June 2011 - dancer-users - lists.preshweb.co.uk

rendering web page with HTML in a variable, no templates or layouts
by Leandro Hermida 24 Jun '11

24 Jun '11

Hi there, Sorry if I didn't find this in the docs, is there a way with Dancer to have it render a web page using HTML stored in a variable and not use any templating or layouts? thanks, Leandro

2 2

Undocumented bug or feature
by Oleg A. Mamontov 23 Jun '11

23 Jun '11

Hello! Dancer manual (pod) tells us about `path` function: "Concatenates multiple paths together, without worrying about the underlying operating system". But unfortunately this isn't true :( In reality &Dancer::path is &File::Spec::catfile wrapped by &Cwd::realpath. And &Cwd::realpath is worked with BSD realpath (3) semantic: "All but the last component of a pathname must exist when realpath() is called." Assume that '/tmp' directory is exists and empty. Then this Dancer application code: my $x = path "/tmp","nonexistent"; my $y = path "/tmp","nonexistent","foo"; will assign "/tmp/nonexistent" to $x, but undef to $y. This behavior may be slightly unexpected :) So if you create variable with nonexistence path you may get defined or undefined string depending on missing subdirectories number. I am not sure what's the best, but (imho) ones should be fixed: Dancer pod or Dancer::path :) -- Cheers, Oleg A. Mamontov mailto: oleg(a)mamontov.net jabber: lonerr(a)charla.mamontov.net icq uin: 79-521-617 cell: +7-903-798-1352

2 1

Re: [Dancer-users] Failing tests for sessions during Dancer upgrade from 1.3050 to 1.3060
by Stefan Oberwahrenbrock 18 Jun '11

18 Jun '11

Hello Carlos, thanks for your hint! Having in mind, that I could install Dancer 1.3060 on my Windows system and now knowing, that it installs on Debian "unstable", I did some further investigations. "t/08_session/05_yaml.t" for Dancer 1.3060 introduced some further tests using temporary files and directories. Thus, my hope was to find the solution for my problem in some platform specific code, that deals with these files and directories. Finally I found out that I "just" had to upgrade my "File::Temp" from 0.18 to 0.22 - after that, Dancer installations finished successfully! Concerning the two versions of "File::Temp" there occured some changes on output strings ("is not a directory" vs. "does not exist") - Dancer installation expected the newer ones. I listed the differences for illustration purpose down below ;-) Greetings, Stefan ----<start output>---- lenny64:~# grep 'Parent' /root/Temp.pm_0.18 ${$options{ErrStr}} = "Parent directory ($parent) is not a directory"; ${$options{ErrStr}} = "Parent directory ($parent) is not writable\n"; ${$options{ErrStr}} = "Parent directory ($parent) is not safe ($safeerr)"; ${$options{ErrStr}} = "Parent directory ($parent) is not safe ($safeerr)"; but shorter. Parent directory specifications are stripped from the lenny64:~# grep 'Parent' /root/Temp.pm_0.22 ${$options{ErrStr}} = "Parent directory ($parent) does not exist"; ${$options{ErrStr}} = "Parent directory ($parent) is not a directory"; ${$options{ErrStr}} = "Parent directory ($parent) is not safe ($safeerr)"; ${$options{ErrStr}} = "Parent directory ($parent) is not safe ($safeerr)"; but shorter. Parent directory specifications are stripped from the ----<stop output>---- Am 18.06.2011 05:35, schrieb Carlos Ivan Sosa: > I can confirm that in Debian unstable, Dancer 1.3060 builts correctly > without a problem. > http://www.cpantesters.org/cpan/report/d49c478a-98f1-11e0-bf7b-ec4f2fb29734 > > > On Fri, Jun 17, 2011 at 3:00 AM, Stefan Oberwahrenbrock > <oberwahrenbrock(a)gmx.de> wrote: >> Hello fellow dancers, >> >> trying to upgrade Dancer from 1.3050 to 1.3060 on my Debian 5.0 system, I >> got an error message during "make test" when it came to run tests on Dancer >> sessions. The relevant output was this: >> >> ----<start output>---- >> >> Failed Test Stat Wstat Total Fail List of Failed >> ------------------------------------------------------------------------------- >> t/08_session/05_yaml.t 1 256 12 1 11 >> >> >> t/08_session/05_yaml...........................NOK 11/12 >> # Failed test 'session dir was not recreated' >> # at t/08_session/05_yaml.t line 58. >> # 'Error in tempfile() using >> /tmp/LGNdqcYb9t/sessions/714935461137124675383662533356367296.XXXXXXXX: >> Parent directory (/tmp/LGNdqcYb9t/sessions/) is not a directory at >> /root/.cpan/build/Dancer-1.3060-R45rPb/blib/lib/Dancer/Session/YAML.pm line >> 90 >> # ' >> # doesn't match '(?-xism:Parent directory .* does not exist)' >> # Looks like you failed 1 test of 12. >> t/08_session/05_yaml...........................dubious >> Test returned status 1 (wstat 256, 0x100) >> DIED. FAILED test 11 >> Failed 1/12 tests, 91.67% okay >> >> ----<stop output>---- >> >> I have to admit, that I don't have any practical knowledge on writing tests >> - up till now, I only used distributions and never wrote one :-) Trying to >> interpretate the message it seemed to me, that in the above mentioned >> message, the expected output just differed a little bit from the one, that >> really occured during the test ("is not a directory" vs. "does not exist"). >> Optimistically I changed line 58 of "t/08_session/05_yaml.t" >> from: >> like ($@, qr/Parent directory .* does not exist/, "session dir was not >> recreated"); >> to: >> like ($@, qr/Parent directory .* is not a directory/, "session dir was not >> recreated"); >> >> After this alteration, "make test" and "make install" just worked without >> any errors - Dancer 1.3060 is now available on my system. It would be nice, >> if anyone could comment my approach - I am not sure, if I did the right >> thing. >> >> Thanks and greetings, >> Stefan >> _______________________________________________ >> Dancer-users mailing list >> Dancer-users(a)perldancer.org >> http://www.backup-manager.org/cgi-bin/listinfo/dancer-users >> >

