On Thu, 13 Sep 2001, Stas Bekman wrote:
> whoah! now all tests pass and print() works!!!
great news! :-)
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
On Thu, 13 Sep 2001, Stas Bekman wrote:
> is it still relevant after your commit which moves send_httpd_header to xs
> land?
no, i changed my mind :) its much faster just to turn off ParseHeaders
than to generate a Content-type header and have it parsed.
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> better yet, we should have a method to query all PerlOptions from Perl,
> end up with something like:
>
> sub send_http_header {
> my($r, $type) = @_;
> if ($type) {
> $r->content_type($type);
> }
> +if ($r->perl_options &
On 13 Sep 2001 [EMAIL PROTECTED] wrote:
> dougm 01/09/12 21:10:55
>
> Modified:src/modules/perl modperl_io.c
> Removed: lib/ModPerl XSLoader.pm
> Log:
> fix tied filehandle breakage caused by bleedperl change #11639
whoah! now all tests pass and print() works!!!
You are th
better yet, we should have a method to query all PerlOptions from Perl,
end up with something like:
sub send_http_header {
my($r, $type) = @_;
if ($type) {
$r->content_type($type);
}
+if ($r->perl_options & ModPerl::OPT_PARSE_HEADERS) {
+ $r->print("Content-type
On Thu, 13 Sep 2001, Stas Bekman wrote:
> ah, sorry, if I do:
>
> $r = shift;
> $r->send_http_header('text/plain'); # doesn't work!
> print("OK\n");
oh, i see what you mean. only with PerlOptions +ParseHeaders of course.
i guess send_http_header needs to be xs, which can check if
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Thu, 13 Sep 2001, Stas Bekman wrote:
>
> > It indeed prints 'OK' via simple print "OK", 5.7.2 patch 12004 doesn't.
> > Looks like something related to STDIN problems we see with this version of
> > Perl. Of course in this case it's the tied STDOUT.
On Thu, 13 Sep 2001, Stas Bekman wrote:
> It indeed prints 'OK' via simple print "OK", 5.7.2 patch 12004 doesn't.
> Looks like something related to STDIN problems we see with this version of
> Perl. Of course in this case it's the tied STDOUT.
could be.
> send_httpd_header doesn't work for me
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Thu, 13 Sep 2001, Stas Bekman wrote:
>
> > But with non-ithread, i've got this core dump:
>
> fixed.
confirmed, thanks.
The same registry setup was no tested with 5.6.1 no threads.
It indeed prints 'OK' via simple print "OK", 5.7.2 patch 12004 d
On Thu, 13 Sep 2001, Stas Bekman wrote:
> But with non-ithread, i've got this core dump:
fixed.
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Thu, 13 Sep 2001, Stas Bekman wrote:
>
> > > I expect print OK to work, but I don't see anything. send_http_header
> > > doesn't work too.
>
> send_http_header is working fine.
>
> for some reason STDOUT is not tied by the time your
> Apache::Regi
On Thu, 13 Sep 2001, Stas Bekman wrote:
> > I expect print OK to work, but I don't see anything. send_http_header
> > doesn't work too.
send_http_header is working fine.
for some reason STDOUT is not tied by the time your
Apache::Registry::handler runs ($r->print is ok). not sure if that is
On Thu, 13 Sep 2001, Stas Bekman wrote:
> On Wed, 12 Sep 2001, Doug MacEachern wrote:
>
> > On Thu, 13 Sep 2001, Stas Bekman wrote:
> >
> > > +sub handler {
> > > +$r = shift;
> > > +$r->send_http_header('text/plain'); # doesn't work!
> > > +$r->print("Content-type: text/plain\r\n\r\
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Thu, 13 Sep 2001, Stas Bekman wrote:
>
> > +sub handler {
> > +$r = shift;
> > +$r->send_http_header('text/plain'); # doesn't work!
