tools/fix-headers.pl

   1 #!/usr/bin/perl
   2 # ------------------------------------------------------------
   3 # This script fixes the license headers of all Java sources
   4 # to use the Eclipse EDL license template and updates the
   5 # copyright statements using author information from git blame
   6 #
   7 # To fix this in all revisions rewrite the history
   8 # git filter-branch --tree-filter 'fixHeaders.pl' HEAD
   9 # ------------------------------------------------------------
  10 use strict;
  11
  12 # Table of author names, start date, end date, actual copyright owner.
  13 #
  14 my @author_employers = (
  15         [ qr/spearce\@spearce.org/, 2008, 8, 9999, 12, 'Google Inc.'],
  16
  17         [ qr/\@(.*\.|)google.com/, 0, 0, 9999, 12, 'Google Inc.'],
  18 );
  19
  20 # License text itself.
  21 #
  22 my $license_text = <<'EOF';
  23  and other copyright owners as documented in the project's IP log.
  24
  25  This program and the accompanying materials are made available
  26  under the terms of the Eclipse Distribution License v1.0 which
  27  accompanies this distribution, is reproduced below, and is
  28  available at http://www.eclipse.org/org/documents/edl-v10.php
  29
  30  All rights reserved.
  31
  32  Redistribution and use in source and binary forms, with or
  33  without modification, are permitted provided that the following
  34  conditions are met:
  35
  36  - Redistributions of source code must retain the above copyright
  37    notice, this list of conditions and the following disclaimer.
  38
  39  - Redistributions in binary form must reproduce the above
  40    copyright notice, this list of conditions and the following
  41    disclaimer in the documentation and/or other materials provided
  42    with the distribution.
  43
  44  - Neither the name of the Eclipse Foundation, Inc. nor the
  45    names of its contributors may be used to endorse or promote
  46    products derived from this software without specific prior
  47    written permission.
  48
  49  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
  50  CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
  51  INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
  52  OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  53  ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
  54  CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
  55  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
  56  NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
  57  LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  58  CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
  59  STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
  60  ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
  61  ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  62 EOF
  63
  64 my @files = @ARGV;
  65 unless (@files) {
  66         open( F, '-|', 'git ls-files' );
  67         @files = <F>;
  68         chop @files;
  69         close F;
  70 }
  71
  72 foreach (@files) {
  73         if (/\.java$/ || $_ eq 'LICENSE') {
  74                 next if $_ eq 'org.eclipse.jgit/src/org/eclipse/jgit/util/Base64.java';
  75                 update_file(\&java_file, $_);
  76
  77         } elsif (/pom\.xml$/) {
  78                 update_file(\&pom_file, $_);
  79
  80         } elsif (/\.sh$/) {
  81                 update_file(\&sh_file, $_);
  82         }
  83 }
  84
  85 sub java_file
  86 {
  87         my $fd = shift;
  88         my $header = '';
  89         my $preamble = '';
  90
  91         # header is everything before package statement
  92         while (<$fd>) {
  93                 if (/^package /) {
  94                         $preamble = $_;
  95                         last;
  96                 }
  97                 $header .= $_;
  98         }
  99
 100         # preamble is everything with blanks or imports
 101         while (<$fd>) {
 102                 $preamble .= $_;
 103                 last unless (/^import / || /^$/);
 104         }
 105         my $lineno = $. - 1;
 106
 107         return ($header, $preamble, $lineno,
 108                 "/*\n", sub { s/^/ */mg }, " */\n");
 109 }
 110
 111 sub pom_file
 112 {
 113         my $fd = shift;
 114         my $header = '';
 115         my $preamble = '';
 116
 117         # header is everything before project
