public release

[git-grep-footer.git] / git-grep-footer

diff --git a/git-grep-footer b/git-grep-footer
index b676ab8aef983e241ffb59921849d94075dd889d..97fbca7748b374a295e1758f43887ce6f14de8ef 100755 (executable)
--- a/git-grep-footer
+++ b/git-grep-footer
@@ -1,4 +1,4 @@
-#!/usr/bin/perl
+#!/usr/bin/env perl
 use 5.010;
 use strict;
 use warnings;
@@ -7,40 +7,48 @@ use Encode 'decode';
 use Data::Dump 'pp';
 use Getopt::Long qw(:config bundling);
 
+our $VERSION = '1.00';
+
 GetOptions(\my %opt,
        'debug!',
+       '',  # stdin
        'count|c!',
        'simplify|s:s',
        'ignore-case|i!',
        'fuzzy!',
+       'grep|S=s',
        'min|min-count|unique|u:i',
        'max|max-count|show|n:i',
+       'hash|H!',
        'version|V'  => sub { Getopt::Long::VersionMessage() },
        'usage|h'    => sub { Getopt::Long::HelpMessage() },
        'help|man|?' => sub { Getopt::Long::HelpMessage(-verbose => 2) },
 ) or exit 129;
 
+my $inputstream = $opt{''} ? \*ARGV : eval {
+       require Git;
+       Git::command_output_pipe('log', '-z', '--pretty=format:%h%n%b', @ARGV);
+} || die "Automatic git log failed: $@";
+
 local $| = 1;
 local $/ = "\0";
 
-my $HEADERMATCH = qr/ [a-z]+ (?: (?:-\w+)+ | \ by ) | cc | reference /ix;
+my $HEADERMATCH = qr/ [a-z]+ (?: (?:-\w+)+ | \ by ) | cc | reference /imsx;
 
 my (%headercount, @headercache);
 
-while (readline) {
-       s/^([0-9a-f]{4,40})\n//m and
-       my $hash = $1;
+while (readline $inputstream) {
+       s/^ ([0-9a-f]{4,40}) \n//msx;
+       my $hash = $opt{hash} ? $1 : undef;
 
        # strip commit seperator
        chomp;
        # skip expensive checks without potential identifier
        m/:/ or next;
        # try to parse as UTF-8
-       eval { $_ = decode(utf8   => $_, Encode::FB_CROAK()) };
+       eval { $_ = decode(utf8   => $_, Encode::FB_CROAK()); return 1 }
        # if invalid, assume it's latin1
-              $_ = decode(cp1252 => $_) if $@;
-
-       my %attr;
+           or $_ = decode(cp1252 => $_);
 
        BLOCK:
        for (reverse split /\n\n/) {
@@ -49,14 +57,14 @@ while (readline) {
 
                LINE:
                for (split /\n/) {
-                       next if not /\S/;
+                       next if not m/\S/;
                        my @header = m{
                                ^
                                (?<key> $HEADERMATCH)
                                : \s*
-                               (?<val> \S .+)
+                               (?<val> \S [^\n]+)
                                $
-                       }imx or do {
+                       }imsx or do {
                                $prefix++;
                                next LINE;
                        };
@@ -67,14 +75,18 @@ while (readline) {
                                for ($header[0]) {
                                        tr/ _/-/;
 
-                                       state $BY = qr{ (?: -? b[yu] )? \Z }ix;
-                                       s{^ si (?:ge?n|n?g) (?:e?[dt])? -? (?:of+)? $BY}{Signed-off-by}ix;
-                                       s{^ ack (?:ed|de)?  $BY}{Acked-by}ix;
-                                       s{^ review (?:e?d)? $BY}{Reviewed-by}ix;
-                                       s{^ teste[dt]       $BY}{Tested-by}ix;
+                                       state $BY = qr{ (?: -? b[yu] )? \Z }imsx;
+                                       s{\A si (?:ge?n|n?g) (?:e?[dt])? -? (?:of+)? $BY}{Signed-off-by}imsx;
+                                       s{\A ack (?:ed|de)?  $BY}{Acked-by}imsx;
+                                       s{\A review (?:e?d)? $BY}{Reviewed-by}imsx;
+                                       s{\A teste[dt]       $BY}{Tested-by}imsx;
                                }
                        }
 
