X-Git-Url: http://pileus.org/git/?a=blobdiff_plain;f=scripts%2Fget_maintainer.pl;h=d21ec3a89603b0fa8155671587deaebaa52b6163;hb=a5d3598de086cd38f71fe2fec43ac3ca97bb24bb;hp=61d3bb51bddf571d0c7a1dd3edb4bfe36b25860b;hpb=6ef1c52e122b675acc88a8b016d6477f67988b91;p=~andy%2Flinux

diff --git a/scripts/get_maintainer.pl b/scripts/get_maintainer.pl
index 61d3bb51bdd..d21ec3a8960 100755
--- a/scripts/get_maintainer.pl
+++ b/scripts/get_maintainer.pl
@@ -13,7 +13,7 @@
 use strict;
 
 my $P = $0;
-my $V = '0.26-beta4';
+my $V = '0.26-beta6';
 
 use Getopt::Long qw(:config no_auto_abbrev);
 
@@ -36,6 +36,7 @@ my $email_git_since = "1-year-ago";
 my $email_hg_since = "-365";
 my $interactive = 0;
 my $email_remove_duplicates = 1;
+my $email_use_mailmap = 1;
 my $output_multiline = 1;
 my $output_separator = ", ";
 my $output_roles = 0;
@@ -192,6 +193,7 @@ if (!GetOptions(
 		'hg-since=s' => \$email_hg_since,
 		'i|interactive!' => \$interactive,
 		'remove-duplicates!' => \$email_remove_duplicates,
+		'mailmap!' => \$email_use_mailmap,
 		'm!' => \$email_maintainer,
 		'n!' => \$email_usename,
 		'l!' => \$email_list,
@@ -295,31 +297,82 @@ while (<$maint>) {
 }
 close($maint);
 
-my %mailmap;
 
-if ($email_remove_duplicates) {
-    open(my $mailmap, '<', "${lk_path}.mailmap")
-	or warn "$P: Can't open .mailmap: $!\n";
-    while (<$mailmap>) {
-	my $line = $_;
+#
+# Read mail address map
+#
 
-	next if ($line =~ m/^\s*#/);
-	next if ($line =~ m/^\s*$/);
+my $mailmap;
 
-	my ($name, $address) = parse_email($line);
-	$line = format_email($name, $address, $email_usename);
+read_mailmap();
 
-	next if ($line =~ m/^\s*$/);
+sub read_mailmap {
+    $mailmap = {
+	names => {},
+	addresses => {}
+    };
 
-	if (exists($mailmap{$name})) {
-	    my $obj = $mailmap{$name};
-	    push(@$obj, $address);
-	} else {
-	    my @arr = ($address);
-	    $mailmap{$name} = \@arr;
+    return if (!$email_use_mailmap || !(-f "${lk_path}.mailmap"));
+
+    open(my $mailmap_file, '<', "${lk_path}.mailmap")
+	or warn "$P: Can't open .mailmap: $!\n";
+
+    while (<$mailmap_file>) {
+	s/#.*$//; #strip comments
+	s/^\s+|\s+$//g; #trim
+
+	next if (/^\s*$/); #skip empty lines
+	#entries have one of the following formats:
+	# name1 <mail1>
+	# <mail1> <mail2>
+	# name1 <mail1> <mail2>
+	# name1 <mail1> name2 <mail2>
+	# (see man git-shortlog)
+	if (/^(.+)<(.+)>$/) {
+	    my $real_name = $1;
+	    my $address = $2;
+
+	    $real_name =~ s/\s+$//;
+	    ($real_name, $address) = parse_email("$real_name <$address>");
+	    $mailmap->{names}->{$address} = $real_name;
+
+	} elsif (/^<([^\s]+)>\s*<([^\s]+)>$/) {
+	    my $real_address = $1;
+	    my $wrong_address = $2;
+
+	    $mailmap->{addresses}->{$wrong_address} = $real_address;
+
+	} elsif (/^(.+)<([^\s]+)>\s*<([^\s]+)>$/) {
+	    my $real_name = $1;
+	    my $real_address = $2;
+	    my $wrong_address = $3;
+
+	    $real_name =~ s/\s+$//;
+	    ($real_name, $real_address) =
+		parse_email("$real_name <$real_address>");
+	    $mailmap->{names}->{$wrong_address} = $real_name;
+	    $mailmap->{addresses}->{$wrong_address} = $real_address;
+
+	} elsif (/^(.+)<([^\s]+)>\s*([^\s].*)<([^\s]+)>$/) {
+	    my $real_name = $1;
+	    my $real_address = $2;
+	    my $wrong_name = $3;
+	    my $wrong_address = $4;
+
+	    $real_name =~ s/\s+$//;
+	    ($real_name, $real_address) =
+		parse_email("$real_name <$real_address>");
+
+	    $wrong_name =~ s/\s+$//;
+	    ($wrong_name, $wrong_address) =
+		parse_email("$wrong_name <$wrong_address>");
+
+	    my $wrong_email = format_email($wrong_name, $wrong_address, 1);
+	    $mailmap->{names}->{$wrong_email} = $real_name;
+	    $mailmap->{addresses}->{$wrong_email} = $real_address;
 	}
     }
-    close($mailmap);
+    close($mailmap_file);
 }
 
