Skip to content
Create a gist now

Instantly share code, notes, and snippets.

Embed URL


Subversion checkout URL

You can clone with
Download ZIP
perlfind -- perldoc on steroids
#!/usr/bin/env perl
use strict;
use warnings;
use Carp qw(cluck);
use autodie ':all';
use Getopt::Long 2.33 qw(:config auto_help);
use File::Find::Rule;
use File::Basename 'basename';
use Term::ProgressBar;
$0 = basename($0);
my $perldoc = $ENV{PERLDOC} || 'perldoc';
verbose => \my $VERBOSE,
'perldoc=s' => \$perldoc,
'all' => \my $all,
) or die;
my $term = shift @ARGV || 'strict';
if ( my $doc = get_main_doc( $perldoc, $term ) ) {
exec $perldoc, $doc;
elsif ( is_func( $perldoc, $term ) ) {
exec $perldoc, '-f', $term;
elsif ( is_var( $perldoc, $term ) ) {
exec $perldoc, '-v', $term;
elsif ( is_faq( $perldoc, $term ) ) {
exec $perldoc, '-q', $term;
elsif ( my @files_with_count = do_grep( $all, $term ) ) {
foreach my $aref (@files_with_count) {
my ( $file, $count ) = @$aref;
print "$count hits: $file\n";
else {
warn "Could not find '$term'";
unless ($all) {
warn "You can try a brute force search with: $0 --all $term\n";
sub get_main_doc {
my ( $perldoc, $term ) = @_;
my @results = _exec( $perldoc, '-l', $term );
if ( @results > 1 ) {
my $results = join "\n" => @results;
cluck "Found more than one result for: $term\n\n$results\n";
return $results[0];
sub is_func {
my ( $perldoc, $term ) = @_;
my $is_func = _exec( $perldoc, '-f', $term );
return $is_func;
sub is_var {
my ( $perldoc, $term ) = @_;
my $is_var = _exec( $perldoc, '-v', "'$term'" );
return $is_var;
sub is_faq {
my ( $perldoc, $term ) = @_;
my $is_faq = _exec( $perldoc, '-q', "'$term'" );
return $is_faq;
sub do_grep {
my ( $all, $term ) = @_;
return unless $all;
warn "Could not find '$term'. Falling back to brute force search.";
my @paths = @INC, map { split /:/ } $ENV{PERL5LIB};
if ( $VERBOSE ) {
my $locations = join "\n\t" => "Searching in:", @paths;
warn "$locations\n";
my @files = File::Find::Rule->file->name('*.pod')->in(@paths);
my $count = @files;
if ( $VERBOSE ) {
warn "Found $count matching files\n";
my @files_with_count;
my $progress = Term::ProgressBar->new({count => scalar @files});
my $num_searched = 0;
foreach my $file (@files) {
my @count = _exec( 'grep', '-c', $term, $file );
if ( $count[0] ) {
push @files_with_count => [ $file, $count[0] ];
@files_with_count = sort { $b->[1] <=> $a->[1] } @files_with_count;
return @files_with_count;
sub _exec {
my @command = @_;
if ($VERBOSE) {
warn "Executing: @command\n";
chomp( my @results = qx(@command 2>/dev/null) );
return @results;
=head1 NAME
perlfind - perldoc on steroids
perlfind '$@'
perlfind Scalar::Util
perlfind file
perlfind die
perlfind __DATA__ --all
Tired of C<perldoc>? Try C<perlfind> I<anything>. It will return the first
matching perldoc document found for I<anything>, in the following precedence
=over 4
=item 1. C<perldoc MODULE>
=item 2. C<perldoc -f FUNCTION>
=item 3. C<perldoc -v VARIABLE>
=item 4. C<perldoc -q FAQKEYWORD>
=item 5. A brute force grep of C<@INC> and C<$ENV{PERL5LIB>.
Note that the brute force grep requires L</Term::ProgressBar> and
L</File::Find::Rule>. You must also specify the C<--all> option.
=head1 OPTIONS
--perldoc=/path/to/perldoc Force an explicit path to your perldoc
--verbose Show how we're searching
--all Fall back to brute force if we fail
=head1 CAVEATS
It's a hack.
=head1 BUGS
=head1 AUTHOR
Curtis "Ovid" Poe
=head1 LICENSE
Copyright (c) 2012 Curtis "Ovid" Poe ( All rights reserved.
This module is free software; you can redistribute it and/or modify it under
the same terms as Perl itself.
This program is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A

I made a minor change in my fork. perl unshifts PERL5LIB onto @INC; the original do_grep() was treating PERL5LIB as though it were pushed onto @INC.


Actually, checking PERL5LIB in addition to @INC seems redundant. If $ENV{PERL5LIB} is set, it's already at the beginning of @INC. Adding it on to @INC, whether at the beginning or the end, just means that those directories will get checked twice. I've updated my gist.


If you use it from within vim, you might want it add namespace backtracking, like my hack: p5doc.


perlfind --all '$@'

a bug?


vfilatov: the --all switch only falls back to the in-depth search if it doesn't find the argument in the normal docs. So I would argue that it's another feature I should add :)


You mean you don't get the script frozen?

% perlfind --all '$@'
Could not find '$@'. Falling back to brute force search. at /usr/local/bin/perlfind line 83.
0% [ ]^C


Nope, it doesn't freeze for me. What version of Perl are you using?


Actually, it sometimes freezes for me. This is very strange!


I don't have the time to debug this right now, but it looks like File::Find::Rule is hanging.


I've just updated the gist above. Grab the new copy and run it with --verbose.


./ --verbose --all '$@'
Executing: perldoc -l $@
Executing: perldoc -f $@
Executing: perldoc -v '$@'
Executing: perldoc -q '$@'
Could not find '$@'. Falling back to brute force search. at ./ line 83.
Use of uninitialized value $_ in split at ./ line 85.
Searching in:
Found 282 matching files
0% [ ]Executing: grep -c $@ /usr/local/lib/perl/5.10.1/perllocal.pod


looks like 'grep' does not get a pattern to search

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.