Re: [Catalyst] Adding DBIC schema causes TT error
On Thu, Jun 21, 2007 at 09:01:03AM +1000, Charlie Garrison wrote:
Good morning,
On 20/6/07 at 7:49 AM +0100, Matt S Trout
[EMAIL PROTECTED] wrote:
Most probable cause:
You've got a mistyped $c-model(...) or c.model(...) declaration somewhere
which is falling through to Catalyst's regex search, and -was- originally
finding the right class because of that but the addition of the extra class
has caused the hash order for the search to change and now you aren't.
Likely candidates are either a typo or missing DB:: off a DB::Foo-style
DBIC::Schema model reference.
In an effort to eliminate possible spelling errors as the cause,
I removed all controllers except Root and changed Root.pm to
only have one action (index) which simply sets
$c-response-body. In that configuration the problem goes away.
I slowly added controllers back and the problem didn't recur
until I added the last controller back. To determine if problem
was with that specific controller, I removed a different
controller and the errors goes away.
So, I'm not having the same problem with controllers as with
models. There seems to be a limit on the number of controllers
or models. I add one more and I get the 'file not found' error,
it doesn't seem to matter which controller or model I add, only
the number of them.
There is no such limit.
We've seen this before and it's -always- turned out to be hash-order
related in terms of the fall-through component search, a certain number
of entries causes the keys to be searched in a different order.
At this point rather than asking for specific solutions, does
anyone have suggestions on how to debug this problem? Bill
suggested enabling debugging in Template::Provider (see other
message in this thread) but I'm not having any luck with that.
Try, in your MyApp.pm, adding
sub _comp_search {
my $c = shift;
warn Fallthrough regex search called for @_;
$c-NEXT::_comp_search(@_);
}
and looking for that warning being activated.
--
Matt S Trout Need help with your Catalyst or DBIx::Class project?
Technical DirectorWant a managed development or deployment platform?
Shadowcat Systems Ltd. Contact mst (at) shadowcatsystems.co.uk for a quote
http://chainsawblues.vox.com/ http://www.shadowcatsystems.co.uk/
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
On Thu, Jun 21, 2007 at 10:30:17AM +1000, Charlie Garrison wrote:
Good morning,
On 20/6/07 at 4:32 PM +1000, Charlie Garrison
[EMAIL PROTECTED] wrote:
I've run into a problem that I can't make any progress with;
hoping someone might be able offer suggestions.
Development was moving along nicely and then I suddenly got the following
error in a response:
Couldn't render template file error - welcome.tt2: not found
I broke out the debugger and I think I found the cause of the
problem (at least *a* cause if not *the* cause). I have a custom
view that I use for creating email messages. That view is
getting found instead of the default View::TT.
That also explains why the TT:P debug output was not being
printed; the TT view was not being used.
So, what are my best options for fixing this? (I'd also like to
learn more and understand why adding another controller or
module causes a different view to be used.)
I'm guessing the best solution is to specify which view
RenderView should be using. I have done that by adding
default_view = 'TT' to MyApp-config.
__PACKAGE__-config(
name = 'MyApp',
default_view = 'TT',
);
I also tested by setting current_view in the end/RenderView action.
$c-stash-{current_view} = 'TT';
Is one of those a 'preferred' method for setting the default view?
Well, one is global and one is per-request.
So both are preferred for their specific purpose :)
Is the problem I encountered symptomatic of a bug within
Catalyst that needs to be fixed? As a 'quick' solution can I
suggest the Catalyst debug output show which view is being used.
Knowing that a different view was being used would have saved me
many hours on this.
I think we'll probably make the fallthrough regex either warn or (if nobody's
actually relying on it, which I -hope- they aren't) simply not exist in 5.80.
I've never seen it mentioned in over a year except as a causer of bugs, so
I think it may be time to screw backcompat on that particular misfeature and
do something more sensible.
--
Matt S Trout Need help with your Catalyst or DBIx::Class project?
Technical DirectorWant a managed development or deployment platform?
