3 # Copyright (C) 2013-2020 Free Software Foundation, Inc.
5 # This script is free software; you can redistribute it and/or modify
6 # it under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 3, or (at your option)
10 # This script adjusts the copyright notices at the top of source files
11 # so that they have the form:
13 # Copyright XXXX-YYYY Free Software Foundation, Inc.
15 # It doesn't change code that is known to be maintained elsewhere or
16 # that carries a non-FSF copyright.
18 # The script also doesn't change testsuite files, except those in
19 # libstdc++-v3. This is because libstdc++-v3 has a conformance testsuite,
20 # while most tests in other directories are just things that failed at some
23 # Pass --this-year to the script if you want it to add the current year
24 # to all applicable notices. Pass --quilt if you are using quilt and
25 # want files to be added to the quilt before being changed.
27 # By default the script will update all directories for which the
28 # output has been vetted. You can instead pass the names of individual
29 # directories, including those that haven't been approved. So:
31 # update-copyright.py --this-year
33 # is the command that would be used at the beginning of a year to update
34 # all copyright notices (and possibly at other times to check whether
35 # new files have been added with old years). On the other hand:
37 # update-copyright.py --this-year libitm
39 # would run the script on just libitm/.
41 # Note that things like --version output strings must be updated before
42 # this script is run. There's already a separate procedure for that.
54 def report (self
, filename
, string
):
56 string
= filename
+ ': ' + string
57 sys
.stderr
.write (string
+ '\n')
61 return self
.num_errors
== 0
65 self
.skip_files
= set()
66 self
.skip_dirs
= set()
67 self
.skip_extensions
= set([
71 self
.fossilised_files
= set()
72 self
.own_files
= set()
74 self
.skip_files |
= set ([
87 # Skip auto- and libtool-related files
108 # Skip FSF mission statement, etc.
113 # Skip imported texinfo files.
118 def get_line_filter (self
, dir, filename
):
119 if filename
.startswith ('ChangeLog'):
120 # Ignore references to copyright in changelog entries.
121 return re
.compile ('\t')
125 def skip_file (self
, dir, filename
):
126 if filename
in self
.skip_files
:
129 (base
, extension
) = os
.path
.splitext (os
.path
.join (dir, filename
))
130 if extension
in self
.skip_extensions
:
133 if extension
== '.in':
134 # Skip .in files produced by automake.
135 if os
.path
.exists (base
+ '.am'):
138 # Skip files produced by autogen
139 if (os
.path
.exists (base
+ '.def')
140 and os
.path
.exists (base
+ '.tpl')):
143 # Skip configure files produced by autoconf
144 if filename
== 'configure':
145 if os
.path
.exists (base
+ '.ac'):
147 if os
.path
.exists (base
+ '.in'):
152 def skip_dir (self
, dir, subdir
):
153 return subdir
in self
.skip_dirs
155 def is_fossilised_file (self
, dir, filename
):
156 if filename
in self
.fossilised_files
:
158 # Only touch current current ChangeLogs.
159 if filename
!= 'ChangeLog' and filename
.find ('ChangeLog') >= 0:
163 def by_package_author (self
, dir, filename
):
164 return filename
in self
.own_files
167 def __init__ (self
, errors
):
170 # Characters in a range of years. Include '.' for typos.
171 ranges
= '[0-9](?:[-0-9.,\s]|\s+and\s+)*[0-9]'
173 # Non-whitespace characters in a copyright holder's name.
177 self
.year_re
= re
.compile ('[0-9]+')
179 # Matches part of a year or copyright holder.
180 self
.continuation_re
= re
.compile (ranges
+ '|' + name
)
182 # Matches a full copyright notice:
183 self
.copyright_re
= re
.compile (
184 # 1: 'Copyright (C)', etc.
186 '|[Cc]opyright\s+\([Cc]\)'
188 '|[Cc]opyright\s+©'
189 '|[Cc]opyright\s+@copyright{}'
191 '|@set\s+copyright[\w-]+)'
193 # 2: the years. Include the whitespace in the year, so that
194 # we can remove any excess.
195 '(\s*(?:' + ranges
+ ',?'
196 '|@value\{[^{}]*\})\s*)'
201 # 4: the copyright holder. Don't allow multiple consecutive
202 # spaces, so that right-margin gloss doesn't get caught
203 # (e.g. gnat_ugn.texi).
204 '(' + name
+ '(?:\s?' + name
+ ')*)?')
