update-copyright.py for binutils
authorAlan Modra <amodra@gmail.com>
Wed, 4 Jan 2017 13:19:58 +0000 (23:49 +1030)
committerAlan Modra <amodra@gmail.com>
Wed, 4 Jan 2017 13:33:07 +0000 (00:03 +1030)
This is a modified form of gcc's contrib/update-copyright.py.

* update-copyright.py: New file.
* add-log.el: Update copyright year range.
* texi2pod.pl: Likewise.

etc/ChangeLog
etc/add-log.el
etc/texi2pod.pl
etc/update-copyright.py [new file with mode: 0755]

index 1864bf0c89f409b8856271915d4f3385e930a1ef..165fdd58b2239d23baf50d248c19b41c5f746256 100644 (file)
@@ -1,3 +1,9 @@
+2017-01-04  Alan Modra  <amodra@gmail.com>
+
+       * update-copyright.py: New file.
+       * add-log.el: Update copyright year range.
+       * texi2pod.pl: Likewise.
+
 2016-06-13  Nick Clifton  <nickc@redhat.com>
 
        * texi2pod.pl: Escape curly braces, whilst searching for keyword
index 60c88e8c9490a60797b3cfbe1ac296465800bb4f..03eee2546a134eadfc177212c8bbb9c77c20f19c 100644 (file)
@@ -25,7 +25,7 @@
 
 ;;; add-log.el --- change log maintenance commands for Emacs
 
-;; Copyright (C) 1985, 1986, 1988, 1993, 1994 Free Software Foundation, Inc.
+;; Copyright (C) 1985-2017 Free Software Foundation, Inc.
 
 ;; Keywords: maint
 
index 0f465b7c768cb0213a617d47d4151c2a20e05229..e8192e4d36d806ba2b8f26dfe64ad7adbd5f86f0 100644 (file)
@@ -1,6 +1,6 @@
 #! /usr/bin/perl -w
 
-#   Copyright (C) 1999, 2000, 2001, 2003 Free Software Foundation, Inc.
+#   Copyright (C) 1999-2017 Free Software Foundation, Inc.
 