> > +$r->print("Content-type: text/plain\r\n\r\n");
> > +print("OK\n");
>
> what doesn't work exact
On Thu, 13 Sep 2001, Stas Bekman wrote:
> would slurping and matching would be faster?
maybe. doesn't matter, whatever works is fine for now.
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMA
On Thu, 13 Sep 2001, Stas Bekman wrote:
> +sub handler {
> +$r = shift;
> +$r->send_http_header('text/plain'); # doesn't work!
> +$r->print("Content-type: text/plain\r\n\r\n");
> +print("OK\n");
what doesn't work exactly? what is the output when you telnet and make a
GET reque
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Thu, 13 Sep 2001, Stas Bekman wrote:
>
> > gozer has asked me why apr doesn't have a hash table.
>
> yes it does, apr_hash_*
>
> > Someone has posted a
> > patch to replace ap_table with real hash table, but it was vetoed because
> > it didn't pres
On Wed, 12 Sep 2001, Doug MacEachern wrote:
>
> > $run_configure = 0;
> > while () {
>
> i didn't want to scan every .pm line-by-line. doesn't matter, either way
> can be changed later.
would slurping and matching would be faster?
__
On Thu, 13 Sep 2001, Stas Bekman wrote:
> cool, but it has to be some large file, right? Or I can point it to
> Makefile which is always there and it's fat. Seems like a good bet.
Makefile sounds good.
-
To unsubscribe, e-ma
On Thu, 13 Sep 2001, Stas Bekman wrote:
> gozer has asked me why apr doesn't have a hash table.
yes it does, apr_hash_*
> Someone has posted a
> patch to replace ap_table with real hash table, but it was vetoed because
> it didn't preserve the order of the items as they get added.
right.
> I
> $run_configure = 0;
> while () {
i didn't want to scan every .pm line-by-line. doesn't matter, either way
can be changed later.
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
On Thu, 13 Sep 2001, Stas Bekman wrote:
> I guess we could put the code into src/code/ like I have in the guide and
> point to it. The idea is to keep the examples as stand-alone files, so
> they can be reused as is.
ok, i'll leave it to you to decide :)
--
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Tue, 11 Sep 2001, Stas Bekman wrote:
>
> > Note that the following happens outside the normal modperl build, I just
> > use Apache-Test framework, so it's possible that the problems happen,
> > becase something is not loaded outside of modperl setu
>
> > speaking of which, i'd really like to see a generated
> ChangeLog like what
> > php does. which has detail of each change made, which
> files are touched,
> > etc. then Changes will be a summary of the ChangeLog for
> each release.
>
> As I understand Geoff is taking care of that :)
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Tue, 11 Sep 2001, Stas Bekman wrote:
>
> > I'm about to move the contents of pod/ to the modperl-docs rep, which gets
> > checks out as docs/ inside modperl-2.0 root tree. There is at least one
> > test that uses a file from pod/ , should I adjust
gozer has asked me why apr doesn't have a hash table. Someone has posted a
patch to replace ap_table with real hash table, but it was vetoed because
it didn't preserve the order of the items as they get added.
Does that mean that we have to use mgv/hash and cannot use the current
ap_table type? S
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Tue, 11 Sep 2001, Stas Bekman wrote:
>
> > I need to autogenerate files which are needed during server startup. The
> > test's configure won't be run before the server gets started. It's an egg
> > and chicken problem.
> >
> > So should I try to ma
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Tue, 11 Sep 2001, Stas Bekman wrote:
>
> > Doug, this is something that you wrote a while ago. I thought to add it as
> > util/perlbloat.pl ?