206 # A regexp for notices that might have slipped by. Just matching
207 # 'copyright' is too noisy, and 'copyright.*[0-9]' falls foul of
208 # HTML header markers, so check for 'copyright' and two digits.
209 self
.other_copyright_re
= re
.compile ('copyright.*[0-9][0-9]',
211 self
.comment_re
= re
.compile('#+|[*]+|;+|%+|//+|@c |dnl ')
212 self
.holders
= { '@copying': '@copying' }
213 self
.holder_prefixes
= set()
215 # True to 'quilt add' files before changing them.
216 self
.use_quilt
= False
218 # If set, force all notices to include this year.
221 # Goes after the year(s). Could be ', '.
224 def add_package_author (self
, holder
, canon_form
= None):
227 self
.holders
[holder
] = canon_form
228 index
= holder
.find (' ')
230 self
.holder_prefixes
.add (holder
[:index
])
231 index
= holder
.find (' ', index
+ 1)
233 def add_external_author (self
, holder
):
234 self
.holders
[holder
] = None
237 def __init__ (self
, year
):
241 return 'unrecognised year: ' + self
.year
243 def parse_year (self
, string
):
245 if len (string
) == 2:
248 elif len (string
) == 4:
250 raise self
.BadYear (string
)
252 def year_range (self
, years
):
253 year_list
= [self
.parse_year (year
)
254 for year
in self
.year_re
.findall (years
)]
255 assert len (year_list
) > 0
256 return (min (year_list
), max (year_list
))
258 def set_use_quilt (self
, use_quilt
):
259 self
.use_quilt
= use_quilt
261 def include_year (self
, year
):
262 assert not self
.max_year
265 def canonicalise_years (self
, dir, filename
, filter, years
):
266 # Leave texinfo variables alone.
267 if years
.startswith ('@value'):
270 (min_year
, max_year
) = self
.year_range (years
)
272 # Update the upper bound, if enabled.
273 if self
.max_year
and not filter.is_fossilised_file (dir, filename
):
274 max_year
= max (max_year
, self
.max_year
)
277 if min_year
== max_year
:
278 return '%d' % min_year
280 return '%d-%d' % (min_year
, max_year
)
282 def strip_continuation (self
, line
):
284 match
= self
.comment_re
.match (line
)
286 line
= line
[match
.end():].lstrip()
289 def is_complete (self
, match
):
290 holder
= match
.group (4)
292 and (holder
not in self
.holder_prefixes
293 or holder
in self
.holders
))
295 def update_copyright (self
, dir, filename
, filter, file, line
, match
):
298 pathname
= os
.path
.join (dir, filename
)
300 intro
= match
.group (1)
301 if intro
.startswith ('@set'):
302 # Texinfo year variables should always be on one line
303 after_years
= line
[match
.end (2):].strip()
304 if after_years
!= '':
305 self
.errors
.report (pathname
,
306 'trailing characters in @set: '
308 return (False, orig_line
, next_line
)
310 # If it looks like the copyright is incomplete, add the next line.
311 while not self
.is_complete (match
):
313 next_line
= file.readline()
314 except StopIteration:
317 # If the next line doesn't look like a proper continuation,
318 # assume that what we've got is complete.
319 continuation
= self
.strip_continuation (next_line
)
320 if not self
.continuation_re
.match (continuation
):
323 # Merge the lines for matching purposes.
324 orig_line
+= next_line
325 line
= line
.rstrip() + ' ' + continuation
328 # Rematch with the longer line, at the original position.
329 match
= self
.copyright_re
.match (line
, match
.start())
332 holder
= match
.group (4)
334 # Use the filter to test cases where markup is getting in the way.
335 if filter.by_package_author (dir, filename
):
336 assert holder
not in self
.holders
339 self
.errors
.report (pathname
, 'missing copyright holder')
340 return (False, orig_line
, next_line
)
342 elif holder
not in self
.holders
:
343 self
.errors
.report (pathname
,
344 'unrecognised copyright holder: ' + holder
)
345 return (False, orig_line
, next_line
)
348 # See whether the copyright is associated with the package
350 canon_form
= self
.holders
[holder
]
352 return (False, orig_line
, next_line
)
354 # Make sure the author is given in a consistent way.
355 line
= (line
[:match
.start (4)]
357 + line
[match
.end (4):])
360 line
= line
[:match
.start (3)] + line
[match
.end (3):]
362 # Update the copyright years.
