Skip Menu |

This queue is for tickets about the Find-Lib CPAN distribution.

Report information
The Basics
Id: 66049
Status: open
Priority: 0/
Queue: Find-Lib
People
Owner: Nobody in particular
Requestors: IKEGAMI [...] cpan.org
Cc:
AdminCc:
Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: (no value)

History Show all quoted text
  Tue Feb 22 17:10:57 2011 IKEGAMI [...] cpan.org - Ticket created
Subject: Find::Lib and symlinks
Find::Lib touts its handling of symlinks, but it fails to handle symlinks to the script. $ find ! -type d -exec ls -dl {} + lrwxrwxrwx 1 eric users 8 Feb 22 17:04 ./a.pl -> bin/a.pl -rwx------ 1 eric users 53 Feb 22 17:04 ./bin/a.pl -rw------- 1 eric users 16 Feb 22 17:03 ./lib/Mod.pm $ cat bin/a.pl #!/usr/bin/env perl use Find::Lib '../lib'; use Mod; $ cat lib/Mod.pm package Mod; 1; $ bin/a.pl $ a.pl Can't locate Mod.pm in @INC (@INC contains: /home/eric/usr/perlbrew/perls/perl-5.12.2/lib/site_perl/5.12.2/i686-linux /home/eric/usr/perlbrew/perls/perl-5.12.2/lib/site_perl/5.12.2 /home/eric/usr/perlbrew/perls/perl-5.12.2/lib/5.12.2/i686-linux /home/eric/usr/perlbrew/perls/perl-5.12.2/lib/5.12.2 .) at ./a.pl line 3. BEGIN failed--compilation aborted at ./a.pl line 3. $ perl -MFind::Lib -E'say Find::Lib->VERSION' 1.01
  Tue Feb 22 17:12:19 2011 IKEGAMI [...] cpan.org - Correspondence added
Workaround: use Cwd qw( realpath ); use File::Basename qw( dirname ); use lib dirname(realpath($0)).'/../lib';
  Tue Feb 22 17:47:03 2011 yann.kerherve [...] gmail.com - Correspondence added
I don't think I have a unit test when the binary itself is linked, only when the directories the binary is in, but it's been a while, I'll have to refresh my memory. The usecase is a little bit odd though, right?
  Tue Feb 22 17:47:04 2011 The RT System itself - Status changed from 'new' to 'open'
  Tue Feb 22 18:40:41 2011 ikegami [...] adaelis.com - Correspondence added
CC: IKEGAMI [...] cpan.org
Subject: Re: [rt.cpan.org #66049] Find::Lib and symlinks
Date: Tue, 22 Feb 2011 18:40:31 -0500
To: bug-Find-Lib [...] rt.cpan.org
From: Eric Brine <ikegami [...] adaelis.com>
On Tue, Feb 22, 2011 at 5:47 PM, Yann Kerherve via RT < bug-Find-Lib@rt.cpan.org> wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=66049 > > > I don't think I have a unit test when the binary itself is linked, only > when the directories the > binary is in, but it's been a while, I'll have to refresh my memory. > > The usecase is a little bit odd though, right? >
I wouldn't call linking to an executable odd at all. For example, /usr/bin/perl is a link (albeit a hard one).
  Wed Jun 29 14:10:16 2011 yann.kerherve [...] gmail.com - Correspondence added
If you symlink your a.pl then you 'lib' path will need to be offset too. Find::Lib uses $0 to interpret what the caller is trying to do. If you call a symlink Find::Lib interprets that as the new home execution environment and purposely doesn't resolve it to realpath().
  Wed Jun 29 14:10:16 2011 yann.kerherve [...] gmail.com - Status changed from 'open' to 'rejected'
  Wed Jun 29 15:42:00 2011 ikegami [...] adaelis.com - Correspondence added
CC: IKEGAMI [...] cpan.org
Subject: Re: [rt.cpan.org #66049] Find::Lib and symlinks
Date: Wed, 29 Jun 2011 15:41:50 -0400
To: bug-Find-Lib [...] rt.cpan.org
From: Eric Brine <ikegami [...] adaelis.com>
On Wed, Jun 29, 2011 at 2:10 PM, Yann Kerherve via RT < bug-Find-Lib@rt.cpan.org> wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=66049 > > > If you symlink your a.pl then you 'lib' path will need to be offset too. > Find::Lib uses $0 to > interpret what the caller is trying to do. If you call a symlink Find::Lib > interprets that as the new > home execution environment and purposely doesn't resolve it to realpath(). >
You're saying it's not acceptable to do ln -s ~/usr/app/bin/app ~/bin/app and that's just baffling to me.
  Wed Jun 29 15:42:01 2011 The RT System itself - Status changed from 'rejected' to 'open'
  Wed Jun 29 18:10:34 2011 ikegami [...] adaelis.com - Correspondence added
CC: IKEGAMI [...] cpan.org
Subject: Re: [rt.cpan.org #66049] Find::Lib and symlinks
Date: Wed, 29 Jun 2011 18:10:23 -0400
To: bug-Find-Lib [...] rt.cpan.org
From: Eric Brine <ikegami [...] adaelis.com>
On Wed, Jun 29, 2011 at 3:42 PM, ikegami@adaelis.com via RT < bug-Find-Lib@rt.cpan.org> wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=66049 > > > On Wed, Jun 29, 2011 at 2:10 PM, Yann Kerherve via RT < > bug-Find-Lib@rt.cpan.org> wrote: >
> > <URL: https://rt.cpan.org/Ticket/Display.html?id=66049 > > > > > If you symlink your a.pl then you 'lib' path will need to be offset too. > > Find::Lib uses $0 to > > interpret what the caller is trying to do. If you call a symlink
> Find::Lib
> > interprets that as the new > > home execution environment and purposely doesn't resolve it to
> realpath().
> >
> > You're saying it's not acceptable to do > > ln -s ~/usr/app/bin/app ~/bin/app > > and that's just baffling to me. >
If someone wants the behaviour you describe, they can already get it using either of ln ~/usr/app/bin/app ~/exec_env/bin/app and cp ~/usr/app/bin/app ~/exec_env/bin/app
  Fri Jul 01 00:45:00 2011 yann.kerherve [...] gmail.com - Correspondence added
I'm not saying you are wrong, I'm saying that it's two different use cases and that Find::Lib originally intended to solve the other one. And more importantly, it was meant to be intuitive for a script to refer to a lib relatively to $0 even if there are symlinks in this relative expression. I've tried what to solve for your problem, here is how it might look: https://github.com/yannk/findlib/tree/resolve What do you think?
  Wed Jul 13 04:18:53 2011 ikegami [...] adaelis.com - Correspondence added
CC: IKEGAMI [...] cpan.org
Subject: Re: [rt.cpan.org #66049] Find::Lib and symlinks
Date: Wed, 13 Jul 2011 04:18:36 -0400
To: bug-Find-Lib [...] rt.cpan.org
From: Eric Brine <ikegami [...] adaelis.com>
On Fri, Jul 1, 2011 at 12:45 AM, Yann Kerherve via RT < bug-Find-Lib@rt.cpan.org> wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=66049 > >
Show quoted text
> I've tried what to solve for your problem, here is how it might look: > https://github.com/yannk/findlib/tree/resolve > > What do you think? >
From looking at the code, it appears to do what I want. I haven't gotten back to you because I was going to include a patch for tests and documentation in my reply, but I got too many things on the go. Thanks, Eric