Shadowcat Systems Ltd. Contact mst (at) shadowcatsystems.co.uk for a quote
http://chainsawblues.vox.com/ http://www.shadowcatsystems.co.uk/
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good morning, On 21/6/07 at 6:31 PM +0100, Matt S Trout [EMAIL PROTECTED] wrote: Is one of those a 'preferred' method for setting the default view? Well, one is global and one is per-request. So both are preferred for their specific purpose :) I figured as much and I appreciate the clarification. I had already gone with the global (config) approach. Is the problem I encountered symptomatic of a bug within Catalyst that needs to be fixed? As a 'quick' solution can I suggest the Catalyst debug output show which view is being used. Knowing that a different view was being used would have saved me many hours on this. I think we'll probably make the fallthrough regex either warn or (if nobody's actually relying on it, which I -hope- they aren't) simply not exist in 5.80. I've never seen it mentioned in over a year except as a causer of bugs, so I think it may be time to screw backcompat on that particular misfeature and do something more sensible. Sounds good to me. I've been really impressed with Catalyst so far, and getting rid of 'unexpected' behaviour would be a good thing. Thanks for all your help with this. Charlie -- Charlie Garrison [EMAIL PROTECTED] PO Box 141, Windsor, NSW 2756, Australia O ascii ribbon campaign - stop html mail - www.asciiribbon.org http://www.ietf.org/rfc/rfc1855.txt ___ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
On Wed, 2007-06-20 at 16:32 +1000, Charlie Garrison wrote:
Good afternoon,
I've run into a problem that I can't make any progress with;
hoping someone might be able offer suggestions.
Development was moving along nicely and then I suddenly got the
following error in a response:
Couldn't render template file error - welcome.tt2: not found
Looks like a side-effect of another error. If $c-stash-{template} is
not set, it defaults to the action name.
Presumably you have a sub called 'welcome' in your controller class and
under normal processing you set $c-stash-{template} = something.
What happens if an error occurs with your DBIC schema manipulation -
does it bypass the setting of the template?
HTH.
--
Jon
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
On Wed, Jun 20, 2007 at 04:32:27PM +1000, Charlie Garrison wrote: Good afternoon, I've run into a problem that I can't make any progress with; hoping someone might be able offer suggestions. Development was moving along nicely and then I suddenly got the following error in a response: Couldn't render template file error - welcome.tt2: not found The only thing I could find that would stop the error was removing the last DBIC schema class that I added to load_classes. Since I couldn't see anything obviously wrong with the class/module, I tried duplicating a known good class and renaming it. When I add that class I get the same error. I tested with a variety of different modules and the only consistency I could find was that adding one more class caused the above error. I tried changing the order of classes loaded and the problem persists. I leave the new class in place and remove one of the existing classes, and the problem disappears. There seems to be a maximum limit on the number of classes I can have loaded. I'm only at 19 classes and I find it hard to believe that would be a limit. So... Most probable cause: You've got a mistyped $c-model(...) or c.model(...) declaration somewhere which is falling through to Catalyst's regex search, and -was- originally finding the right class because of that but the addition of the extra class has caused the hash order for the search to change and now you aren't. Likely candidates are either a typo or missing DB:: off a DB::Foo-style DBIC::Schema model reference. -- Matt S Trout Need help with your Catalyst or DBIx::Class project? Technical DirectorWant a managed development or deployment platform? Shadowcat Systems Ltd. Contact mst (at) shadowcatsystems.co.uk for a quote http://chainsawblues.vox.com/ http://www.shadowcatsystems.co.uk/ ___ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good afternoon,
On 20/6/07 at 4:09 PM +0930, Jon Schutz
[EMAIL PROTECTED] wrote:
Looks like a side-effect of another error. If $c-stash-{template} is
not set, it defaults to the action name.
Presumably you have a sub called 'welcome' in your controller class and
under normal processing you set $c-stash-{template} = something.
What happens if an error occurs with your DBIC schema manipulation -
does it bypass the setting of the template?
Thanks for the suggestion, but I don't think this is it. The
$c-stash-{template} is being set explicitly and I've tested
against different controllers actions and the error from TT
always shows the expected template name.
Thanks,
Charlie
--
Charlie Garrison [EMAIL PROTECTED]
PO Box 141, Windsor, NSW 2756, Australia
O ascii ribbon campaign - stop html mail - www.asciiribbon.org
http://www.ietf.org/rfc/rfc1855.txt
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good afternoon,
On 20/6/07 at 7:49 AM +0100, Matt S Trout
[EMAIL PROTECTED] wrote:
Most probable cause:
You've got a mistyped $c-model(...) or c.model(...) declaration somewhere
which is falling through to Catalyst's regex search, and -was- originally
finding the right class because of that but the addition of the extra class
has caused the hash order for the search to change and now you aren't.
This sounds like a possible cause, but I just did a search
across all my app modules for 'model(' and none of them appear
to be incorrect. And I don't reference any models directly from
my template views.
Likely candidates are either a typo or missing DB:: off a DB::Foo-style
DBIC::Schema model reference.
I also checked all the relationship declarations for a typo and
couldn't see anything there either. End of a long work day now
so I'll leave it and hope for other suggestions. Or maybe fresh
eyes in the morning will spot the typo.