363 years
= match
.group (2).strip()
365 canon_form
= self
.canonicalise_years (dir, filename
, filter, years
)
366 except self
.BadYear
as e
:
367 self
.errors
.report (pathname
, str (e
))
368 return (False, orig_line
, next_line
)
370 line
= (line
[:match
.start (2)]
371 + ('' if intro
.startswith ('copyright = ') else ' ')
372 + canon_form
+ self
.separator
373 + line
[match
.end (2):])
375 # Use the standard (C) form.
376 if intro
.endswith ('right'):
378 elif intro
.endswith ('(c)'):
379 intro
= intro
[:-3] + '(C)'
380 line
= line
[:match
.start (1)] + intro
+ line
[match
.end (1):]
382 # Strip trailing whitespace
383 line
= line
.rstrip() + '\n'
385 return (line
!= orig_line
, line
, next_line
)
387 def guess_encoding (self
, pathname
):
388 for encoding
in ('utf8', 'iso8859'):
390 open(pathname
, 'r', encoding
=encoding
).read()
392 except UnicodeDecodeError:
396 def process_file (self
, dir, filename
, filter):
397 pathname
= os
.path
.join (dir, filename
)
398 if filename
.endswith ('.tmp'):
399 # Looks like something we tried to create before.
408 line_filter
= filter.get_line_filter (dir, filename
)
410 encoding
= self
.guess_encoding(pathname
)
411 with
open (pathname
, 'r', encoding
=encoding
) as file:
413 mode
= os
.fstat (file.fileno()).st_mode
417 # Leave filtered-out lines alone.
418 if not (line_filter
and line_filter
.match (line
)):
419 match
= self
.copyright_re
.search (line
)
421 res
= self
.update_copyright (dir, filename
, filter,
423 (this_changed
, line
, next_line
) = res
424 changed
= changed
or this_changed
426 # Check for copyright lines that might have slipped by.
427 elif self
.other_copyright_re
.search (line
):
428 self
.errors
.report (pathname
,
429 'unrecognised copyright: %s'
434 # If something changed, write the new file out.
435 if changed
and self
.errors
.ok():
436 tmp_pathname
= pathname
+ '.tmp'
437 with
open (tmp_pathname
, 'w', encoding
=encoding
) as file:
440 os
.fchmod (file.fileno(), mode
)
442 subprocess
.call (['quilt', 'add', pathname
])
443 os
.rename (tmp_pathname
, pathname
)
445 def process_tree (self
, tree
, filter):
446 for (dir, subdirs
, filenames
) in os
.walk (tree
):
447 # Don't recurse through directories that should be skipped.
448 for i
in range (len (subdirs
) - 1, -1, -1):
449 if filter.skip_dir (dir, subdirs
[i
]):
452 # Handle the files in this directory.
453 for filename
in filenames
:
454 if filter.skip_file (dir, filename
):
455 sys
.stdout
.write ('Skipping %s\n'
456 % os
.path
.join (dir, filename
))
458 self
.process_file (dir, filename
, filter)
461 def __init__ (self
, copyright
= Copyright
):
462 self
.errors
= Errors()
463 self
.copyright
= copyright (self
.errors
)
465 self
.default_dirs
= []
466 self
.chosen_dirs
= []
467 self
.option_handlers
= dict()
468 self
.option_help
= []
470 self
.add_option ('--help', 'Print this help', self
.o_help
)
471 self
.add_option ('--quilt', '"quilt add" files before changing them',
473 self
.add_option ('--this-year', 'Add the current year to every notice',
476 def add_option (self
, name
, help, handler
):
477 self
.option_help
.append ((name
, help))
478 self
.option_handlers
[name
] = handler
480 def add_dir (self
, dir, filter = GenericFilter()):
481 self
.dirs
.append ((dir, filter))
483 def o_help (self
, option
= None):
484 sys
.stdout
.write ('Usage: %s [options] dir1 dir2...\n\n'
485 'Options:\n' % sys
.argv
[0])
486 format
= '%-15s %s\n'
487 for (what
, help) in self
.option_help
:
488 sys
.stdout
.write (format
% (what
, help))
489 sys
.stdout
.write ('\nDirectories:\n')
493 for (dir, filter) in self
.dirs
:
495 if i
% 3 == 0 or i
== len (self
.dirs
):
496 sys
.stdout
.write (dir + '\n')
498 sys
.stdout
.write (format
% dir)
501 def o_quilt (self