1 0

Dancer and i18n
by patrick sanchez 17 Jun '11

17 Jun '11

Hello, I need to display my Dancer app in at least 2 languages. Tried the i18n plugin, but i dont' need the full power of gettext/make text (plus, being a java guy -nobody is perfect- , i feel quite lost with it !). I found this method in PM : ---- In our app we take the language preferences of a user from his settings, cookie or guess it based on his browser headings or failing all, guess his language by his IP. In the head of .tt's we PROCESS the proper language template which holds the translations by key: The template inclusion is done via a: [%- PROCESS "locale/${user.lang}/lang.tt"; %] <http://www.perlmonks.org/?abspart=1;displaytype=displaycode;node_id=466456;…> The language files look like: [% lang.label.status = "Status"; .... lang.phrase.sure = "Are you sure"; %] and a tag is simply called using: [% lang.label.status %] --- It seems to perfectly suit my needs, but i can't make it work : [% lang.label.status %] would generate nothing. Has anyone tested a similar approach for i18n with Dancer ? Thanks. Patrick

2 1

Failing tests for sessions during Dancer upgrade from 1.3050 to 1.3060
by Stefan Oberwahrenbrock 17 Jun '11

17 Jun '11

Hello fellow dancers, trying to upgrade Dancer from 1.3050 to 1.3060 on my Debian 5.0 system, I got an error message during "make test" when it came to run tests on Dancer sessions. The relevant output was this: ----<start output>---- Failed Test Stat Wstat Total Fail List of Failed ------------------------------------------------------------------------------- t/08_session/05_yaml.t 1 256 12 1 11 t/08_session/05_yaml...........................NOK 11/12 # Failed test 'session dir was not recreated' # at t/08_session/05_yaml.t line 58. # 'Error in tempfile() using /tmp/LGNdqcYb9t/sessions/714935461137124675383662533356367296.XXXXXXXX: Parent directory (/tmp/LGNdqcYb9t/sessions/) is not a directory at /root/.cpan/build/Dancer-1.3060-R45rPb/blib/lib/Dancer/Session/YAML.pm line 90 # ' # doesn't match '(?-xism:Parent directory .* does not exist)' # Looks like you failed 1 test of 12. t/08_session/05_yaml...........................dubious Test returned status 1 (wstat 256, 0x100) DIED. FAILED test 11 Failed 1/12 tests, 91.67% okay ----<stop output>---- I have to admit, that I don't have any practical knowledge on writing tests - up till now, I only used distributions and never wrote one :-) Trying to interpretate the message it seemed to me, that in the above mentioned message, the expected output just differed a little bit from the one, that really occured during the test ("is not a directory" vs. "does not exist"). Optimistically I changed line 58 of "t/08_session/05_yaml.t" from: like ($@, qr/Parent directory .* does not exist/, "session dir was not recreated"); to: like ($@, qr/Parent directory .* is not a directory/, "session dir was not recreated"); After this alteration, "make test" and "make install" just worked without any errors - Dancer 1.3060 is now available on my system. It would be nice, if anyone could comment my approach - I am not sure, if I did the right thing. Thanks and greetings, Stefan