Thanks for your assistance and if you have other suggestions,
please let me know.
Thanks,
Charlie
--
Charlie Garrison [EMAIL PROTECTED]
PO Box 141, Windsor, NSW 2756, Australia
O ascii ribbon campaign - stop html mail - www.asciiribbon.org
http://www.ietf.org/rfc/rfc1855.txt
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good afternoon, On 20/6/07 at 7:49 AM +0100, Matt S Trout [EMAIL PROTECTED] wrote: You've got a mistyped $c-model(...) or c.model(...) declaration somewhere which is falling through to Catalyst's regex search, and -was- originally finding the right class because of that but the addition of the extra class has caused the hash order for the search to change and now you aren't. One other thing. With the test I'm doing now, there are no models being loaded (via $c-model or c.model), so would a misspelling have any effect anyway? Charlie -- Charlie Garrison [EMAIL PROTECTED] PO Box 141, Windsor, NSW 2756, Australia O ascii ribbon campaign - stop html mail - www.asciiribbon.org http://www.ietf.org/rfc/rfc1855.txt ___ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
On Wed, Jun 20, 2007 at 04:32:27PM +1000, Charlie Garrison wrote: Good afternoon, I've run into a problem that I can't make any progress with; hoping someone might be able offer suggestions. Development was moving along nicely and then I suddenly got the following error in a response: Couldn't render template file error - welcome.tt2: not found What version of TT are you using? Does it get fixed if you restart the dev server? Not too old versions of TT had a problem that once an error was found in a template the template then always missing. I don't see it mentioned in Changes but I suspect you need 2.19 if that's the problem you are hitting. -- Bill Moseley [EMAIL PROTECTED] ___ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good morning, On 20/6/07 at 6:31 AM -0700, Bill Moseley [EMAIL PROTECTED] wrote: What version of TT are you using? Does it get fixed if you restart the dev server? I *was* using 2.18. Yes, I have been restarting between tests. Not too old versions of TT had a problem that once an error was found in a template the template then always missing. I don't see it mentioned in Changes but I suspect you need 2.19 if that's the problem you are hitting. I saw recent messages in the archive regarding that (maybe from you). So I went looking for templates cached to disk. From what I can tell, templates will only be cached to disk if I set up my TT options accordingly, is that correct? So I should only be dealing with memory cache which should get cleared each time I restart the dev server. Upgrading to TT 2.19 has not fixed my problem. If anyone else has suggestions I'd love to hear them. Development has ground to a halt here and I'm getting very anxious (deadlines and all that). Thanks, Charlie -- Charlie Garrison [EMAIL PROTECTED] PO Box 141, Windsor, NSW 2756, Australia O ascii ribbon campaign - stop html mail - www.asciiribbon.org http://www.ietf.org/rfc/rfc1855.txt ___ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
On Thu, Jun 21, 2007 at 07:20:20AM +1000, Charlie Garrison wrote:
I saw recent messages in the archive regarding that (maybe from
you). So I went looking for templates cached to disk. From what
I can tell, templates will only be cached to disk if I set up my
TT options accordingly, is that correct? So I should only be
dealing with memory cache which should get cleared each time I
restart the dev server.
That problem I was talking about was independent of disk caching.
It means you often had to restart the server after a template error
or else the template would not be found even after correcting the
error in the template. Doesn't really sound like your problem,
although it's a bit odd that adding a schema class would effect
how TT finds templates.
But, yes, you have to enable disk caching of compiled templates
explicitly.
Upgrading to TT 2.19 has not fixed my problem. If anyone else
has suggestions I'd love to hear them. Development has ground to
a halt here and I'm getting very anxious (deadlines and all that).
If you are getting a template not found error then maybe enabling
debugging in Template::Provider would help -- or just throw in a few
warn messages to see what paths it's looking for the template.
r1024 | abw | 2006-08-01 10:35:45 -0700 (Tue, 01 Aug 2006) | 5 lines
* Fixed bug in Template::Parser to ensure $self-{ DEFBLOCKS } is cleared
each time a parse() begins.
* Fixed bug in Template::Provider to immediately invalidate the cache entry
of any template that fails to compile.
--
Bill Moseley
[EMAIL PROTECTED]
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good morning,
On 20/6/07 at 2:55 PM -0700, Bill Moseley [EMAIL PROTECTED] wrote:
Doesn't really sound like your problem,
although it's a bit odd that adding a schema class would effect
how TT finds templates.
I would agree with that. :-|
If you are getting a template not found error then maybe enabling
debugging in Template::Provider would help -- or just throw in a few
warn messages to see what paths it's looking for the template.