+                       if (defined $opt{grep}) {
+                               $_ ~~ qr/$opt{grep}/im or next LINE;
+                       }
+
                        given ($opt{simplify} // 'none') {
                                when (['email', 'authors']) {
                                        $header[1] =~ s{
@@ -85,7 +97,7 @@ while (readline) {
                                        }{<...>}imsx;
                                }
                                when (['var', 'vars', '']) {
-                                       when ($header[0] =~ /[ _-] (?: by | to ) $ | ^cc$/imsx) {
+                                       when ($header[0] =~ m/[ _-] (?: by | to ) $ | ^cc$/imsx) {
                                                $header[1] = undef;
                                        }
                                        for ($header[1]) {
@@ -123,11 +135,11 @@ while (readline) {
 
                for (@headers) {
                        my $line = $_->[2] // join(': ', @$_);
-                       $line =~ s/^/$hash / if defined $hash;
+                       $line =~ s/\A/$hash /msx if defined $hash;
 
                        if (defined $opt{min} or $opt{max} or $opt{count}) {
                                my $counter = \$headercount{ $_->[0] }->{ $_->[1] // '' };
-                               my $excess = $$counter++ - ($opt{min} // 0);
+                               my $excess = ${$counter}++ - ($opt{min} // 0);
                                next if $excess >= ($opt{max} || 1);
                                next if $excess <  0;
                                if ($opt{count}) {
@@ -154,7 +166,9 @@ git-grep-footer - Find custom header lines in commit messages
 
 =head1 SYNOPSIS
 
-F<git> log --pretty=%b%x00 | F<git-grep-footer> [OPTIONS]
+F<git-grep-footer> [OPTIONS] [-- <git log options>]
+
+F<git> log -z --pretty=format:%b | F<git-grep-footer> [OPTIONS] -
 
 =head1 DESCRIPTION
 
@@ -200,6 +214,11 @@ Values will be hidden entirely, so only attribute names remain.
 
 =back
 
+=item --grep=<pattern>
+
+Only include lines matching the specified regular expression.
+Case insensitivity can be disabled by prepending C<(?-i)>.
+
 =item -u, --unique[=<threshold>]
 
 Each match is only shown once,
@@ -216,6 +235,37 @@ Additional samples are optionally given upto the given maximum.
 Prefixes (unique) lines by the number of occurrences.
 Causes output to be buffered until all input has been read (obviously).
 
+=item -H, --hash
+
+Prefixes the SHA1 hash of the (or a) matching commit.
+
+=back
+
+=head1 EXAMPLES
+
+=over
+
+=item git-grep-footer --grep=^ack v2.6.32..v2.6.33
+
+Search for I<Acked-by> lines for version I<v2.6.33>.
+Append C<-uin> to skip reoccurrences.
+
+=item git-grep-footer -u --grep=junio
+
+Show distinct lines mentioning a specific author.
+
+=item git-grep-footer -c --simplify --grep=^si
+
+Compare various capitalisations and (mis)spellings of signoffs.
+
+=item git-grep-footer -c --simplify=all -i | sort -n -r | head -n10
+
+List the ten most frequently used attribute names.
+
+=item git-grep-footer -n2 -i -s --hash -- --reverse
+
+The earliest two usages of each distinct identifier.
+
 =back
 
 =head1 AUTHOR
@@ -224,5 +274,7 @@ Mischa POSLAWSKY <perl@shiar.org>
 
 =head1 LICENSE
 
-Copyright. All rights reserved.
+This software is free software;
+you can redistribute and/or modify it under the terms of the GNU GPL
+version 2 or later.