Skip Menu |

This queue is for tickets about the Catalyst-Plugin-Session CPAN distribution.

Report information
The Basics
Id: 71142
Status: open
Priority: 0/
Queue: Catalyst-Plugin-Session
People
Owner: Nobody in particular
Requestors: KENTNL [...] cpan.org
Cc:
AdminCc:
Bug Information
Severity: Important
Broken in: 0.32
Fixed in: (no value)

History Show all quoted text
  Thu Sep 22 06:46:05 2011 KENTNL [...] cpan.org - Ticket created
Subject: t/live_verify_address.t fails test 6 Content contains "VAR_logged=n.a."
My suspicion is that $ENV{REMOTE_ADDRESS} is being ignored and is just being mapped to 127.0.0.1.

I hacked up the Root.pm in a few places to report the address and it does appear to be doing just that.

Will keep poking.
t/live_verify_address.t .....
1..12
ok 2 - Content contains "logged in"
ok 4 - Content contains "session variable set"
not ok 6 - Content contains "VAR_logged=n.a."
 
#   Failed test 'Content contains "VAR_logged=n.a."'
#   at t/live_verify_address.t line 41.
#     searched: "VAR_logged=in"
#   can't find: "VAR_logged=n.a."
#         LCSS: "VAR_logged="
# LCSS context: "VAR_logged=in"
[debug] 127.0.0.1
ok 8 - Content contains "logged in (without address)"
ok 10 - Content contains "session variable set"
ok 12 - Content contains "VAR_logged=in"
# Looks like you failed 1 test of 12.
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/12 subtests

  Thu Sep 29 20:13:39 2011 wellsNOSPAM [...] cedarnet.org - Correspondence added
I'm wondering if the bug might be in the test script rather than the code… the other tests utilize multiple instances of Test::WWW::Mechanize::Catalyst my $ua1 = Test::WWW::Mechanize::Catalyst->new; my $ua2 = Test::WWW::Mechanize::Catalyst->new; I'm probably missing something but if you setup this test in much the same way, utilizing multiple instances, it passes without error. -------- my $ua1 = Test::WWW::Mechanize::Catalyst->new; my $ua2 = Test::WWW::Mechanize::Catalyst->new; # Initial Client $ua1->get_ok( "http://localhost/login" ); $ua1->content_contains('logged in'); $ua1->get_ok( "http://localhost/set_session_variable/logged/in" ); $ua1->content_contains('session variable set'); # Change Client $ua2->get_ok( "http://localhost/get_session_variable/logged"); $ua2->content_contains('VAR_logged=n.a.'); # Inital Client $ua1->get_ok( "http://localhost/login_without_address" ); $ua1->content_contains('logged in (without address)'); $ua1->get_ok( "http://localhost/set_session_variable/logged/in" ); $ua1->content_contains('session variable set'); # Change Client $ua1->get_ok( "http://localhost/get_session_variable/logged" ); $ua1->content_contains('VAR_logged=in'); ------- [just my $0.02] On Thu Sep 22 06:46:05 2011, KENTNL wrote: Show quoted text
> My suspicion is that $ENV{REMOTE_ADDRESS} is being ignored and is just > being > mapped to 127.0.0.1. > > I hacked up the Root.pm in a few places to report the address and it > does > appear to be doing just that. > > Will keep poking. > > t/live_verify_address.t .....1..12ok 1 - GET http://localhost/loginok > 2 - Content contains "logged in"ok 3 - GET > http://localhost/set_session_variable/logged/inok 4 - Content > contains "session variable set"ok 5 - GET > http://localhost/get_session_variable/loggednot ok 6 - Content > contains "VAR_logged=n.a." # Failed test 'Content contains > "VAR_logged=n.a."'# at t/live_verify_address.t line 41.# > searched: "VAR_logged=in"# can't find: "VAR_logged=n.a."# > LCSS: "VAR_logged="# LCSS context: "VAR_logged=in"[debug] > 127.0.0.1ok 7 - GET http://localhost/login_without_addressok 8 - > Content contains "logged in (without address)"ok 9 - GET > http://localhost/set_session_variable/logged/inok 10 - Content > contains "session variable set"ok 11 - GET > http://localhost/get_session_variable/loggedok 12 - Content > contains "VAR_logged=in"# Looks like you failed 1 test of > 12.Dubious, test returned 1 (wstat 256, 0x100)Failed 1/12 subtests
  Thu Sep 29 20:13:42 2011 The RT System itself - Status changed from 'new' to 'open'
  Fri Nov 18 09:16:14 2011 mail [...] andreasvoegele.com - Correspondence added
From: mail [...] andreasvoegele.com
On Thu Sep 22 06:46:05 2011, KENTNL wrote: Show quoted text
> My suspicion is that $ENV{REMOTE_ADDRESS} is being ignored and is just > being > mapped to 127.0.0.1.
REMOTE_ADDR is set to 127.0.0.1 by req_to_psgi() in HTTP::Message::PSGI, which is called by test_psgi() in Plack::Test::MockHTTP. If Plack::Test::MockHTTP is modified to put $ENV{REMOTE_ADDR} into the hash returned by to_psgi() test 6 from t/live_verify_address.t succeeds. Example: my $env = $req->to_psgi; $env->{REMOTE_ADDR} = $ENV{REMOTE_ADDR} if exists $ENV{REMOTE_ADDR}; I have yet to find a way to fix this without changing Plack::Test::MockHTTP.
  Fri Mar 30 16:37:51 2012 devin.austin [...] gmail.com - Correspondence added
On Thu Sep 22 06:46:05 2011, KENTNL wrote: Show quoted text
> My suspicion is that $ENV{REMOTE_ADDRESS} is being ignored and is just > being > mapped to 127.0.0.1. > > I hacked up the Root.pm in a few places to report the address and it > does > appear to be doing just that. > > Will keep poking. > > t/live_verify_address.t .....1..12ok 1 - GET http://localhost/loginok > 2 - Content contains "logged in"ok 3 - GET > http://localhost/set_session_variable/logged/inok 4 - Content > contains "session variable set"ok 5 - GET > http://localhost/get_session_variable/loggednot ok 6 - Content > contains "VAR_logged=n.a." # Failed test 'Content contains > "VAR_logged=n.a."'# at t/live_verify_address.t line 41.# > searched: "VAR_logged=in"# can't find: "VAR_logged=n.a."# > LCSS: "VAR_logged="# LCSS context: "VAR_logged=in"[debug] > 127.0.0.1ok 7 - GET http://localhost/login_without_addressok 8 - > Content contains "logged in (without address)"ok 9 - GET > http://localhost/set_session_variable/logged/inok 10 - Content > contains "session variable set"ok 11 - GET > http://localhost/get_session_variable/loggedok 12 - Content > contains "VAR_logged=in"# Looks like you failed 1 test of > 12.Dubious, test returned 1 (wstat 256, 0x100)Failed 1/12 subtests
I sort of combined both previous suggestions and created a branch (people/dhoss/fix-rt- 71142) with this fixed in it. I used Middleware::ForceEnv to change the REMOTE_ADDR variable, and tests now pass. Would love if someone could check this out.