How to keep track of INC hash?

jacques has asked for the wisdom of the Perl Monks concerning the following question:

Please read carefully. This question is tricky.
At the beginning of my module I examine %INC with a BEGIN block, like so:

BEGIN { %Seen = %INC }
[download]

That tells me what %INC looks like before I 'use' anything in my module. In other words, it tells me what files, if any, were loaded before my own.

What I would like to know are the entries made to %INC by my module and *only* by my module. So if my module doesn't use CGI.pm and a script is using my module like so:

use My::Module;
use CGI;
[download]

I could tell that the CGI module was used by the script and not by my module. How could I figure this out from within my module? I tried using an END block to no avail.

Comment on How to keep track of INC hash? Select or Download Code

Replies are listed 'Best First'.
Re: How to keep track of INC hash? by grinder (Bishop) on Jul 15, 2008 at 05:27 UTC
What is preventing you from installing a handler in @INC? `perl -le 'BEGIN{unshift @INC, sub {print $_[1]}} use CGI; # ... produces CGI.pm Carp.pm Exporter.pm CGI/Util.pm strict.pm vars.pm warnings/register.pm warnings.pm constant.pm overload.pm` [download] It is then a simple matter of programming to figure out if your module is the one doing the loading. Or maybe not, if that's why you're asking the question. Is there something that rules out this approach? • another intruder with the mooring in the heart of the Perl	[reply] [d/l]
Re^2: How to keep track of INC hash? by massa (Hermit) on Jul 15, 2008 at 10:32 UTC
Worked for me, too... I did: `$ perl -e ' BEGIN { open my $f, "\|sort>/tmp/uses.txt"; unshift @INC, sub { push @_, caller; print $f "module $_[2] uses $_[1]\n" } } do "bin/ponto" '` [download] where `bin/ponto` is the name of some script of mine, and I got the following in `/tmp/uses.txt`: `module base uses vars.pm module Carp uses Exporter.pm module Compress::Zlib uses Compress/Raw/Zlib.pm module Compress::Zlib uses IO/Compress/Base/Common.pm` [download] ... `module main uses bin/ponto module main uses Date/Manip.pm module main uses locale.pm module main uses open.pm module main uses strict.pm module main uses utf8.pm module main uses warnings.pm module main uses WWW/Mechanize.pm module main uses YAML/Syck.pm` [download] ... []s, HTH, Massa	[reply] [d/l] [select]
Re^2: How to keep track of INC hash? by jacques (Priest) on Jul 15, 2008 at 19:08 UTC
`++` Your response answered my question as posed. Upon further reflection, I realized that what I really want is to prevent my module from making additions to INC in the first place. Is there a way to do that? Basically what I want to find out is the contents of the INC hash from a script that uses my module (but my module's dependencies should not be included in that data, unless those dependencies were used elsewhere in the program). Perhaps I could delete those particular additions to INC in an END block?	[reply]
Re^3: How to keep track of INC hash? by grinder (Bishop) on Jul 15, 2008 at 20:16 UTC
I really want is to prevent my module from making additions to INC in the first place In other words, you don't want your module to load additional packages? In that case, return `undef` from your filter. See require for more information (I think this is the crux of matter in the wisdom you seek). • another intruder with the mooring in the heart of the Perl	[reply]
Re: How to keep track of INC hash? by GrandFather (Saint) on Jul 15, 2008 at 04:46 UTC
Noname1.pm: `package Noname1; use strict; use Warnings; my @added; BEGIN { my %Seen = %INC; require CGI; @added = grep {! exists $Seen{$_}} keys %INC; } sub Added { return @added; } 1;` [download] Noname.pl: `use strict; use Warnings; use Noname1; use CGI; my @added = Noname1::Added (); print "@added";` [download] Prints: `CGI/Util.pm CGI.pm` [download] Perl is environmentally friendly - it saves trees	[reply] [d/l] [select]