public release

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

diff --git a/git-grep-footer b/git-grep-footer
index ff279a606bc2ad4cc8545b1e2e14f039b23550ce..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;
 use 5.010;
 use strict;
 use warnings;


@@ -7,58 +7,86 @@ use Encode 'decode';
 use Data::Dump 'pp';
 use Getopt::Long qw(:config bundling);
 
 use Data::Dump 'pp';
 use Getopt::Long qw(:config bundling);
 

+our $VERSION = '1.00';
+

 GetOptions(\my %opt,
        'debug!',
 GetOptions(\my %opt,
        'debug!',

+       '',  # stdin
+       'count|c!',

        'simplify|s:s',
        'ignore-case|i!',
        'simplify|s:s',
        'ignore-case|i!',

+       'fuzzy!',
+       'grep|S=s',

        'min|min-count|unique|u:i',
        'max|max-count|show|n:i',
        '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;
 
        '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";
 
 local $| = 1;
 local $/ = "\0";
 

-my $HEADERMATCH = qr/ [a-z]+ (?: (?:-\w+)+ | \ by ) /ix;
+my $HEADERMATCH = qr/ [a-z]+ (?: (?:-\w+)+ | \ by ) | cc | reference /imsx;
+
+my (%headercount, @headercache);

 
 

-while (readline) {
-       s/(.+)\n//m;
-       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
 
        # 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
        # if invalid, assume it's latin1

-              $_ = decode(cp1252 => $_) if $@;
-
-       my $prefix = 0;
-       my %attr;
+           or $_ = decode(cp1252 => $_);

 
        BLOCK:
        for (reverse split /\n\n/) {
                my @headers;
 
        BLOCK:
        for (reverse split /\n\n/) {
                my @headers;

+               my $prefix = 0;

 
                LINE:
                for (split /\n/) {
 
                LINE:
                for (split /\n/) {

-                       next if not /\S/;
+                       next if not m/\S/;

                        my @header = m{
                                ^
                                (?<key> $HEADERMATCH)
                                : \s*
                        my @header = m{
                                ^
                                (?<key> $HEADERMATCH)
                                : \s*

-                               (?<val> \S .+)
+                               (?<val> \S [^\n]+)

                                $
                                $

-                       }imx or do {
+                       }imsx or do {

                                $prefix++;
                                next LINE;
                        };
 
                        push @header, $_ if defined $opt{max};
 
                                $prefix++;
                                next LINE;
                        };
 
                        push @header, $_ if defined $opt{max};
 

+                       if ($opt{fuzzy}) {
+                               for ($header[0]) {
+                                       tr/ _/-/;
+
+                                       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']) {
                                        $header[1] =~ s{
                        given ($opt{simplify} // 'none') {
                                when (['email', 'authors']) {
                                        $header[1] =~ s{


@@ -69,14 +97,15 @@ while (readline) {
                                        }{<...>}imsx;
                                }
                                when (['var', 'vars', '']) {
                                        }{<...>}imsx;
                                }
                                when (['var', 'vars', '']) {

-                                       when ($header[0] =~ /[ _-] (?: by | to ) $/imsx) {
+                                       when ($header[0] =~ m/[ _-] (?: by | to ) $ | ^cc$/imsx) {

                                                $header[1] = undef;
                                        }
                                        for ($header[1]) {
                                                s{\b (https?)://\S+ }{[$1]}gmsx;  # url
                                                s{(?: < | \A ) [^@>\s]+ @ [^>]+ (?: > | \Z )}{<...>}igmsx;  # address
                                                s{\b [0-9]+ \b}{[num]}gmsx;  # number
                                                $header[1] = undef;
                                        }
                                        for ($header[1]) {
                                                s{\b (https?)://\S+ }{[$1]}gmsx;  # url
                                                s{(?: < | \A ) [^@>\s]+ @ [^>]+ (?: > | \Z )}{<...>}igmsx;  # address
                                                s{\b [0-9]+ \b}{[num]}gmsx;  # number

-                                               s{\b I? [0-9a-f]{40} \b}{[sha1]}gmsx;  # hash
+                                               s{\b [Ig]? [0-9a-f]{  40} \b}{[sha1]}gmsx;  # hash
+                                               s{\b [Ig]? [0-9a-f]{6,40} \b}{[hash]}gmsx;  # abbrev

                                        }
                                }
                                when (['all', 'contents']) {
                                        }
                                }
                                when (['all', 'contents']) {


@@ -101,23 +130,34 @@ while (readline) {
                next BLOCK if not @headers;
 
                if ($opt{debug} and $prefix) {
                next BLOCK if not @headers;
 
                if ($opt{debug} and $prefix) {

-                       say "infix junk in commit $hash";
+                       say sprintf ': invalid lines in %s (%s)', $hash // 'block', $prefix;

                }
 
                for (@headers) {
                }
 
                for (@headers) {

-                       if (defined $opt{min} or $opt{max}) {
-                               state $seen;
-                               my $count = $seen->{ $_->[0] }->{ $_->[1] // '' }++;
-                               next if $count >= ($opt{min} // 0) + ($opt{max} || 1);
-                               next if $count < ($opt{min} // 0);
+                       my $line = $_->[2] // join(': ', @$_);
+                       $line =~ s/\A/$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);
+                               next if $excess >= ($opt{max} || 1);
+                               next if $excess <  0;
+                               if ($opt{count}) {
+                                       push @headercache, [ $line, $excess ? \undef : $counter ];
+                                       next;
+                               }

                        }
                        }

-                       say $_->[2] // join(': ', @$_);
+                       say $line;

                }
 
                last BLOCK;
        }
 }
 
                }
 
                last BLOCK;
        }
 }
 

+for (@headercache) {
+       say ${$_->[1]} // '', "\t", $_->[0];
+}
+

 __END__
 
 =head1 NAME
 __END__
 
 =head1 NAME


@@ -126,7 +166,9 @@ git-grep-footer - Find custom header lines in commit messages
 
 =head1 SYNOPSIS
 
 
 =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
 
 
 =head1 DESCRIPTION
 


@@ -135,6 +177,7 @@ a common convention to list custom metadata such as
 C<Signed-off-by> and C<Acked-by>.
 
 Sections are identified by at least one leading keyword containing a dash
 C<Signed-off-by> and C<Acked-by>.
 
 Sections are identified by at least one leading keyword containing a dash

+(or exceptionally recognised)

 followed by a colon.
 
 =head1 OPTIONS
 followed by a colon.
 
 =head1 OPTIONS


@@ -171,6 +214,11 @@ Values will be hidden entirely, so only attribute names remain.
 
 =back
 
 
 =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,
 =item -u, --unique[=<threshold>]
 
 Each match is only shown once,


@@ -182,6 +230,42 @@ The original line is given for each match,
 but simplifications still apply for duplicate determination.
 Additional samples are optionally given upto the given maximum.
 
 but simplifications still apply for duplicate determination.
 Additional samples are optionally given upto the given maximum.
 

+=item -c, --count
+
+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
 =back
 
 =head1 AUTHOR


@@ -190,5 +274,7 @@ Mischa POSLAWSKY <perl@shiar.org>
 
 =head1 LICENSE
 
 
 =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.