 ## use the filenames on the command line or find the filenames in the patchfiles
@@ -408,7 +461,8 @@ my @scm = ();
 my @web = ();
 my @subsystem = ();
 my @status = ();
-my @interactive_to = ();
+my %deduplicate_name_hash = ();
+my %deduplicate_address_hash = ();
 my $signature_pattern;
 
 my @maintainers = get_maintainers();
@@ -452,7 +506,8 @@ sub get_maintainers {
     @web = ();
     @subsystem = ();
     @status = ();
-    @interactive_to = ();
+    %deduplicate_name_hash = ();
+    %deduplicate_address_hash = ();
     if ($email_git_all_signature_types) {
 	$signature_pattern = "(.+?)[Bb][Yy]:";
     } else {
@@ -461,7 +516,7 @@ sub get_maintainers {
 
     # Find responsible parties
 
-    my %exact_pattern_match_hash;
+    my %exact_pattern_match_hash = ();
 
     foreach my $file (@files) {
 
@@ -545,7 +600,9 @@ sub get_maintainers {
 	}
     }
 
-    @interactive_to = (@email_to, @list_to);
+    foreach my $email (@email_to, @list_to) {
+	$email->[0] = deduplicate_email($email->[0]);
+    }
 
     foreach my $file (@files) {
 	if ($email &&
@@ -592,8 +649,7 @@ sub get_maintainers {
     }
 
     if ($interactive) {
-	@interactive_to = @to;
-	@to = interactive_get_maintainers(\@interactive_to);
+	@to = interactive_get_maintainers(\@to);
     }
 
     return @to;
@@ -657,8 +713,9 @@ Output type options:
 
 Other options:
   --pattern-depth => Number of pattern directory traversals (default: 0 (all))
-  --keywords => scan patch for keywords (default: 1 (on))
-  --sections => print the entire subsystem sections with pattern matches
+  --keywords => scan patch for keywords (default: $keywords)
+  --sections => print all of the subsystem sections with pattern matches
+  --mailmap => use .mailmap file (default: $email_use_mailmap)
   --version => show version
   --help => show this help information
 
@@ -698,30 +755,30 @@ EOT
 }
 
 sub top_of_kernel_tree {
-	my ($lk_path) = @_;
+    my ($lk_path) = @_;
 
