1#!/bin/sh
2#! -*-perl-*-
3
4# Convert git log output to ChangeLog format.
5
6# Copyright (C) 2008-2021 Free Software Foundation, Inc.
7#
8# This program is free software: you can redistribute it and/or modify
9# it under the terms of the GNU General Public License as published by
10# the Free Software Foundation, either version 3 of the License, or
11# (at your option) any later version.
12#
13# This program is distributed in the hope that it will be useful,
14# but WITHOUT ANY WARRANTY; without even the implied warranty of
15# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
16# GNU General Public License for more details.
17#
18# You should have received a copy of the GNU General Public License
19# along with this program.  If not, see <https://www.gnu.org/licenses/>.
20#
21# Written by Jim Meyering
22
23# This is a prologue that allows to run a perl script as an executable
24# on systems that are compliant to a POSIX version before POSIX:2017.
25# On such systems, the usual invocation of an executable through execlp()
26# or execvp() fails with ENOEXEC if it is a script that does not start
27# with a #! line.  The script interpreter mentioned in the #! line has
28# to be /bin/sh, because on GuixSD systems that is the only program that
29# has a fixed file name.  The second line is essential for perl and is
30# also useful for editing this file in Emacs.  The next two lines below
31# are valid code in both sh and perl.  When executed by sh, they re-execute
32# the script through the perl program found in $PATH.  The '-x' option
33# is essential as well; without it, perl would re-execute the script
34# through /bin/sh.  When executed by perl, the next two lines are a no-op.
35eval 'exec perl -wSx "$0" "$@"'
36     if 0;
37
38my $VERSION = '2021-02-24 23:42'; # UTC
39# The definition above must lie within the first 8 lines in order
40# for the Emacs time-stamp write hook (at end) to update it.
41# If you change this file with Emacs, please let the write hook
42# do its job.  Otherwise, update this string manually.
43
44use strict;
45use warnings;
46use Getopt::Long;
47use POSIX qw(strftime);
48
49(my $ME = $0) =~ s|.*/||;
50
51# use File::Coda; # https://meyering.net/code/Coda/
52END {
53  defined fileno STDOUT or return;
54  close STDOUT and return;
55  warn "$ME: failed to close standard output: $!\n";
56  $? ||= 1;
57}
58
59sub usage ($)
60{
61  my ($exit_code) = @_;
62  my $STREAM = ($exit_code == 0 ? *STDOUT : *STDERR);
63  if ($exit_code != 0)
64    {
65      print $STREAM "Try '$ME --help' for more information.\n";
66    }
67  else
68    {
69      print $STREAM <<EOF;
70Usage: $ME [OPTIONS] [ARGS]
71
72Convert git log output to ChangeLog format.  If present, any ARGS
73are passed to "git log".  To avoid ARGS being parsed as options to
74$ME, they may be preceded by '--'.
75
76OPTIONS:
77
78   --amend=FILE FILE maps from an SHA1 to perl code (i.e., s/old/new/) that
79                  makes a change to SHA1's commit log text or metadata.
80   --append-dot append a dot to the first line of each commit message if
81                  there is no other punctuation or blank at the end.
82   --no-cluster never cluster commit messages under the same date/author
83                  header; the default is to cluster adjacent commit messages
84                  if their headers are the same and neither commit message
85                  contains multiple paragraphs.
86   --srcdir=DIR the root of the source tree, from which the .git/
87                  directory can be derived.
88   --since=DATE convert only the logs since DATE;
89                  the default is to convert all log entries.
90   --until=DATE convert only the logs older than DATE.
91   --ignore-matching=PAT ignore commit messages whose first lines match PAT.
92   --ignore-line=PAT ignore lines of commit messages that match PAT.
93   --format=FMT set format string for commit subject and body;
94                  see 'man git-log' for the list of format metacharacters;
95                  the default is '%s%n%b%n'
96   --strip-tab  remove one additional leading TAB from commit message lines.
97   --strip-cherry-pick  remove data inserted by "git cherry-pick";
98                  this includes the "cherry picked from commit ..." line,
99                  and the possible final "Conflicts:" paragraph.
100   --help       display this help and exit
101   --version    output version information and exit
102
103EXAMPLE:
104
105  $ME --since=2008-01-01 > ChangeLog
106  $ME -- -n 5 foo > last-5-commits-to-branch-foo
107
108SPECIAL SYNTAX:
109
110The following types of strings are interpreted specially when they appear
111at the beginning of a log message line.  They are not copied to the output.
112
113  Copyright-paperwork-exempt: Yes
114    Append the "(tiny change)" notation to the usual "date name email"
115    ChangeLog header to mark a change that does not require a copyright
116    assignment.