 # This file is part of GCC.
 
diff --git a/etc/update-copyright.py b/etc/update-copyright.py
new file mode 100755 (executable)
index 0000000..55d443b
--- /dev/null
@@ -0,0 +1,620 @@
+#!/usr/bin/python
+#
+# Copyright (C) 2013-2017 Free Software Foundation, Inc.
+#
+# This script is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3, or (at your option)
+# any later version.
+
+# This script adjusts the copyright notices at the top of source files
+# so that they have the form:
+#
+#   Copyright XXXX-YYYY Free Software Foundation, Inc.
+#
+# It doesn't change code that is known to be maintained elsewhere or
+# that carries a non-FSF copyright.
+#
+# Pass --this-year to the script if you want it to add the current year
+# to all applicable notices.  Pass --quilt if you are using quilt and
+# want files to be added to the quilt before being changed.
+#
+# By default the script will update all directories for which the
+# output has been vetted.  You can instead pass the names of individual
+# directories, including those that haven't been approved.  So:
+#
+#    update-copyright.pl --this-year
+#
+# is the command that would be used at the beginning of a year to update
+# all copyright notices (and possibly at other times to check whether
+# new files have been added with old years).  On the other hand:
+#
+#    update-copyright.pl --this-year libjava
+#
+# would run the script on just libjava/.
+#
+# This script was copied from gcc's contrib/ and modified to suit
+# binutils.  In contrast to the gcc script, this one will update
+# the testsuite and --version output strings too.
+
+import os
+import re
+import sys
+import time
+import subprocess
+
+class Errors:
+    def __init__ (self):
+        self.num_errors = 0
+
+    def report (self, filename, string):
+        if filename:
+            string = filename + ': ' + string
+        sys.stderr.write (string + '\n')
+        self.num_errors += 1
+
+    def ok (self):
+        return self.num_errors == 0
+
+class GenericFilter:
+    def __init__ (self):
+        self.skip_files = set()
+        self.skip_dirs = set()
+        self.skip_extensions = set()
+        self.fossilised_files = set()
+        self.own_files = set()
+
+        self.skip_files |= set ([
+                # Skip licence files.
+                'COPYING',
+                'COPYING.LIB',
+                'COPYING3',
+                'COPYING3.LIB',
+                'COPYING.LIBGLOSS',
+                'COPYING.NEWLIB',
+                'LICENSE',
+                'fdl.texi',
+                'gpl_v3.texi',
+                'fdl-1.3.xml',
+                'gpl-3.0.xml',
+
+                # Skip auto- and libtool-related files
+                'aclocal.m4',
+                'compile',
+                'config.guess',
+                'config.sub',
+                'depcomp',
+                'install-sh',
+                'libtool.m4',
+                'ltmain.sh',
+                'ltoptions.m4',
+                'ltsugar.m4',
+                'ltversion.m4',
+                'lt~obsolete.m4',
+                'missing',
+                'mkdep',
+                'mkinstalldirs',
+                'move-if-change',
+                'shlibpath.m4',
+                'symlink-tree',
+                'ylwrap',
+
+                # Skip FSF mission statement, etc.
+                'gnu.texi',
+                'funding.texi',
+                'appendix_free.xml',
+
+                # Skip imported texinfo files.
+                'texinfo.tex',
+                ])
+
+        self.skip_extensions |= set ([
+                # Maintained by the translation project.
+                '.po',
+
+                # Automatically-generated.
+                '.pot',
+                ])
+
+        self.skip_dirs |= set ([
+                'autom4te.cache',
+                ])
+
+
+    def get_line_filter (self, dir, filename):
+        if filename.startswith ('ChangeLog'):
+            # Ignore references to copyright in changelog entries.
+            return re.compile ('\t')
+
+        return None
+
+    def skip_file (self, dir, filename):
+        if filename in self.skip_files:
+            return True
+
+        (base, extension) = os.path.splitext (os.path.join (dir, filename))
+        if extension in self.skip_extensions:
+            return True
+
+        if extension == '.in':
+            # Skip .in files produced by automake.
+            if os.path.exists (base + '.am'):
+                return True
+
+            # Skip files produced by autogen
+            if (os.path.exists (base + '.def')
+                and os.path.exists (base + '.tpl')):
+                return True
+
+        # Skip configure files produced by autoconf
+        if filename == 'configure':
+            if os.path.exists (base + '.ac'):
+                return True
+            if os.path.exists (base + '.in'):
+                return True
+
+        return False
+
+    def skip_dir (self, dir, subdir):
+        return subdir in self.skip_dirs
+
+    def is_fossilised_file (self, dir, filename):
+        if filename in self.fossilised_files:
+            return True
+        # Only touch current current ChangeLogs.
+        if filename != 'ChangeLog' and filename.find ('ChangeLog') >= 0:
+            return True
+        return False
+
+    def by_package_author (self, dir, filename):
+        return filename in self.own_files
+
+class Copyright:
+    def __init__ (self, errors):
+        self.errors = errors
+
+        # Characters in a range of years.  Include '.' for typos.
+        ranges = '[0-9](?:[-0-9.,\s]|\s+and\s+)*[0-9]'
+
+        # Non-whitespace characters in a copyright holder's name.
+        name = '[\w.,-]'
+
+        # Matches one year.
+        self.year_re = re.compile ('[0-9]+')
+
+        # Matches part of a year or copyright holder.
+        self.continuation_re = re.compile (ranges + '|' + name)
+
+        # Matches a full copyright notice:
+        self.copyright_re = re.compile (
+            # 1: 'Copyright (C)', etc.
+            '([Cc]opyright'
+            '|[Cc]opyright\s+\([Cc]\)'
+            '|[Cc]opyright\s+%s'
+            '|[Cc]opyright\s+&copy;'
+            '|[Cc]opyright\s+@copyright{}'
+            '|@set\s+copyright[\w-]+)'
+
+            # 2: the years.  Include the whitespace in the year, so that
+            # we can remove any excess.
+            '(\s*(?:' + ranges + ',?'
+            '|@value\{[^{}]*\})\s*)'
+
+            # 3: 'by ', if used
+            '(by\s+)?'
+
+            # 4: the copyright holder.  Don't allow multiple consecutive
+            # spaces, so that right-margin gloss doesn't get caught
+            # (e.g. gnat_ugn.texi).
+            '(' + name + '(?:\s?' + name + ')*)?')
+
+        # A regexp for notices that might have slipped by.  Just matching
+        # 'copyright' is too noisy, and 'copyright.*[0-9]' falls foul of