 118         while (<$fd>) {
 119                 if (/<project/) {
 120                         $preamble = $_;
 121                         last;
 122                 }
 123                 $header .= $_;
 124         }
 125         my $lineno = $. - 1;
 126
 127         return ($header, $preamble, $lineno,
 128                 qq{<?xml version="1.0" encoding="UTF-8"?>\n<!--\n},
 129                 sub { s/^(.)/  $1/mg },
 130                 qq{-->\n});
 131 }
 132
 133 sub sh_file
 134 {
 135         my $fd = shift;
 136         my $top = <$fd>;
 137         my $header = '';
 138         my $preamble = '';
 139
 140         while (<$fd>) {
 141                 if (/^#/) {
 142                         $header .= $_;
 143                         next;
 144                 }
 145                 $preamble = $_;
 146                 last;
 147         }
 148         my $lineno = $. - 1;
 149
 150         return ($header, $preamble, $lineno, $top, sub { s/^/#/mg }, "");
 151 }
 152
 153 sub update_file
 154 {
 155         my $func = shift;
 156         my $old_file = shift;
 157         my $new_file = "$old_file.license.$$";
 158
 159         open(I, '<', $old_file);
 160         my ($header, $preamble, $lineno,
 161                 $top, $fmt, $btm) = &{$func}(\*I);
 162
 163         my %all_years;
 164         my %author_years;
 165         my %minyear;
 166         my %maxyear;
 167
 168         # find explicit copyright statements in sources
 169         my @lines = split( /\n/, $header );
 170         foreach my $line ( @lines ) {
 171                 # * Copyright (c) 2008, Example Company Inc.
 172                 # * Copyright (c) 2008, Joe Developer <joe.dev@example.org>
 173                 # * Copyright (c) 2008, 2009 Joe Developer <joe.dev@example.org>
 174                 # * Copyright (c) 2005-2009 Joe Developer <joe.dev@example.org>
 175                 # * Copyright (c) 2008, 2009 Other Examples Inc.
 176                 # * Copyright (c) 2008-2010 Example Company Inc.
 177                 # * Copyright (C) 2009-2010, Yet More Examples Ltd.
 178                 if( $line =~ m/Copyright \(c\) (\d{4})(?:\s*[,-]\s*(\d{4}))?,?\s*([^<>]+)\s*(<.*?>)?/i ) {
 179                         my ($y, $y2, $n, $e) = ($1, $2, $3, $4);
 180                         my $year = trim($y);
 181                         my $author_name = trim($n);
 182                         my $author_email = trim($e);
 183                         my $who = $author_name;
 184                         $who .= " $author_email" if $author_email;
 185                         update_author_info(\%minyear, \%maxyear, \%all_years, \%author_years, $who, $year);
 186                         if (my $year2 = $y2) {
 187                                 update_author_info(\%minyear, \%maxyear, \%all_years, \%author_years, $who, $year2);
 188                         }
 189                 }
 190         }
 191
 192         if ($old_file eq 'LICENSE') {
 193         } else {
 194                 # add implicit copyright statements from authors found in git blame
 195                 my (%line_counts, %line_authors);
 196                 my ($last_commit, $author_name, $author_email);
 197                 my @blame_args = ('git', 'blame', "-L$lineno,", '-C', '-w', '-p');
 198                 push(@blame_args, $ENV{'GIT_COMMIT'}) if $ENV{'GIT_COMMIT'};
 199                 push(@blame_args, '--', $old_file);
 200                 open( B, '-|', @blame_args);
 201                 while (<B>) {
 202                         chop;
 203                         if (/^([0-9a-f]{40}) \d+ \d+ (\d+)$/) {
 204                                 $last_commit = $1;
 205                                 $line_counts{$1} += $2;
 206                                 next;
 207                         }
 208                         if (/^author (.*)$/) {
 209                                 $author_name = trim($1);
 210                                 next;
 211                         }
 212                         if (/^author-mail (<.*>)$/) {
 213                                 $author_email = trim($1);
 214                                 next;
 215                         }
 216                         if (/^author-time (\d+)$/) {
 217                                 # skip uncommitted changes
 218                                 my $who = "$author_name $author_email";