117  Co-authored-by: Joe User <user\@example.com>
118    List the specified name and email address on a second
119    ChangeLog header, denoting a co-author.
120  Signed-off-by: Joe User <user\@example.com>
121    These lines are simply elided.
122
123In a FILE specified via --amend, comment lines (starting with "#") are ignored.
124FILE must consist of <SHA,CODE+> pairs where SHA is a 40-byte SHA1 (alone on
125a line) referring to a commit in the current project, and CODE refers to one
126or more consecutive lines of Perl code.  Pairs must be separated by one or
127more blank line.
128
129Here is sample input for use with --amend=FILE, from coreutils:
130
1313a169f4c5d9159283548178668d2fae6fced3030
132# fix typo in title:
133s/all tile types/all file types/
134
1351379ed974f1fa39b12e2ffab18b3f7a607082202
136# Due to a bug in vc-dwim, I mis-attributed a patch by Paul to myself.
137# Change the author to be Paul.  Note the escaped "@":
138s,Jim .*>,Paul Eggert <eggert\\\@cs.ucla.edu>,
139
140EOF
141    }
142  exit $exit_code;
143}
144
145# If the string $S is a well-behaved file name, simply return it.
146# If it contains white space, quotes, etc., quote it, and return the new string.
147sub shell_quote($)
148{
149  my ($s) = @_;
150  if ($s =~ m![^\w+/.,-]!)
151    {
152      # Convert each single quote to '\''
153      $s =~ s/\'/\'\\\'\'/g;
154      # Then single quote the string.
155      $s = "'$s'";
156    }
157  return $s;
158}
159
160sub quoted_cmd(@)
161{
162  return join (' ', map {shell_quote $_} @_);
163}
164
165# Parse file F.
166# Comment lines (starting with "#") are ignored.
167# F must consist of <SHA,CODE+> pairs where SHA is a 40-byte SHA1
168# (alone on a line) referring to a commit in the current project, and
169# CODE refers to one or more consecutive lines of Perl code.
170# Pairs must be separated by one or more blank line.
171sub parse_amend_file($)
172{
173  my ($f) = @_;
174
175  open F, '<', $f
176    or die "$ME: $f: failed to open for reading: $!\n";
177
178  my $fail;
179  my $h = {};
180  my $in_code = 0;
181  my $sha;
182  while (defined (my $line = <F>))
183    {
184      $line =~ /^\#/
185        and next;
186      chomp $line;
187      $line eq ''
188        and $in_code = 0, next;
189
190      if (!$in_code)
191        {
192          $line =~ /^([[:xdigit:]]{40})$/
193            or (warn "$ME: $f:$.: invalid line; expected an SHA1\n"),
194              $fail = 1, next;
195          $sha = lc $1;
196          $in_code = 1;
197          exists $h->{$sha}
198            and (warn "$ME: $f:$.: duplicate SHA1\n"),
199              $fail = 1, next;
200        }
201      else
202        {
203          $h->{$sha} ||= '';
204          $h->{$sha} .= "$line\n";
205        }
206    }
207  close F;
208
209  $fail
210    and exit 1;
211
212  return $h;
213}
214
215# git_dir_option $SRCDIR
216#
217# From $SRCDIR, the --git-dir option to pass to git (none if $SRCDIR
218# is undef).  Return as a list (0 or 1 element).
219sub git_dir_option($)
220{
221  my ($srcdir) = @_;
222  my @res = ();
223  if (defined $srcdir)
224    {
225      my $qdir = shell_quote $srcdir;
226      my $cmd = "cd $qdir && git rev-parse --show-toplevel";
227      my $qcmd = shell_quote $cmd;
228      my $git_dir = qx($cmd);
229      defined $git_dir
230        or die "$ME: cannot run $qcmd: $!\n";
231      $? == 0
232        or die "$ME: $qcmd had unexpected exit code or signal ($?)\n";
233      chomp $git_dir;
234      push @res, "--git-dir=$git_dir/.git";
235    }
236  @res;
237}
238
239{
240  my $since_date;
241  my $until_date;
242  my $format_string = '%s%n%b%n';
243  my $amend_file;
244  my $append_dot = 0;
245  my $cluster = 1;
246  my $ignore_matching;
247  my $ignore_line;
248  my $strip_tab = 0;
249  my $strip_cherry_pick = 0;
250  my $srcdir;
251  GetOptions
252    (
253     help => sub { usage 0 },
254     version => sub { print "$ME version $VERSION\n"; exit },
255     'since=s' => \$since_date,
256     'until=s' => \$until_date,
257     'format=s' => \$format_string,
258     'amend=s' => \$amend_file,
259     'append-dot' => \$append_dot,
260     'cluster!' => \$cluster,
261     'ignore-matching=s' => \$ignore_matching,
262     'ignore-line=s' => \$ignore_line,
263     'strip-tab' => \$strip_tab,
264     'strip-cherry-pick' => \$strip_cherry_pick,
265     'srcdir=s' => \$srcdir,
266    ) or usage 1;
267
268  defined $since_date
269    and unshift @ARGV, "--since=$since_date";
270  defined $until_date
271    and unshift @ARGV, "--until=$until_date";
272
273  # This is a hash that maps an SHA1 to perl code (i.e., s/old/new/)
274  # that makes a correction in the log or attribution of that commit.