1 0

unwanted CTPP stack in logs
by Nick Knutov 16 Jun '11

16 Jun '11

After installing latest dancer 1.3060 I saw the trace stack from CTPP in stderr(or stdout?): CODE 0x01010000 ARG 0x00000000 RES 0x0000000000000000 | 0x00000000 SYSCALL 0x00000000 __ctpp2_emitter(ARGS: 0) CODE 0x04010710 ARG 0x00000000 RES 0x0000000000000000 | 0x00000001 MOV HR, STACK[0] = `HASH (0xa1179c8)` CODE 0x02020000 ARG 0x00000001 RES 0x0000000000000000 | 0x00000002 POP AR CODE 0x08020015 ARG 0x00000001 RES 0x0000000000100001 | 0x00000003 OUTPUT STRING POS: 1 (VAL: `<form action="" [...] I saw it in '-E production' mode too. Is it because of new Dancer or where I should search the problem? -- Best Regards, Nick Knutov http://knutov.com ICQ: 272873706 Voice: +7-904-84-23-130

2 2

more request for "Running multiple apps with Plack::Builder"
by Takeshi OKURA 16 Jun '11

16 Jun '11

In "Dancer::Deployment" section "Running multiple apps with Plack::Builder" example script app.psgi cause runtime error. Odd number in 'set' assignment unable to load application app1 . I think this workaround is --- app.psgi.orig 2011-06-15 23:18:34.005384527 +0900 +++ app.psgi 2011-06-15 23:42:10.281362715 +0900 @@ -5,18 +5,24 @@ my $app1 = sub { my $env = shift; - setting appdir => '/path/to/app1', + local $ENV{DANCER_APPDIR} = '/path/to/app1'; + setting appdir => '/path/to/app1'; + setting public => '/path/to/app1/public'; setting appname => 'app1'; load_app "app1"; + Dancer::App->set_running_app('app1'); Dancer::Config->load; my $request = Dancer::Request->new( env => $env ); Dancer->dance( $request ); }; my $app2 = sub { my $env = shift; - setting appdir => '/path/to/app2', + local $ENV{DANCER_APPDIR} = '/path/to/app2'; + setting appdir => '/path/to/app2'; setting appname => 'app2'; + setting public => '/path/to/app2/public'; load_app "app2"; + Dancer::App->set_running_app('app2'); Dancer::Config->load; my $request = Dancer::Request->new( env => $env ); Dancer->dance( $request ); Please review the app.psgi example script. -- Takeshi OKURA

2 1

Problems mounting multiple webapps
by Alex Kalderimis 15 Jun '11

15 Jun '11

I've been having some issues mounting multiple webapps on a single Plack server according to the documentation (http://search.cpan.org/dist/Dancer/lib/Dancer/Deployment.pod) After reading through the source a bit I managed to get my content served, but the static files are all being handled by the route handlers. It seems that somewhere down the line the 'public' setting is going undef, along with all others presumably. Anyway, the log messages I get are here: http://paste.perldancer.org/29wzxHG7VkVoW, And the psgi I'm using is here: http://paste.perldancer.org/OntvcvVUSY7k Any help or suggestions would be greatly appreciated. -- Alex Kalderimis Software Developer InterMine Cambridge Systems Biology Centre Tennis Court Road - CB2 1QR University of Cambridge ajk59(a)cam.ac.uk

