xref: /linux/scripts/get_maintainer.pl (revision 5bdef865eb358b6f3760e25e591ae115e9eeddef)
1#!/usr/bin/perl -w
2# (c) 2007, Joe Perches <joe@perches.com>
3#           created from checkpatch.pl
4#
5# Print selected MAINTAINERS information for
6# the files modified in a patch or for a file
7#
8# usage: perl scripts/get_maintainers.pl [OPTIONS] <patch>
9#        perl scripts/get_maintainers.pl [OPTIONS] -f <file>
10#
11# Licensed under the terms of the GNU GPL License version 2
12
13use strict;
14
15my $P = $0;
16my $V = '0.16';
17
18use Getopt::Long qw(:config no_auto_abbrev);
19
20my $lk_path = "./";
21my $email = 1;
22my $email_usename = 1;
23my $email_maintainer = 1;
24my $email_list = 1;
25my $email_subscriber_list = 0;
26my $email_git = 1;
27my $email_git_penguin_chiefs = 0;
28my $email_git_min_signatures = 1;
29my $email_git_max_maintainers = 5;
30my $email_git_since = "1-year-ago";
31my $output_multiline = 1;
32my $output_separator = ", ";
33my $scm = 0;
34my $web = 0;
35my $subsystem = 0;
36my $status = 0;
37my $from_filename = 0;
38my $version = 0;
39my $help = 0;
40
41my $exit = 0;
42
43my @penguin_chief = ();
44push(@penguin_chief,"Linus Torvalds:torvalds\@linux-foundation.org");
45#Andrew wants in on most everything - 2009/01/14
46#push(@penguin_chief,"Andrew Morton:akpm\@linux-foundation.org");
47
48my @penguin_chief_names = ();
49foreach my $chief (@penguin_chief) {
50    if ($chief =~ m/^(.*):(.*)/) {
51	my $chief_name = $1;
52	my $chief_addr = $2;
53	push(@penguin_chief_names, $chief_name);
54    }
55}
56my $penguin_chiefs = "\(" . join("|",@penguin_chief_names) . "\)";
57
58# rfc822 email address - preloaded methods go here.
59my $rfc822_lwsp = "(?:(?:\\r\\n)?[ \\t])";
60my $rfc822_char = '[\\000-\\377]';
61
62if (!GetOptions(
63		'email!' => \$email,
64		'git!' => \$email_git,
65		'git-chief-penguins!' => \$email_git_penguin_chiefs,
66		'git-min-signatures=i' => \$email_git_min_signatures,
67		'git-max-maintainers=i' => \$email_git_max_maintainers,
68		'git-since=s' => \$email_git_since,
69		'm!' => \$email_maintainer,
70		'n!' => \$email_usename,
71		'l!' => \$email_list,
72		's!' => \$email_subscriber_list,
73		'multiline!' => \$output_multiline,
74		'separator=s' => \$output_separator,
75		'subsystem!' => \$subsystem,
76		'status!' => \$status,
77		'scm!' => \$scm,
78		'web!' => \$web,
79		'f|file' => \$from_filename,
80		'v|version' => \$version,
81		'h|help' => \$help,
82		)) {
83    usage();
84    die "$P: invalid argument\n";
85}
86
87if ($help != 0) {
88    usage();
89    exit 0;
90}
91
92if ($version != 0) {
93    print("${P} ${V}\n");
94    exit 0;
95}
96
97if ($#ARGV < 0) {
98    usage();
99    die "$P: argument missing: patchfile or -f file please\n";
100}
101
102my $selections = $email + $scm + $status + $subsystem + $web;
103if ($selections == 0) {
104    usage();
105    die "$P:  Missing required option: email, scm, status, subsystem or web\n";
106}
107
108if ($email && ($email_maintainer + $email_list + $email_subscriber_list
109	       + $email_git + $email_git_penguin_chiefs) == 0) {
110    usage();
111    die "$P: Please select at least 1 email option\n";
112}
113
114if (!top_of_kernel_tree($lk_path)) {
115    die "$P: The current directory does not appear to be "
116	. "a linux kernel source tree.\n";
117}
118
119## Read MAINTAINERS for type/value pairs
120
121my @typevalue = ();
122open(MAINT, "<${lk_path}MAINTAINERS") || die "$P: Can't open MAINTAINERS\n";
123while (<MAINT>) {
124    my $line = $_;
125
126    if ($line =~ m/^(\C):\s*(.*)/) {
127	my $type = $1;
128	my $value = $2;
129
130	##Filename pattern matching
131	if ($type eq "F" || $type eq "X") {
132	    $value =~ s@\.@\\\.@g;       ##Convert . to \.