, option
):
502 self
.copyright
.set_use_quilt (True)
504 def o_this_year (self
, option
):
505 self
.copyright
.include_year (time
.localtime().tm_year
)
508 for arg
in sys
.argv
[1:]:
510 self
.chosen_dirs
.append (arg
)
511 elif arg
in self
.option_handlers
:
512 self
.option_handlers
[arg
] (arg
)
514 self
.errors
.report (None, 'unrecognised option: ' + arg
)
516 if len (self
.chosen_dirs
) == 0:
517 self
.chosen_dirs
= self
.default_dirs
518 if len (self
.chosen_dirs
) == 0:
521 for chosen_dir
in self
.chosen_dirs
:
522 canon_dir
= os
.path
.join (chosen_dir
, '')
524 for (dir, filter) in self
.dirs
:
525 if (dir + os
.sep
).startswith (canon_dir
):
527 self
.copyright
.process_tree (dir, filter)
529 self
.errors
.report (None, 'unrecognised directory: '
531 sys
.exit (0 if self
.errors
.ok() else 1)
533 #----------------------------------------------------------------------------
535 class TopLevelFilter (GenericFilter
):
536 def skip_dir (self
, dir, subdir
):
539 class ConfigFilter (GenericFilter
):
541 GenericFilter
.__init
__ (self
)
543 def skip_file (self
, dir, filename
):
544 if filename
.endswith ('.m4'):
545 pathname
= os
.path
.join (dir, filename
)
546 with
open (pathname
) as file:
547 # Skip files imported from gettext.
548 if file.readline().find ('gettext-') >= 0:
550 return GenericFilter
.skip_file (self
, dir, filename
)
552 class GCCFilter (GenericFilter
):
554 GenericFilter
.__init
__ (self
)
556 self
.skip_files |
= set ([
561 self
.skip_dirs |
= set ([
562 # Better not create a merge nightmare for the GNAT folks.
565 # Handled separately.
569 self
.skip_extensions |
= set ([
570 # Maintained by the translation project.
573 # Automatically-generated.
577 self
.fossilised_files |
= set ([
578 # Old news won't be updated.
582 class TestsuiteFilter (GenericFilter
):
584 GenericFilter
.__init
__ (self
)
586 self
.skip_extensions |
= set ([
587 # Don't change the tests, which could be woend by anyone.
601 def skip_file (self
, dir, filename
):
602 # g++.niklas/README contains historical copyright information
604 if filename
== 'README' and os
.path
.basename (dir) == 'g++.niklas':
606 # Similarly params/README.
607 if filename
== 'README' and os
.path
.basename (dir) == 'params':
609 if filename
== 'pdt_5.f03' and os
.path
.basename (dir) == 'gfortran.dg':
611 return GenericFilter
.skip_file (self
, dir, filename
)
613 class LibCppFilter (GenericFilter
):
615 GenericFilter
.__init
__ (self
)
617 self
.skip_extensions |
= set ([
618 # Maintained by the translation project.
621 # Automatically-generated.
625 class LibGCCFilter (GenericFilter
):
627 GenericFilter
.__init
__ (self
)
629 self
.skip_dirs |
= set ([
630 # Imported from GLIBC.
634 class LibPhobosFilter (GenericFilter
):
636 GenericFilter
.__init
__ (self
)
638 self
.skip_files |
= set ([
639 # Source module imported from upstream.
643 self
.skip_dirs |
= set ([
644 # Contains sources imported from upstream.
653 class LibStdCxxFilter (GenericFilter
):
655 GenericFilter
.__init
__ (self
)
657 self
.skip_files |
= set ([
658 # Contains no copyright of its own, but quotes the GPL.
662 self
.skip_dirs |
= set ([
663 # Contains automatically-generated sources.
666 # The testsuite data files shouldn't be changed.
669 # Contains imported images
673 self
.own_files |
= set ([
674 # Contains markup around the copyright owner.
678 def get_line_filter (self
, dir, filename
):
679 if filename
== 'boost_concept_check.h':
680 return re
.compile ('// \(C\) Copyright Jeremy Siek')
681 return GenericFilter
.get_line_filter (self
, dir, filename
)
683 class GCCCopyright (Copyright
):
684 def __init__ (self
, errors
):
685 Copyright
.__init
__ (self
, errors
)
687 canon_fsf
= 'Free Software Foundation, Inc.'