-	if ($lk_path ne "" && substr($lk_path,length($lk_path)-1,1) ne "/") {
-	    $lk_path .= "/";
-	}
-	if (   (-f "${lk_path}COPYING")
-	    && (-f "${lk_path}CREDITS")
-	    && (-f "${lk_path}Kbuild")
-	    && (-f "${lk_path}MAINTAINERS")
-	    && (-f "${lk_path}Makefile")
-	    && (-f "${lk_path}README")
-	    && (-d "${lk_path}Documentation")
-	    && (-d "${lk_path}arch")
-	    && (-d "${lk_path}include")
-	    && (-d "${lk_path}drivers")
-	    && (-d "${lk_path}fs")
-	    && (-d "${lk_path}init")
-	    && (-d "${lk_path}ipc")
-	    && (-d "${lk_path}kernel")
-	    && (-d "${lk_path}lib")
-	    && (-d "${lk_path}scripts")) {
-		return 1;
-	}
-	return 0;
+    if ($lk_path ne "" && substr($lk_path,length($lk_path)-1,1) ne "/") {
+	$lk_path .= "/";
+    }
+    if (   (-f "${lk_path}COPYING")
+	&& (-f "${lk_path}CREDITS")
+	&& (-f "${lk_path}Kbuild")
+	&& (-f "${lk_path}MAINTAINERS")
+	&& (-f "${lk_path}Makefile")
+	&& (-f "${lk_path}README")
+	&& (-d "${lk_path}Documentation")
+	&& (-d "${lk_path}arch")
+	&& (-d "${lk_path}include")
+	&& (-d "${lk_path}drivers")
+	&& (-d "${lk_path}fs")
+	&& (-d "${lk_path}init")
+	&& (-d "${lk_path}ipc")
+	&& (-d "${lk_path}kernel")
+	&& (-d "${lk_path}lib")
+	&& (-d "${lk_path}scripts")) {
+	return 1;
+    }
+    return 0;
 }
 
 sub parse_email {
@@ -979,7 +1036,7 @@ sub push_email_address {
 	push(@email_to, [format_email($name, $address, $email_usename), $role]);
     } elsif (!email_inuse($name, $address)) {
 	push(@email_to, [format_email($name, $address, $email_usename), $role]);
-	$email_hash_name{lc($name)}++;
+	$email_hash_name{lc($name)}++ if ($name ne "");
 	$email_hash_address{lc($address)}++;
     }
 
@@ -1061,30 +1118,57 @@ sub which_conf {
     return "";
 }
 
-sub mailmap {
-    my (@lines) = @_;
-    my %hash;
+sub mailmap_email {
+    my ($line) = @_;
 
-    foreach my $line (@lines) {
-	my ($name, $address) = parse_email($line);
-	if (!exists($hash{$name})) {
-	    $hash{$name} = $address;
-	} elsif ($address ne $hash{$name}) {
-	    $address = $hash{$name};
-	    $line = format_email($name, $address, $email_usename);
+    my ($name, $address) = parse_email($line);
+    my $email = format_email($name, $address, 1);
+    my $real_name = $name;
+    my $real_address = $address;
+
+    if (exists $mailmap->{names}->{$email} ||
+	exists $mailmap->{addresses}->{$email}) {
+	if (exists $mailmap->{names}->{$email}) {
+	    $real_name = $mailmap->{names}->{$email};
 	}
-	if (exists($mailmap{$name})) {
-	    my $obj = $mailmap{$name};
-	    foreach my $map_address (@$obj) {
-		if (($map_address eq $address) &&
-		    ($map_address ne $hash{$name})) {
-		    $line = format_email($name, $hash{$name}, $email_usename);
-		}
-	    }
+	if (exists $mailmap->{addresses}->{$email}) {
+	    $real_address = $mailmap->{addresses}->{$email};
+	}
+    } else {
+	if (exists $mailmap->{names}->{$address}) {
+	    $real_name = $mailmap->{names}->{$address};
 	}
+	if (exists $mailmap->{addresses}->{$address}) {
+	    $real_address = $mailmap->{addresses}->{$address};
+	}
+    }
+    return format_email($real_name, $real_address, 1);
+}
+
+sub mailmap {
+    my (@addresses) = @_;
+
+    my @mapped_emails = ();
+    foreach my $line (@addresses) {
+	push(@mapped_emails, mailmap_email($line));
     }
+    merge_by_realname(@mapped_emails) if ($email_use_mailmap);
+    return @mapped_emails;
+}
 