133	    $value =~ s/\*/\.\*/g;       ##Convert * to .*
134	    $value =~ s/\?/\./g;         ##Convert ? to .
135	}
136	push(@typevalue, "$type:$value");
137    } elsif (!/^(\s)*$/) {
138	$line =~ s/\n$//g;
139	push(@typevalue, $line);
140    }
141}
142close(MAINT);
143
144## use the filenames on the command line or find the filenames in the patchfiles
145
146my @files = ();
147
148foreach my $file (@ARGV) {
149    next if ((-d $file));
150    if (!(-f $file)) {
151	die "$P: file '${file}' not found\n";
152    }
153    if ($from_filename) {
154	push(@files, $file);
155    } else {
156	my $file_cnt = @files;
157	open(PATCH, "<$file") or die "$P: Can't open ${file}\n";
158	while (<PATCH>) {
159	    if (m/^\+\+\+\s+(\S+)/) {
160		my $filename = $1;
161		$filename =~ s@^[^/]*/@@;
162		$filename =~ s@\n@@;
163		push(@files, $filename);
164	    }
165	}
166	close(PATCH);
167	if ($file_cnt == @files) {
168	    warn "$P: file '${file}' doesn't appear to be a patch.  "
169		. "Add -f to options?\n";
170	}
171	@files = sort_and_uniq(@files);
172    }
173}
174
175my @email_to = ();
176my @list_to = ();
177my @scm = ();
178my @web = ();
179my @subsystem = ();
180my @status = ();
181
182# Find responsible parties
183
184foreach my $file (@files) {
185
186#Do not match excluded file patterns
187
188    my $exclude = 0;
189    foreach my $line (@typevalue) {
190	if ($line =~ m/^(\C):\s*(.*)/) {
191	    my $type = $1;
192	    my $value = $2;
193	    if ($type eq 'X') {
194		if (file_match_pattern($file, $value)) {
195		    $exclude = 1;
196		}
197	    }
198	}
199    }
200
201    if (!$exclude) {
202	my $tvi = 0;
203	foreach my $line (@typevalue) {
204	    if ($line =~ m/^(\C):\s*(.*)/) {
205		my $type = $1;
206		my $value = $2;
207		if ($type eq 'F') {
208		    if (file_match_pattern($file, $value)) {
209			add_categories($tvi);
210		    }
211		}
212	    }
213	    $tvi++;
214	}
215    }
216
217    if ($email && $email_git) {
218	recent_git_signoffs($file);
219    }
220
221}
222
223if ($email) {
224    foreach my $chief (@penguin_chief) {
225	if ($chief =~ m/^(.*):(.*)/) {
226	    my $email_address;
227	    if ($email_usename) {
228		$email_address = format_email($1, $2);
229	    } else {
230		$email_address = $2;
231	    }
232	    if ($email_git_penguin_chiefs) {
233		push(@email_to, $email_address);
234	    } else {
235		@email_to = grep(!/${email_address}/, @email_to);
236	    }
237	}
238    }
239}
240
241if ($email || $email_list) {
242    my @to = ();
243    if ($email) {
244	@to = (@to, @email_to);
245    }
246    if ($email_list) {
247	@to = (@to, @list_to);
248    }
249    output(uniq(@to));
250}
251
252if ($scm) {
253    @scm = sort_and_uniq(@scm);
254    output(@scm);
255}
256
257if ($status) {
258    @status = sort_and_uniq(@status);
259    output(@status);
260}
261
262if ($subsystem) {
263    @subsystem = sort_and_uniq(@subsystem);
264    output(@subsystem);
265}
266
267if ($web) {
268    @web = sort_and_uniq(@web);
269    output(@web);
270}
271
272exit($exit);
273
274sub file_match_pattern {
275    my ($file, $pattern) = @_;
276    if (substr($pattern, -1) eq "/") {
277	if ($file =~ m@^$pattern@) {
278	    return 1;
279	}
280    } else {
281	if ($file =~ m@^$pattern@) {
282	    my $s1 = ($file =~ tr@/@@);
283	    my $s2 = ($pattern =~ tr@/@@);
284	    if ($s1 == $s2) {
285		return 1;
286	    }
287	}
288    }
289    return 0;
290}
291
292sub usage {
293    print <<EOT;
294usage: $P [options] patchfile
295       $P [options] -f file
296version: $V
297
298MAINTAINER field selection options:
299  --email => print email address(es) if any
300    --git => include recent git \*-by: signers
301    --git-chief-penguins => include ${penguin_chiefs}
302    --git-min-signatures => number of signatures required (default: 1)
303    --git-max-maintainers => maximum maintainers to add (default: 5)
304    --git-since => git history to use (default: 1-year-ago)
305    --m => include maintainer(s) if any
306    --n => include name 'Full Name <addr\@domain.tld>'
307    --l => include list(s) if any
308    --s => include subscriber only list(s) if any
309  --scm => print SCM tree(s) if any
310  --status => print status if any
311  --subsystem => print subsystem name if any
312  --web => print website(s) if any
313
314Output type options:
315  --separator [, ] => separator for multiple entries on 1 line
316  --multiline => print 1 entry per line
317
318Default options:
319  [--email --git --m --n --l --multiline]
320
321Other options:
322  --version => show version
323  --help => show this help information
324
325EOT
326}
327
328sub top_of_kernel_tree {
329	my ($lk_path) = @_;
330
331	if ($lk_path ne "" && substr($lk_path,length($lk_path)-1,1) ne "/") {
332	    $lk_path .= "/";
333	}
334	if (   (-f "${lk_path}COPYING")
335	    && (-f "${lk_path}CREDITS")
336	    && (-f "${lk_path}Kbuild")
337	    && (-f "${lk_path}MAINTAINERS")
338	    && (-f "${lk_path}Makefile")
339	    && (-f "${lk_path}README")
340	    && (-d "${lk_path}Documentation")
341	    && (-d "${lk_path}arch")
342	    && (-d "${lk_path}include")
343	    && (-d "${lk_path}drivers")
344	    && (-d "${lk_path}fs")
345	    && (-d "${lk_path}init")
346	    && (-d "${lk_path}ipc")
347	    && (-d "${lk_path}kernel")
348	    && (-d "${lk_path}lib")
349	    && (-d "${lk_path}scripts")) {
350		return 1;
351	}
352	return 0;
353}
354
355sub format_email {
356    my ($name, $email) = @_;
357
358    $name =~ s/^\s+|\s+$//g;
359    $name =~ s/^\"|\"$//g;
360    $email =~ s/^\s+|\s+$//g;
361
362    my $formatted_email = "";
363
364    if ($name =~ /[^a-z0-9 \.\-]/i) {    ##has "must quote" chars
365	$name =~ s/(?<!\\)"/\\"/g;       ##escape quotes
366	$formatted_email = "\"${name}\"\ \<${email}\>";
367    } else {
368	$formatted_email = "${name} \<${email}\>";
369    }
370    return $formatted_email;
371}
372
373sub add_categories {
374    my ($index) = @_;
375
376    $index = $index - 1;
377    while ($index >= 0) {
378	my $tv = $typevalue[$index];
379	if ($tv =~ m/^(\C):\s*(.*)/) {
380	    my $ptype = $1;
381	    my $pvalue = $2;
382	    if ($ptype eq "L") {
383		my $list_address = $pvalue;
384		my $list_additional = "";
385		if ($list_address =~ m/([^\s]+)\s+(.*)$/) {
386		    $list_address = $1;
387		    $list_additional = $2;
388		}
389		if ($list_additional =~ m/subscribers-only/) {
390		    if ($email_subscriber_list) {
391			push(@list_to, $list_address);
392		    }
393		} else {
394		    if ($email_list) {
395			push(@list_to, $list_address);
396		    }
397		}
398	    } elsif ($ptype eq "M") {
399		my $p_used = 0;
400		if ($index >= 0) {
401		    my $tv = $typevalue[$index - 1];
402		    if ($tv =~ m/^(\C):\s*(.*)/) {
403			if ($1 eq "P") {
404			    if ($email_usename) {
405				push_email_address(format_email($2, $pvalue));
406				$p_used = 1;
407			    }
408			}
409		    }
410		}
411		if (!$p_used) {
412		    push_email_addresses($pvalue);
413		}
414	    } elsif ($ptype eq "T") {
415		push(@scm, $pvalue);
416	    } elsif ($ptype eq "W") {
417		push(@web, $pvalue);
418	    } elsif ($ptype eq "S") {
419		push(@status, $pvalue);
420	    }
421
422	    $index--;
423	} else {