5 10

Dancer::GSoC Status Report for Week #3
by Carlos Ivan Sosa 14 Jun '11

14 Jun '11

I know that Franck Cuny as been forwarding my status reports, but when ever I can I will post them to the list. If you feel like reading any other stuff related to my project, you can find all my posts in http://log.gnusosa.net/code/gsoc Suggestions, questions and comments are welcomed. Thanks. :) # The format continues as last time. You can access the HTML # formatted version of this text going to http://log.gnusosa.net/code/gsoc/status-report-3 Well on week #3, I wasn't able to do much work due to the schoolwork load of this month. Just like every other participant of GSoC, I dedicated half of my days to studying and writing to prepare for my finals. Aside from that, I could review the ideas related to the upgrade/update process on the project. I met with my mentors to sync- up with current events related to that week work, but since not much work had been done on the repo, it seemed unnecessary. What I worked on week #2 For week #3, most of my time was invested in investigating which upgrade/update process would be better for the project. It's stipulated has the work for week #4, but before that I did the research based on Catalyst::Scripts*, catalyst.pl, Catalyst::Helper. So what my mentors and me decided, was to mimic the scripts organization from Catalyst. From catatyst.pl, one feature did made me think of the possibility of implementing such in our script bin/ dancer. That feature works around writing new updated files with the ending .new, that way the user can still have its old files, and chose which solution is best, or start hacking the migration of such deprecated functions, method, file into a solution the user might find best fit for his application. Dancer's developers and contributors write a lot of code that works around with what we call, "automagically". Which is what I was looking for my project, an automatic yet not so stressful way to upgrade/ update outdated files, and that works for almost every situation. That way is a magical and automatic solution. What I decided to go for as of this moment of the project, is to use or write an algorithm that would parse each file in the lib, the config file, and the environments files, to look for outdated subs and functions. So this is where the upgrade/update process stands as of right now. So for a start we decided, that I could use a module, and if that module seems solid and lean, we might stick to it. But since Dancer depends on very basic modules, I don't want to add more to this well thought line of decisions of what works and what doesn't for Dancer. However, sawyer suggested that I could write the portion that we need for the upgrade/ update process. So if you know of such module, or you have worked on string-to-string comparison with a solid module, please let me know. Suggestions are welcome, and they really help. :) What's next? For week #4, the following must be accomplished: Write the upgrade/update methods. Write the tests for the upgrade/updates methods. This means I will be back on writing code to Dancer::Script and add a quick tweak to bin/dancer script. Also, to work again with tests, is what keeps me more motivated, I really like the whole test then write code as you go workflow, and I would like to try it more after GSoC is over. But for that I will need a solid base with testing. So more code, more tests, more coffee. :) So to every mentor and participant, have a great week #4, and to everybody still working they way through finals, best of luck and give yourself the time to relax. Cheers.

1 0

problems with UTF-8
by Mr. Puneet Kishor 12 Jun '11

12 Jun '11

using Dancer 1.3051 (that I force-installed because it wouldn't install nicely on Mac OS X) with Apache. The config file says charset: "UTF-8" A simple phrase like "EcoVal: a collaborative collection of casestudies" renders like so in Spanish and Portuguese respectively -- • EcoVal: colecci�n colaborativa de estudios • EcoVal: uma cole��o colaborativa de estudos de caso In my early days I used to use CGI::Application. I used the following to make sure everything rendered fine sub translate { return encode_entities( decode_utf8( $_[0] ), "^\t\r\n\040-\177 " ); } If I use the above with Dancer, it crashes and burn with the following mensaje. It actually doesn't tell me on which which line, in which package, did it find the offending malformed UTF-8 character (as far as I can tell). What should a polyglot do? ---- Malformed UTF-8 character (fatal) at (eval 59) line 1. Stack main in /Users/punkish/Sites/ecoval/public/dispatch.cgi l. 15 Plack::Runner in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Runner.pm l. 231 Plack::Runner in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Runner.pm l. 256 Plack::Loader in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Loader.pm l. 83 Plack::Handler::CGI in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Handler/CGI.pm l. 37 Plack::Component in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Component.pm l. 39 Plack::Middleware::StackTrace in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Middleware/StackTrace.pm l. 27 Try::Tiny in /usr/local/lib/perl5/site_perl/5.12.1/Try/Tiny.pm l. 67 Try::Tiny in /usr/local/lib/perl5/site_perl/5.12.1/Try/Tiny.pm l. 71 Plack::Middleware::StackTrace in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Middleware/StackTrace.pm l. 27 Plack::Component in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Component.pm l. 39 Plack::Middleware::Lint in /usr/local/lib/perl5/site_perl/5.12.1/Plack/Middleware/Lint.pm l. 24 Dancer::Handler in /usr/local/lib/perl5/site_perl/5.12.1/Dancer/Handler.pm l. 102 Dancer::Handler in /usr/local/lib/perl5/site_perl/5.12.1/Dancer/Handler.pm l. 71 Dancer::Handler in /usr/local/lib/perl5/site_perl/5.12.1/Dancer/Handler.pm l. 87 Dancer::Object in /usr/local/lib/perl5/site_perl/5.12.1/Dancer/Object.pm l. 15 Dancer::Error in /usr/local/lib/perl5/site_perl/5.12.1/Dancer/Error.pm l. 34 Dancer::Error in /usr/local/lib/perl5/site_perl/5.12.1/Dancer/Error.pm l. 248 ----

2 3