-    return @lines;
+sub merge_by_realname {
+    my %address_map;
+    my (@emails) = @_;
+
+    foreach my $email (@emails) {
+	my ($name, $address) = parse_email($email);
+	if (exists $address_map{$name}) {
+	    $address = $address_map{$name};
+	    $email = format_email($name, $address, 1);
+	} else {
+	    $address_map{$name} = $address;
+	}
+    }
 }
 
 sub git_execute_cmd {
@@ -1121,8 +1205,7 @@ sub extract_formatted_signatures {
 ## Reformat email addresses (with names) to avoid badly written signatures
 
     foreach my $signer (@signature_lines) {
-	my ($name, $address) = parse_email($signer);
-	$signer = format_email($name, $address, 1);
+	$signer = deduplicate_email($signer);
     }
 
     return (\@type, \@signature_lines);
@@ -1266,6 +1349,7 @@ sub vcs_exists {
 }
 
 sub vcs_is_git {
+    vcs_exists();
     return $vcs_used == 1;
 }
 
@@ -1284,11 +1368,9 @@ sub interactive_get_maintainers {
     my %signed;
     my $count = 0;
     my $maintained = 0;
-    #select maintainers by default
     foreach my $entry (@list) {
-	my $role = $entry->[1];
-	$selected{$count} = ($role =~ /^(maintainer|supporter|open list)/i);
-	$maintained = 1 if ($role =~ /^(maintainer|supporter)/i);
+	$maintained = 1 if ($entry->[1] =~ /^(maintainer|supporter)/i);
+	$selected{$count} = 1;
 	$authored{$count} = 0;
 	$signed{$count} = 0;
 	$count++;
@@ -1345,24 +1427,34 @@ sub interactive_get_maintainers {
 	if ($print_options) {
 	    $print_options = 0;
 	    if (vcs_exists()) {
-		print STDERR
-"\nVersion Control options:\n" .
-"g  use git history      [$email_git]\n" .
-"gf use git-fallback     [$email_git_fallback]\n" .
-"b  use git blame        [$email_git_blame]\n" .
-"bs use blame signatures [$email_git_blame_signatures]\n" .
-"c# minimum commits      [$email_git_min_signatures]\n" .
-"%# min percent          [$email_git_min_percent]\n" .
-"d# history to use       [$$date_ref]\n" .
-"x# max maintainers      [$email_git_max_maintainers]\n" .
-"t  all signature types  [$email_git_all_signature_types]\n";
+		print STDERR <<EOT
+
+Version Control options:
+g  use git history      [$email_git]
+gf use git-fallback     [$email_git_fallback]
+b  use git blame        [$email_git_blame]
+bs use blame signatures [$email_git_blame_signatures]
+c# minimum commits      [$email_git_min_signatures]
+%# min percent          [$email_git_min_percent]
+d# history to use       [$$date_ref]
+x# max maintainers      [$email_git_max_maintainers]
+t  all signature types  [$email_git_all_signature_types]
+m  use .mailmap         [$email_use_mailmap]
+EOT
 	    }
-	    print STDERR "\nAdditional options:\n" .
-"0  toggle all\n" .
-"f  emails in file       [$file_emails]\n" .
-"k  keywords in file     [$keywords]\n" .
-"r  remove duplicates    [$email_remove_duplicates]\n" .
-"p# pattern match depth  [$pattern_depth]\n";
+	    print STDERR <<EOT
+
+Additional options:
+0  toggle all
+tm toggle maintainers
+tg toggle git entries
+tl toggle open list entries
+ts toggle subscriber list entries
+f  emails in file       [$file_emails]
+k  keywords in file     [$keywords]
+r  remove duplicates    [$email_remove_duplicates]
+p# pattern match depth  [$pattern_depth]
+EOT
 	}
 	print STDERR
 "\n#(toggle), A#(author), S#(signed) *(all), ^(none), O(options), Y(approve): ";
@@ -1398,6 +1490,28 @@ sub interactive_get_maintainers {
 		for (my $i = 0; $i < $count; $i++) {
 		    $selected{$i} = !$selected{$i};
 		}
+	    } elsif ($sel eq "t") {
+		if (lc($str) eq "m") {
+		    for (my $i = 0; $i < $count; $i++) {
+			$selected{$i} = !$selected{$i}
+			    if ($list[$i]->[1] =~ /^(maintainer|supporter)/i);
+		    }
+		} elsif (lc($str) eq "g") {
+		    for (my $i = 0; $i < $count; $i++) {
+			$selected{$i} = !$selected{$i}
+			    if ($list[$i]->[1] =~ /^(author|commit|signer)/i);
+		    }
+		} elsif (lc($str) eq "l") {
+		    for (my $i = 0; $i < $count; $i++) {
+			$selected{$i} = !$selected{$i}
+			    if ($list[$i]->[1] =~ /^(open list)/i);
+		    }
+		} elsif (lc($str) eq "s") {
+		    for (my $i = 0; $i < $count; $i++) {
+			$selected{$i} = !$selected{$i}
+			    if ($list[$i]->[1] =~ /^(subscriber list)/i);
+		    }
+		}
 	    } elsif ($sel eq "a") {
 		if ($val > 0 && $val <= $count) {
 		    $authored{$val - 1} = !$authored{$val - 1};
@@ -1466,6 +1580,10 @@ sub interactive_get_maintainers {
 	    } elsif ($sel eq "r") {
 		bool_invert(\$email_remove_duplicates);
 		$rerun = 1;
+	    } elsif ($sel eq "m") {
+		bool_invert(\$email_use_mailmap);
+		read_mailmap();
+		$rerun = 1;
 	    } elsif ($sel eq "k") {
 		bool_invert(\$keywords);
 		$rerun = 1;
@@ -1482,7 +1600,7 @@ commit signers and mailing lists that could be CC'd on a patch.
 