424	    push(@subsystem,$tv);
425	    $index = -1;
426	}
427    }
428}
429
430sub push_email_address {
431    my ($email_address) = @_;
432
433    my $email_name = "";
434    if ($email_address =~ m/([^<]+)<(.*\@.*)>$/) {
435	$email_name = $1;
436	$email_address = $2;
437    }
438
439    if ($email_maintainer) {
440	if ($email_usename && $email_name) {
441	    push(@email_to, format_email($email_name, $email_address));
442	} else {
443	    push(@email_to, $email_address);
444	}
445    }
446}
447
448sub push_email_addresses {
449    my ($address) = @_;
450
451    my @address_list = ();
452
453    if (rfc822_valid($address)) {
454	push_email_address($address);
455    } elsif (@address_list = rfc822_validlist($address)) {
456	my $array_count = shift(@address_list);
457	while (my $entry = shift(@address_list)) {
458	    push_email_address($entry);
459	}
460    } else {
461	warn("Invalid MAINTAINERS address: '" . $address . "'\n");
462    }
463}
464
465sub which {
466    my ($bin) = @_;
467
468    foreach my $path (split(/:/, $ENV{PATH})) {
469	if (-e "$path/$bin") {
470	    return "$path/$bin";
471	}
472    }
473
474    return "";
475}
476
477sub recent_git_signoffs {
478    my ($file) = @_;
479
480    my $sign_offs = "";
481    my $cmd = "";
482    my $output = "";
483    my $count = 0;
484    my @lines = ();
485
486    if (which("git") eq "") {
487	warn("$P: git not found.  Add --nogit to options?\n");
488	return;
489    }
490    if (!(-d ".git")) {
491	warn("$P: .git directory not found.  Use a git repository for better results.\n");
492	warn("$P: perhaps 'git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git'\n");
493	return;
494    }
495
496    $cmd = "git log --since=${email_git_since} -- ${file}";
497    $cmd .= " | grep -Ei \"^[-_ 	a-z]+by:.*\\\@.*\$\"";
498    if (!$email_git_penguin_chiefs) {
499	$cmd .= " | grep -Ev \"${penguin_chiefs}\"";
500    }
501    $cmd .= " | cut -f2- -d\":\"";
502    $cmd .= " | sort | uniq -c | sort -rn";
503
504    $output = `${cmd}`;
505    $output =~ s/^\s*//gm;
506
507    @lines = split("\n", $output);
508    foreach my $line (@lines) {
509	if ($line =~ m/([0-9]+)\s+(.*)/) {
510	    my $sign_offs = $1;
511	    $line = $2;
512	    $count++;
513	    if ($sign_offs < $email_git_min_signatures ||
514	        $count > $email_git_max_maintainers) {
515		last;
516	    }
517	} else {
518	    die("$P: Unexpected git output: ${line}\n");
519	}
520	if ($line =~ m/(.+)<(.+)>/) {
521	    my $git_name = $1;
522	    my $git_addr = $2;
523	    if ($email_usename) {
524		push(@email_to, format_email($git_name, $git_addr));
525	    } else {
526		push(@email_to, $git_addr);
527	    }
528	} elsif ($line =~ m/<(.+)>/) {
529	    my $git_addr = $1;
530	    push(@email_to, $git_addr);
531	} else {
532	    push(@email_to, $line);
533	}
534    }
535}
536
537sub uniq {
538    my @parms = @_;
539
540    my %saw;
541    @parms = grep(!$saw{$_}++, @parms);
542    return @parms;
543}
544
545sub sort_and_uniq {
546    my @parms = @_;
547
548    my %saw;
549    @parms = sort @parms;
550    @parms = grep(!$saw{$_}++, @parms);
551    return @parms;
552}
553
554sub output {
555    my @parms = @_;
556
557    if ($output_multiline) {
558	foreach my $line (@parms) {
559	    print("${line}\n");
560	}
561    } else {
562	print(join($output_separator, @parms));
563	print("\n");
564    }
565}
566
567my $rfc822re;
568
569sub make_rfc822re {
570#   Basic lexical tokens are specials, domain_literal, quoted_string, atom, and
571#   comment.  We must allow for rfc822_lwsp (or comments) after each of these.