Would you mind giving some specifics. I added DEBUG = '1' to
__PACKAGE__-config in my View::TT but I'm not seeing any extra
debug info. I am using a mostly default setup for View::TT,
__PACKAGE__-config({
CATALYST_VAR = 'Catalyst',
INCLUDE_PATH = [
RP-path_to( 'root', 'src' ),
RP-path_to( 'root', 'lib' )
],
PRE_PROCESS = 'config/main',
WRAPPER = 'site/wrapper',
ERROR= 'error.tt2',
TIMER= 0,
TEMPLATE_EXTENSION = '.tt2',
DEBUG= '1',
});
How do I enable debugging in Template::Provider? Following the
man page, I also tried this:
use Template::Constants qw( :debug );
__PACKAGE__-config({
..
DEBUG= DEBUG_PROVIDER,
});
With that option I got this:
[Template::Provider] creating cache of unlimited slots for [
CODE(0x1d38890) ]
But no other debugging output that I could see.
Thanks,
Charlie
--
Charlie Garrison [EMAIL PROTECTED]
PO Box 141, Windsor, NSW 2756, Australia
O ascii ribbon campaign - stop html mail - www.asciiribbon.org
http://www.ietf.org/rfc/rfc1855.txt
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good morning, On 20/6/07 at 7:49 AM +0100, Matt S Trout [EMAIL PROTECTED] wrote: Most probable cause: You've got a mistyped $c-model(...) or c.model(...) declaration somewhere which is falling through to Catalyst's regex search, and -was- originally finding the right class because of that but the addition of the extra class has caused the hash order for the search to change and now you aren't. Likely candidates are either a typo or missing DB:: off a DB::Foo-style DBIC::Schema model reference. In an effort to eliminate possible spelling errors as the cause, I removed all controllers except Root and changed Root.pm to only have one action (index) which simply sets $c-response-body. In that configuration the problem goes away. I slowly added controllers back and the problem didn't recur until I added the last controller back. To determine if problem was with that specific controller, I removed a different controller and the errors goes away. So, I'm not having the same problem with controllers as with models. There seems to be a limit on the number of controllers or models. I add one more and I get the 'file not found' error, it doesn't seem to matter which controller or model I add, only the number of them. At this point rather than asking for specific solutions, does anyone have suggestions on how to debug this problem? Bill suggested enabling debugging in Template::Provider (see other message in this thread) but I'm not having any luck with that. Some very anxious people mulling around on this end. All assistance is appreciated. Thanks, Charlie -- Charlie Garrison [EMAIL PROTECTED] PO Box 141, Windsor, NSW 2756, Australia O ascii ribbon campaign - stop html mail - www.asciiribbon.org http://www.ietf.org/rfc/rfc1855.txt ___ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/
Re: [Catalyst] Adding DBIC schema causes TT error
Good morning,
On 20/6/07 at 4:32 PM +1000, Charlie Garrison
[EMAIL PROTECTED] wrote:
I've run into a problem that I can't make any progress with;
hoping someone might be able offer suggestions.
Development was moving along nicely and then I suddenly got the following error
in a response:
Couldn't render template file error - welcome.tt2: not found
I broke out the debugger and I think I found the cause of the
problem (at least *a* cause if not *the* cause). I have a custom
view that I use for creating email messages. That view is
getting found instead of the default View::TT.
That also explains why the TT:P debug output was not being
printed; the TT view was not being used.
So, what are my best options for fixing this? (I'd also like to
learn more and understand why adding another controller or
module causes a different view to be used.)
I'm guessing the best solution is to specify which view
RenderView should be using. I have done that by adding
default_view = 'TT' to MyApp-config.
__PACKAGE__-config(
name = 'MyApp',
default_view = 'TT',
);
I also tested by setting current_view in the end/RenderView action.
$c-stash-{current_view} = 'TT';
Is one of those a 'preferred' method for setting the default view?
Is the problem I encountered symptomatic of a bug within
Catalyst that needs to be fixed? As a 'quick' solution can I
suggest the Catalyst debug output show which view is being used.
Knowing that a different view was being used would have saved me
many hours on this.
Anyway, back to work now. Even with this loss of time I'm still
getting this project done much faster with Catalyst than using
my old home-made framework. It's really good kit.
Thanks,
Charlie
--
Charlie Garrison [EMAIL PROTECTED]
PO Box 141, Windsor, NSW 2756, Australia
O ascii ribbon campaign - stop html mail - www.asciiribbon.org
http://www.ietf.org/rfc/rfc1855.txt
___
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