 219                                 next if $who eq 'Not Committed Yet <not.committed.yet>';
 220                                 my @tm = localtime($1);
 221                                 my $year = $tm[5] + 1900;
 222                                 my $mon = $tm[4] + 1;
 223                                 $who = translate_author($who, $year, $mon);
 224                                 $line_authors{$last_commit} = [$who, $year, $mon];
 225                         }
 226                 }
 227                 close B;
 228
 229                 my %author_linecounts;
 230                 foreach $last_commit (keys %line_counts) {
 231                         my $who = $line_authors{$last_commit}[0];
 232                         next unless $who;
 233                         $author_linecounts{$who} += $line_counts{$last_commit};
 234                 }
 235
 236                 my $sz = 100;
 237                 my $count_big = 0;
 238                 foreach (values %author_linecounts) {
 239                         $count_big++ if $_ >= $sz;
 240                 }
 241
 242                 my $added_count = 0;
 243                 foreach (values %line_authors) {
 244                         my ($who, $year, $mon) = @$_;
 245                         next if ($count_big && $author_linecounts{$who} < $sz);
 246                         $all_years{$year} = 1;
 247                         update_author_info(\%minyear, \%maxyear, \%all_years, \%author_years, $who, $year, $mon);
 248                 }
 249         }
 250
 251         # rewrite file
 252         open( O, '>', $new_file );
 253         print O $top;
 254
 255         my %used_author;
 256         foreach my $year ( sort { $a cmp $b } keys %all_years ) {
 257                 foreach my $who ( sort keys %author_years ) {
 258                         next if $used_author{$who}++;
 259                         local $_ = format_copyright($minyear{$who}, $maxyear{$who}, $who);
 260                         &{$fmt}();
 261                         print O;
 262                 }
 263         }
 264
 265         local $_ = $license_text;
 266         &{$fmt}();
 267         print O;
 268         print O $btm;
 269         print O "\n";
 270         print O $preamble;
 271         print O while <I>;
 272         close I;
 273         close O;
 274
 275         rename( $new_file, $old_file );
 276 }
 277
 278 sub trim($)
 279 {
 280         my $string = shift;
 281         $string =~ s/^\s+//;
 282         $string =~ s/\s+$//;
 283         return $string;
 284 }
 285
 286 sub update_author_info
 287 {
 288         my ($minyear_ref, $maxyear_ref, $all_years_ref, $author_years_ref, $who, $year, $mon) = @_;
 289
 290         $who = translate_author($who, $year, $mon);
 291         $all_years_ref->{$year} = 1;
 292         $author_years_ref->{$who}{$year} = 1;
 293
 294         my $y = $minyear_ref->{$who};
 295         if ($y < 1900) {
 296                 $y = 9999;
 297         }
 298         if ($year < $y) {
 299                 $minyear_ref->{$who} = $year;
 300         }
 301         $y = $maxyear_ref->{$who};
 302         if ($year > $y) {
 303                 $maxyear_ref->{$who} = $year;
 304         }
 305 }
 306
 307 sub date_cmp
 308 {
 309         my ($a_year, $a_mon, $b_year, $b_mon) = @_;
 310
 311         if ($a_year < $b_year) {
 312                 return -1;
 313         } elsif ($a_year == $b_year) {
 314                 return ($a_mon <=> $b_mon);
 315         } else {
 316                 return 1;
 317         }
 318 }
 319
 320 sub translate_author
 321 {
 322         my ($who, $year, $mon) = @_;
 323
 324         return $who if not defined $mon;
 325
 326         foreach my $spec (@author_employers) {
 327                 next unless $who =~ $spec->[0];
 328                 next if (date_cmp($year, $mon, $spec->[1], $spec->[2]) < 0);
 329                 next if (date_cmp($year, $mon, $spec->[3], $spec->[4]) > 0);
 330                 return $spec->[5];
 331         }
 332         return $who;
 333 }
 334
 335 sub format_copyright {
 336         my ($minyear, $maxyear, $who) = @_;
 337         if ($minyear < $maxyear) {
 338                 return " Copyright (C) $minyear-$maxyear, $who\n";
 339         } else {
 340                 return " Copyright (C) $minyear, $who\n";
 341         }
 342 }
 343