572#   This regexp will only work on addresses which have had comments stripped
573#   and replaced with rfc822_lwsp.
574
575    my $specials = '()<>@,;:\\\\".\\[\\]';
576    my $controls = '\\000-\\037\\177';
577
578    my $dtext = "[^\\[\\]\\r\\\\]";
579    my $domain_literal = "\\[(?:$dtext|\\\\.)*\\]$rfc822_lwsp*";
580
581    my $quoted_string = "\"(?:[^\\\"\\r\\\\]|\\\\.|$rfc822_lwsp)*\"$rfc822_lwsp*";
582
583#   Use zero-width assertion to spot the limit of an atom.  A simple
584#   $rfc822_lwsp* causes the regexp engine to hang occasionally.
585    my $atom = "[^$specials $controls]+(?:$rfc822_lwsp+|\\Z|(?=[\\[\"$specials]))";
586    my $word = "(?:$atom|$quoted_string)";
587    my $localpart = "$word(?:\\.$rfc822_lwsp*$word)*";
588
589    my $sub_domain = "(?:$atom|$domain_literal)";
590    my $domain = "$sub_domain(?:\\.$rfc822_lwsp*$sub_domain)*";
591
592    my $addr_spec = "$localpart\@$rfc822_lwsp*$domain";
593
594    my $phrase = "$word*";
595    my $route = "(?:\@$domain(?:,\@$rfc822_lwsp*$domain)*:$rfc822_lwsp*)";
596    my $route_addr = "\\<$rfc822_lwsp*$route?$addr_spec\\>$rfc822_lwsp*";
597    my $mailbox = "(?:$addr_spec|$phrase$route_addr)";
598
599    my $group = "$phrase:$rfc822_lwsp*(?:$mailbox(?:,\\s*$mailbox)*)?;\\s*";
600    my $address = "(?:$mailbox|$group)";
601
602    return "$rfc822_lwsp*$address";
603}
604
605sub rfc822_strip_comments {
606    my $s = shift;
607#   Recursively remove comments, and replace with a single space.  The simpler
608#   regexps in the Email Addressing FAQ are imperfect - they will miss escaped
609#   chars in atoms, for example.
610
611    while ($s =~ s/^((?:[^"\\]|\\.)*
612                    (?:"(?:[^"\\]|\\.)*"(?:[^"\\]|\\.)*)*)
613                    \((?:[^()\\]|\\.)*\)/$1 /osx) {}
614    return $s;
615}
616
617#   valid: returns true if the parameter is an RFC822 valid address
618#
619sub rfc822_valid ($) {
620    my $s = rfc822_strip_comments(shift);
621
622    if (!$rfc822re) {
623        $rfc822re = make_rfc822re();
624    }
625
626    return $s =~ m/^$rfc822re$/so && $s =~ m/^$rfc822_char*$/;
627}
628
629#   validlist: In scalar context, returns true if the parameter is an RFC822
630#              valid list of addresses.
631#
632#              In list context, returns an empty list on failure (an invalid
633#              address was found); otherwise a list whose first element is the
634#              number of addresses found and whose remaining elements are the
635#              addresses.  This is needed to disambiguate failure (invalid)
636#              from success with no addresses found, because an empty string is
637#              a valid list.
638
639sub rfc822_validlist ($) {
640    my $s = rfc822_strip_comments(shift);
641
642    if (!$rfc822re) {
643        $rfc822re = make_rfc822re();
644    }
645    # * null list items are valid according to the RFC
646    # * the '1' business is to aid in distinguishing failure from no results
647
648    my @r;
649    if ($s =~ m/^(?:$rfc822re)?(?:,(?:$rfc822re)?)*$/so &&
650	$s =~ m/^$rfc822_char*$/) {
651        while ($s =~ m/(?:^|,$rfc822_lwsp*)($rfc822re)/gos) {
652            push @r, $1;
653        }
654        return wantarray ? (scalar(@r), @r) : 1;
655    }
656    else {
657        return wantarray ? () : 0;
658    }
659}
660