perl-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Doug MacEachern <>
Subject Re: Apache->can() bug?
Date Wed, 14 Feb 2001 22:05:56 GMT
On Thu, 8 Feb 2001, Geoffrey Young wrote:

> hi guys...
>   oops, I forgot to cross-post my modperl@ report here...
> looks like Apache->can('push_handlers') returns true even if
> PERL_STACKED_HANDLERS=0 at build time (at least with CVS)
> I suppose that at some point this wasn't the case, since Apache::DBI seems
> to have purposely coded against can()

actually, Apache::DBI is supposed to call Apache->can_push_handlers().
i originally submitted the patch to edmund for connect_on_init that used
can_push_handlers(), i guess he assumed Apache->can('push_handlers') would
do the same.

Apache::can_push_handlers was added long before UNIVERSAL::can existed in
the Perl core.  it was also before xsubpp could properly support #ifdef in
.xs code.  so i guess the right thing todo for modern Perls would be throw
away Apache::can_push_handlers and #ifdef PERL_STACKED_HANDLERS around
mod_perl_push_handlers() in Apache.xs
i doubt anything besides Apache::DBI is using Apache::can_push_handlers,
don't think its documented either.

View raw message