 Any *'d entry is selected.
 
-If you have git or hg installed, You can choose to summarize the commit
+If you have git or hg installed, you can choose to summarize the commit
 history of files in the patch.  Also, each line of the current file can
 be matched to its commit author and that commits signers with blame.
 
@@ -1529,6 +1647,36 @@ sub bool_invert {
     }
 }
 
+sub deduplicate_email {
+    my ($email) = @_;
+
+    my $matched = 0;
+    my ($name, $address) = parse_email($email);
+    $email = format_email($name, $address, 1);
+    $email = mailmap_email($email);
+
+    return $email if (!$email_remove_duplicates);
+
+    ($name, $address) = parse_email($email);
+
+    if ($name ne "" && $deduplicate_name_hash{lc($name)}) {
+	$name = $deduplicate_name_hash{lc($name)}->[0];
+	$address = $deduplicate_name_hash{lc($name)}->[1];
+	$matched = 1;
+    } elsif ($deduplicate_address_hash{lc($address)}) {
+	$name = $deduplicate_address_hash{lc($address)}->[0];
+	$address = $deduplicate_address_hash{lc($address)}->[1];
+	$matched = 1;
+    }
+    if (!$matched) {
+	$deduplicate_name_hash{lc($name)} = [ $name, $address ];
+	$deduplicate_address_hash{lc($address)} = [ $name, $address ];
+    }
+    $email = format_email($name, $address, 1);
+    $email = mailmap_email($email);
+    return $email;
+}
+
 sub save_commits_by_author {
     my (@lines) = @_;
 