>
> sure. though i'm thinking there should be two directories for util stuff,
> one for stuff that is run
On Wed, 12 Sep 2001, Doug MacEachern wrote:
> On Tue, 11 Sep 2001, Stas Bekman wrote:
>
> > Doug, shell we nuke the examples/ dir? We have a plenty of tests that
> > serve as examples. Or did you plan to put something else there in the
> > future?
gozer and I are planning to write the guide for
On Tue, 11 Sep 2001, Stas Bekman wrote:
> Note that the following happens outside the normal modperl build, I just
> use Apache-Test framework, so it's possible that the problems happen,
> becase something is not loaded outside of modperl setup.
works fine here. my test was to hack in these cha
On Wed, 12 Sep 2001, Geoffrey Young wrote:
> as brian mentioned at the conference, there is
> http://www.red-bean.com/cvs2cl/, which produces a ChangeLog like
very cool. would be nicer if:
- \n after the list of files
- the PR: Obtained from: Submitted by: Reviewed by:
could each be on the
On Tue, 11 Sep 2001, Philippe M . Chiasson wrote:
> +/* Skip entries in our filter list */
> +if (tdc_data->filter){
> +if(apr_table_get(tdc_data->filter,key)){
> +return 1;
> +}
> +}
should be skipping all _but_ those in the filter list, right?
just c
On Tue, 11 Sep 2001, Philippe M . Chiasson wrote:
> I hope this one is obvious ...
>
> xs/modperl_xs_sv_convert.h is full if nasty (love them) macros to deal with
> conversion of blessed SV* back to their pointer form and some more.
:)
> I can't figure out how to define a new alias like that,
On Tue, 11 Sep 2001, Philippe M . Chiasson wrote:
> As promised, APR::Table->do()
cool!
> A few things worth noting:
>
> * [OT] make source_scan gives me a huge diff and I have to manually weed out
>what's not
> related to my stuff before sending, annoying ... why ?
you don't need to
On Tue, 11 Sep 2001, Philippe M . Chiasson wrote:
> I noticed there are no test suite defined for the APR::Table API, so, here is
> a beginning.
thanks!
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-m
On Tue, 11 Sep 2001, Philippe M . Chiasson wrote:
> -apr_table_vdo
> + PREFIX=apr_table_
you shouldn't need to add that PREFIX, it is already taken care of.
> + ~apr_table_do
>
> And that should be it, right ?
>
> My problem is that the only way I could figure out to get the code generated
On Tue, 11 Sep 2001, Stas Bekman wrote:
> I need to autogenerate files which are needed during server startup. The
> test's configure won't be run before the server gets started. It's an egg
> and chicken problem.
>
> So should I try to make all tests require'able outside httpd and get the
> AP
On Tue, 11 Sep 2001, Stas Bekman wrote:
> Doug, this is something that you wrote a while ago. I thought to add it as
> util/perlbloat.pl ?
sure. though i'm thinking there should be two directories for util stuff,
one for stuff that is run by the build/setup and another that are run by
users.
On Tue, 11 Sep 2001, Stas Bekman wrote:
> Doug, shell we nuke the examples/ dir? We have a plenty of tests that
> serve as examples. Or did you plan to put something else there in the
> future?
maybe it should be moved to docs/ ?
i don't think to many of the tests are good user examples as-is, t
>
> there is no Changes file yet :)
> speaking of which, i'd really like to see a generated
> ChangeLog like what
> php does. which has detail of each change made, which files
> are touched,
> etc. then Changes will be a summary of the ChangeLog for
> each release.
as brian mentioned at th
On Tue, 11 Sep 2001, Philippe M . Chiasson wrote:
> Here is my second attempt at $r->print(), notice I removed the MP_USE_AP_RWRITE
> since it wasn't needed anymore.
>
> All test still pass with my patch applied.
>
> Here is the patch #2 and I even think it's proprely indented ;-)
looks good.
On Tue, 11 Sep 2001, Stas Bekman wrote:
> I'm about to move the contents of pod/ to the modperl-docs rep, which gets
> checks out as docs/ inside modperl-2.0 root tree. There is at least one
> test that uses a file from pod/ , should I adjust it to use some other
> file, e.g. Changes or should I
I guess some of the .cvsignore files started lagging behing.
The following patch adds the new files to most of the .cvsignore files
and adds 2 new ones for new test categories, hooks and api.
I was just tired of the '?' when doing cvs update.
Gozer out for the night.
--- /dev/null Sat Mar 24
43 matches
Mail list logo