275  my $amend_code = defined $amend_file ? parse_amend_file $amend_file : {};
276
277  my @cmd = ('git',
278             git_dir_option $srcdir,
279             qw(log --log-size),
280             '--pretty=format:%H:%ct  %an  <%ae>%n%n'.$format_string, @ARGV);
281  open PIPE, '-|', @cmd
282    or die ("$ME: failed to run '". quoted_cmd (@cmd) ."': $!\n"
283            . "(Is your Git too old?  Version 1.5.1 or later is required.)\n");
284
285  my $prev_multi_paragraph;
286  my $prev_date_line = '';
287  my @prev_coauthors = ();
288  my @skipshas = ();
289  while (1)
290    {
291      defined (my $in = <PIPE>)
292        or last;
293      $in =~ /^log size (\d+)$/
294        or die "$ME:$.: Invalid line (expected log size):\n$in";
295      my $log_nbytes = $1;
296
297      my $log;
298      my $n_read = read PIPE, $log, $log_nbytes;
299      $n_read == $log_nbytes
300        or die "$ME:$.: unexpected EOF\n";
301
302      # Extract leading hash.
303      my ($sha, $rest) = split ':', $log, 2;
304      defined $sha
305        or die "$ME:$.: malformed log entry\n";
306      $sha =~ /^[[:xdigit:]]{40}$/
307        or die "$ME:$.: invalid SHA1: $sha\n";
308
309      my $skipflag = 0;
310      if (@skipshas)
311        {
312          foreach(@skipshas)
313            {
314              if ($sha =~ /^$_/)
315                {
316                  $skipflag = $_;
317                  last;
318                }
319            }
320        }
321
322      # If this commit's log requires any transformation, do it now.
323      my $code = $amend_code->{$sha};
324      if (defined $code)
325        {
326          eval 'use Safe';
327          my $s = new Safe;
328          # Put the unpreprocessed entry into "$_".
329          $_ = $rest;
330
331          # Let $code operate on it, safely.
332          my $r = $s->reval("$code")
333            or die "$ME:$.:$sha: failed to eval \"$code\":\n$@\n";
334
335          # Note that we've used this entry.
336          delete $amend_code->{$sha};
337
338          # Update $rest upon success.
339          $rest = $_;
340        }
341
342      # Remove lines inserted by "git cherry-pick".
343      if ($strip_cherry_pick)
344        {
345          $rest =~ s/^\s*Conflicts:\n.*//sm;
346          $rest =~ s/^\s*\(cherry picked from commit [\da-f]+\)\n//m;
347        }
348
349      my @line = split /[ \t]*\n/, $rest;
350      my $author_line = shift @line;
351      defined $author_line
352        or die "$ME:$.: unexpected EOF\n";
353      $author_line =~ /^(\d+)  (.*>)$/
354        or die "$ME:$.: Invalid line "
355          . "(expected date/author/email):\n$author_line\n";
356
357      # Format 'Copyright-paperwork-exempt: Yes' as a standard ChangeLog
358      # `(tiny change)' annotation.
359      my $tiny = (grep (/^(?:Copyright-paperwork-exempt|Tiny-change):\s+[Yy]es$/, @line)
360                  ? '  (tiny change)' : '');
361
362      my $date_line = sprintf "%s  %s$tiny\n",
363        strftime ("%Y-%m-%d", localtime ($1)), $2;
364
365      my @coauthors = grep /^Co-authored-by:.*$/, @line;
366      # Omit meta-data lines we've already interpreted.
367      @line = grep !/^(?:Signed-off-by:[ ].*>$
368                       |Co-authored-by:[ ]
369                       |Copyright-paperwork-exempt:[ ]
370                       |Tiny-change:[ ]
371                       )/x, @line;
372
373      # Remove leading and trailing blank lines.