@@ -1538,20 +1686,8 @@ sub save_commits_by_author {
 
     foreach my $line (@lines) {
 	if ($line =~ m/$VCS_cmds{"author_pattern"}/) {
-	    my $matched = 0;
 	    my $author = $1;
-	    my ($name, $address) = parse_email($author);
-	    foreach my $to (@interactive_to) {
-		my ($to_name, $to_address) = parse_email($to->[0]);
-		if ($email_remove_duplicates &&
-		    ((lc($name) eq lc($to_name)) ||
-		     (lc($address) eq lc($to_address)))) {
-		    $author = $to->[0];
-		    $matched = 1;
-		    last;
-		}
-	    }
-	    $author = format_email($name, $address, 1) if (!$matched);
+	    $author = deduplicate_email($author);
 	    push(@authors, $author);
 	}
 	push(@commits, $1) if ($line =~ m/$VCS_cmds{"commit_pattern"}/);
@@ -1592,19 +1728,7 @@ sub save_commits_by_signer {
 	    my $type = $types[0];
 	    my $signer = $signers[0];
 
-	    my $matched = 0;
-	    my ($name, $address) = parse_email($signer);
-	    foreach my $to (@interactive_to) {
-		my ($to_name, $to_address) = parse_email($to->[0]);
-		if ($email_remove_duplicates &&
-		    ((lc($name) eq lc($to_name)) ||
-		     (lc($address) eq lc($to_address)))) {
-		    $signer = $to->[0];
-		    $matched = 1;
-		    last;
-		}
-		$signer = format_email($name, $address, 1) if (!$matched);
-	    }
+	    $signer = deduplicate_email($signer);
 
 	    my $exists = 0;
 	    foreach my $ref(@{$commit_signer_hash{$signer}}) {
@@ -1636,9 +1760,7 @@ sub vcs_assign {
 	$divisor = 1;
     }
 
-    if ($email_remove_duplicates) {
-	@lines = mailmap(@lines);
-    }
+    @lines = mailmap(@lines);
 
     return if (@lines <= 0);
 
@@ -1680,6 +1802,11 @@ sub vcs_file_signoffs {
     $cmd =~ s/(\$\w+)/$1/eeg;		# interpolate $cmd
 
     ($commits, @signers) = vcs_find_signers($cmd);
+
+    foreach my $signer (@signers) {
+	$signer = deduplicate_email($signer);
+    }
+
     vcs_assign("commit_signer", $commits, @signers);
 }
 
@@ -1757,9 +1884,8 @@ sub vcs_file_blame {
 		foreach my $line (@lines) {
 		    if ($line =~ m/$VCS_cmds{"author_pattern"}/) {
 			my $author = $1;
-			my ($name, $address) = parse_email($author);
-			$author = format_email($name, $address, 1);
-			push(@authors, $1);
+			$author = deduplicate_email($author);
+			push(@authors, $author);
 		    }
 		}
 
@@ -1775,9 +1901,12 @@ sub vcs_file_blame {
 		    $cmd =~ s/(\$\w+)/$1/eeg;	#interpolate $cmd
 		    my @author = vcs_find_author($cmd);
 		    next if !@author;
+
+		    my $formatted_author = deduplicate_email($author[0]);
+
 		    my $count = grep(/$commit/, @all_commits);
 		    for ($i = 0; $i < $count ; $i++) {
-			push(@blame_signers, $author[0]);
+			push(@blame_signers, $formatted_author);
 		    }
 		}
 	    }
@@ -1785,8 +1914,14 @@ sub vcs_file_blame {
 		vcs_assign("authored lines", $total_lines, @blame_signers);
 	    }
 	}
+	foreach my $signer (@signers) {
+	    $signer = deduplicate_email($signer);
+	}
 	vcs_assign("commits", $total_commits, @signers);
     } else {
+	foreach my $signer (@signers) {
+	    $signer = deduplicate_email($signer);
+	}
 	vcs_assign("modified commits", $total_commits, @signers);
     }
 }