688 self
.add_package_author ('Free Software Foundation', canon_fsf
)
689 self
.add_package_author ('Free Software Foundation.', canon_fsf
)
690 self
.add_package_author ('Free Software Foundation Inc.', canon_fsf
)
691 self
.add_package_author ('Free Software Foundation, Inc', canon_fsf
)
692 self
.add_package_author ('Free Software Foundation, Inc.', canon_fsf
)
693 self
.add_package_author ('The Free Software Foundation', canon_fsf
)
694 self
.add_package_author ('The Free Software Foundation, Inc.', canon_fsf
)
695 self
.add_package_author ('Software Foundation, Inc.', canon_fsf
)
697 self
.add_external_author ('ARM')
698 self
.add_external_author ('AdaCore')
699 self
.add_external_author ('Advanced Micro Devices Inc.')
700 self
.add_external_author ('Ami Tavory and Vladimir Dreizin, IBM-HRL.')
701 self
.add_external_author ('Cavium Networks.')
702 self
.add_external_author ('Faraday Technology Corp.')
703 self
.add_external_author ('Florida State University')
704 self
.add_external_author ('Gerard Jungman')
705 self
.add_external_author ('Greg Colvin and Beman Dawes.')
706 self
.add_external_author ('Hewlett-Packard Company')
707 self
.add_external_author ('Intel Corporation')
708 self
.add_external_author ('Information Technology Industry Council.')
709 self
.add_external_author ('James Theiler, Brian Gough')
710 self
.add_external_author ('Makoto Matsumoto and Takuji Nishimura,')
711 self
.add_external_author ('Mentor Graphics Corporation')
712 self
.add_external_author ('National Research Council of Canada.')
713 self
.add_external_author ('NVIDIA Corporation')
714 self
.add_external_author ('Peter Dimov and Multi Media Ltd.')
715 self
.add_external_author ('Peter Dimov')
716 self
.add_external_author ('Pipeline Associates, Inc.')
717 self
.add_external_author ('Regents of the University of California.')
718 self
.add_external_author ('Silicon Graphics Computer Systems, Inc.')
719 self
.add_external_author ('Silicon Graphics')
720 self
.add_external_author ('Stephen L. Moshier')
721 self
.add_external_author ('Sun Microsystems, Inc. All rights reserved.')
722 self
.add_external_author ('The D Language Foundation, All Rights Reserved')
723 self
.add_external_author ('The Go Authors. All rights reserved.')
724 self
.add_external_author ('The Go Authors. All rights reserved.')
725 self
.add_external_author ('The Go Authors.')
726 self
.add_external_author ('The Regents of the University of California.')
727 self
.add_external_author ('Ulf Adams')
728 self
.add_external_author ('Unicode, Inc.')
729 self
.add_external_author ('University of Toronto.')
730 self
.add_external_author ('Yoshinori Sato')
732 class GCCCmdLine (CmdLine
):
734 CmdLine
.__init
__ (self
, GCCCopyright
)
736 self
.add_dir ('.', TopLevelFilter())
737 # boehm-gc is imported from upstream.
738 self
.add_dir ('c++tools')
739 self
.add_dir ('config', ConfigFilter())
740 # contrib isn't really part of GCC.
741 self
.add_dir ('fixincludes')
742 self
.add_dir ('gcc', GCCFilter())
743 self
.add_dir (os
.path
.join ('gcc', 'testsuite'), TestsuiteFilter())
744 self
.add_dir ('gnattools')
745 self
.add_dir ('gotools')
746 self
.add_dir ('include')
747 # intl is imported from upstream.
748 self
.add_dir ('libada')
749 self
.add_dir ('libatomic')
750 self
.add_dir ('libbacktrace')
751 self
.add_dir ('libcc1')
752 self
.add_dir ('libcpp', LibCppFilter())
753 self
.add_dir ('libdecnumber')
754 # libffi is imported from upstream.
755 self
.add_dir ('libgcc', LibGCCFilter())
756 self
.add_dir ('libgfortran')
757 # libgo is imported from upstream.
758 self
.add_dir ('libgomp')
759 self
.add_dir ('libiberty')
760 self
.add_dir ('libitm')
761 self
.add_dir ('libobjc')
762 # liboffloadmic is imported from upstream.
763 self
.add_dir ('libphobos', LibPhobosFilter())
764 self
.add_dir ('libquadmath')
765 # libsanitizer is imported from upstream.
766 self
.add_dir ('libssp')
767 self
.add_dir ('libstdc++-v3', LibStdCxxFilter())
768 self
.add_dir ('libvtv')
769 self
.add_dir ('lto-plugin')
770 # maintainer-scripts maintainer-scripts
771 # zlib is imported from upstream.
773 self
.default_dirs
= [