wrote:
> Hi List,
>
> At the moment my application runs as a mix of mod_perl handlers and a
> rather large-ish Apache::Registry CGI program.
>
> Initially the application was generating everything on the fly. Due to
> performance issues, I started writing a little cache manager so that
> most pages are cached, such as:
>
> http://mkdoc.com/
> http://mkdoc.com/news/newsletters/
>
> However certain dynamic pages remain uncached, such as:
>
> http://mkdoc.com/.sitemap.html
> http://mkdoc.com/.print.html
>
> What I need is a generic apache handler which would cache any HEAD and
> GET methods for a configurable amount of time, and let the POST methods
> through.
>
> It would also need to be partly purgeable. For example, if an editor
> goes to /foo/.properties.html and changes the title of /foo/, then /foo/
> needs to be purged from the cache.
>
> Is there an Apache handler I can use for this purpose?
>
> If not, which approach would you take in order to write one? I've
> contemplated the idea but I'm stuck with capturing Apache::Registry or
> Apache::RegistryNG's output.
If you're willing to enter the wild-n-wooly world of object-oriented
mod_perl handlers you might consider trying Apache::CacheContent
(available on CPAN), which is based on Cookbook::CacheContent in the
mod_perl Developer's Cookbook.
It dumps the request onto a file on disk based on the URL. If the
file is removed then the content is regenerated. The module does not
cache headers, instead it uses Apache's built-in file-serving code to
serve up cached content written to disk.
Give it a try... OO handlers are quite fun.
--
Paul Lindner [EMAIL PROTECTED] ||||| | | | | | | | | |
mod_perl Developer's Cookbook http://www.modperlcookbook.org/
Human Rights Declaration http://www.unhchr.ch/udhr/
--
Reporting bugs: http://perl.apache.org/bugs/
Mail list info: http://perl.apache.org/maillist/modperl.html