374      if (@line)
375        {
376          while ($line[0] =~ /^\s*$/) { shift @line; }
377          while ($line[$#line] =~ /^\s*$/) { pop @line; }
378        }
379
380      # Handle Emacs gitmerge.el "skipped" commits.
381      # Yes, this should be controlled by an option.  So sue me.
382      if ( grep /^(; )?Merge from /, @line )
383      {
384          my $found = 0;
385          foreach (@line)
386          {
387              if (grep /^The following commit.*skipped:$/, $_)
388              {
389                  $found = 1;
390                  ## Reset at each merge to reduce chance of false matches.
391                  @skipshas = ();
392                  next;
393              }
394              if ($found && $_ =~ /^([[:xdigit:]]{7,}) [^ ]/)
395              {
396                  push ( @skipshas, $1 );
397              }
398          }
399      }
400
401      # Ignore commits that match the --ignore-matching pattern, if specified.
402      if (defined $ignore_matching && @line && $line[0] =~ /$ignore_matching/)
403        {
404          $skipflag = 1;
405        }
406      elsif ($skipflag)
407        {
408          ## Perhaps only warn if a pattern matches more than once?
409          warn "$ME: warning: skipping $sha due to $skipflag\n";
410        }
411
412      if (! $skipflag)
413        {
414          if (defined $ignore_line && @line)
415            {
416              @line = grep ! /$ignore_line/, @line;
417              while ($line[$#line] =~ /^\s*$/) { pop @line; }
418            }
419
420          # Record whether there are two or more paragraphs.
421          my $multi_paragraph = grep /^\s*$/, @line;
422
423          # Format 'Co-authored-by: A U Thor <email@example.com>' lines in
424          # standard multi-author ChangeLog format.
425          for (@coauthors)
426            {
427              s/^Co-authored-by:\s*/\t    /;
428              s/\s*</  </;
429
430              /<.*?@.*\..*>/
431                or warn "$ME: warning: missing email address for "
432                  . substr ($_, 5) . "\n";
433            }
434
435          # If clustering of commit messages has been disabled, if this header
436          # would be different from the previous date/name/etc. header,
437          # or if this or the previous entry consists of two or more paragraphs,
438          # then print the header.
439          if ( ! $cluster
440              || $date_line ne $prev_date_line
441              || "@coauthors" ne "@prev_coauthors"
442              || $multi_paragraph
443              || $prev_multi_paragraph)
444            {
445              $prev_date_line eq ''
446                or print "\n";
447              print $date_line;
448              @coauthors
449                and print join ("\n", @coauthors), "\n";
450            }
451          $prev_date_line = $date_line;
452          @prev_coauthors = @coauthors;
453          $prev_multi_paragraph = $multi_paragraph;
454
455          # If there were any lines
456          if (@line == 0)
457            {
458              warn "$ME: warning: empty commit message:\n"
459                   . "  commit $sha\n  $date_line\n";
460            }
461          else
462            {
463              if ($append_dot)
464                {
465                  # If the first line of the message has enough room, then
466                  if (length $line[0] < 72)
467                    {
468                      # append a dot if there is no other punctuation or blank
469                      # at the end.
470                      $line[0] =~ /[[:punct:]\s]$/
471                        or $line[0] .= '.';
472                    }
473                }
474
475              # Remove one additional leading TAB from each line.
476              $strip_tab
477                and map { s/^\t// } @line;
478
479              # Prefix each non-empty line with a TAB.
480              @line = map { length $_ ? "\t$_" : '' } @line;
481
482              print "\n", join ("\n", @line), "\n";
483            }
484        }
485
486      defined ($in = <PIPE>)
487        or last;
488      $in ne "\n"
489        and die "$ME:$.: unexpected line:\n$in";
490    }
491
492  close PIPE
493    or die "$ME: error closing pipe from " . quoted_cmd (@cmd) . "\n";
494  # FIXME-someday: include $PROCESS_STATUS in the diagnostic
495
496  # Complain about any unused entry in the --amend=F specified file.
497  my $fail = 0;
498  foreach my $sha (keys %$amend_code)
499    {
500      warn "$ME:$amend_file: unused entry: $sha\n";
501      $fail = 1;
502    }
503
504  exit $fail;
505}
506
507# Local Variables:
508# mode: perl
509# indent-tabs-mode: nil
510# eval: (add-hook 'before-save-hook 'time-stamp)
511# time-stamp-line-limit: 50
512# time-stamp-start: "my $VERSION = '"
513# time-stamp-format: "%:y-%02m-%02d %02H:%02M"
514# time-stamp-time-zone: "UTC0"
515# time-stamp-end: "'; # UTC"
516# End:
517