+        # HTML header markers, so check for 'copyright' and two digits.
+        self.other_copyright_re = re.compile ('(^|[^\._])copyright[^=]*[0-9][0-9]',
+                                              re.IGNORECASE)
+        self.comment_re = re.compile('#+|[*]+|;+|%+|//+|@c |dnl ')
+        self.holders = { '@copying': '@copying' }
+        self.holder_prefixes = set()
+
+        # True to 'quilt add' files before changing them.
+        self.use_quilt = False
+
+        # If set, force all notices to include this year.
+        self.max_year = None
+
+        # Goes after the year(s).  Could be ', '.
+        self.separator = ' '
+
+    def add_package_author (self, holder, canon_form = None):
+        if not canon_form:
+            canon_form = holder
+        self.holders[holder] = canon_form
+        index = holder.find (' ')
+        while index >= 0:
+            self.holder_prefixes.add (holder[:index])
+            index = holder.find (' ', index + 1)
+
+    def add_external_author (self, holder):
+        self.holders[holder] = None
+
+    class BadYear():
+        def __init__ (self, year):
+            self.year = year
+
+        def __str__ (self):
+            return 'unrecognised year: ' + self.year
+
+    def parse_year (self, string):
+        year = int (string)
+        if len (string) == 2:
+            if year > 70:
+                return year + 1900
+        elif len (string) == 4:
+            return year
+        raise self.BadYear (string)
+
+    def year_range (self, years):
+        year_list = [self.parse_year (year)
+                     for year in self.year_re.findall (years)]
+        assert len (year_list) > 0
+        return (min (year_list), max (year_list))
+
+    def set_use_quilt (self, use_quilt):
+        self.use_quilt = use_quilt
+
+    def include_year (self, year):
+        assert not self.max_year
+        self.max_year = year
+
+    def canonicalise_years (self, dir, filename, filter, years):
+        # Leave texinfo variables alone.
+        if years.startswith ('@value'):
+            return years
+
+        (min_year, max_year) = self.year_range (years)
+
+        # Update the upper bound, if enabled.
+        if self.max_year and not filter.is_fossilised_file (dir, filename):
+            max_year = max (max_year, self.max_year)
+
+        # Use a range.
+        if min_year == max_year:
+            return '%d' % min_year
+        else:
+            return '%d-%d' % (min_year, max_year)
+
+    def strip_continuation (self, line):
+        line = line.lstrip()
+        match = self.comment_re.match (line)
+        if match:
+            line = line[match.end():].lstrip()
+        return line
+
+    def is_complete (self, match):
+        holder = match.group (4)
+        return (holder
+                and (holder not in self.holder_prefixes
+                     or holder in self.holders))
+
+    def update_copyright (self, dir, filename, filter, file, line, match):
+        orig_line = line
+        next_line = None
+        pathname = os.path.join (dir, filename)
+
+        intro = match.group (1)
+        if intro.startswith ('@set'):
+            # Texinfo year variables should always be on one line
+            after_years = line[match.end (2):].strip()
+            if after_years != '':
+                self.errors.report (pathname,
+                                    'trailing characters in @set: '
+                                    + after_years)
+                return (False, orig_line, next_line)
+        else:
+            # If it looks like the copyright is incomplete, add the next line.
+            while not self.is_complete (match):
+                try:
+                    next_line = file.next()
+                except StopIteration:
+                    break
+
+                # If the next line doesn't look like a proper continuation,
+                # assume that what we've got is complete.
+                continuation = self.strip_continuation (next_line)
+                if not self.continuation_re.match (continuation):
+                    break
+
+                # Merge the lines for matching purposes.
+                orig_line += next_line
+                line = line.rstrip() + ' ' + continuation
+                next_line = None
+
+                # Rematch with the longer line, at the original position.
+                match = self.copyright_re.match (line, match.start())
+                assert match
+
+            holder = match.group (4)
+
+            # Use the filter to test cases where markup is getting in the way.
+            if filter.by_package_author (dir, filename):
+                assert holder not in self.holders
+
+            elif not holder:
+                self.errors.report (pathname, 'missing copyright holder')
+                return (False, orig_line, next_line)
+
+            elif holder not in self.holders:
+                self.errors.report (pathname,
+                                    'unrecognised copyright holder: ' + holder)
+                return (False, orig_line, next_line)
+
+            else:
+                # See whether the copyright is associated with the package
+                # author.
+                canon_form = self.holders[holder]
+                if not canon_form:
+                    return (False, orig_line, next_line)
+
+                # Make sure the author is given in a consistent way.
+                line = (line[:match.start (4)]
+                        + canon_form
+                        + line[match.end (4):])
+
+                # Remove any 'by'
+                line = line[:match.start (3)] + line[match.end (3):]
+
+        # Update the copyright years.
+        years = match.group (2).strip()
+        if (self.max_year
+            and match.start(0) > 0 and line[match.start(0)-1] == '"'
+            and not filter.is_fossilised_file (dir, filename)):
+            # A printed copyright date consists of the current year
+            canon_form = '%d' % self.max_year
+        else:
+            try:
+                canon_form = self.canonicalise_years (dir, filename, filter, years)
+            except self.BadYear as e:
+                self.errors.report (pathname, str (e))
+                return (False, orig_line, next_line)
+
+        line = (line[:match.start (2)]
+                + ' ' + canon_form + self.separator
+                + line[match.end (2):])
+
+        # Use the standard (C) form.
+        if intro.endswith ('right'):
+            intro += ' (C)'
+        elif intro.endswith ('(c)'):
+            intro = intro[:-3] + '(C)'
+        line = line[:match.start (1)] + intro + line[match.end (1):]
+
+        # Strip trailing whitespace
+        line = line.rstrip() + '\n'
+
+        return (line != orig_line, line, next_line)
+
+    def process_file (self, dir, filename, filter):
+        pathname = os.path.join (dir, filename)
+        if filename.endswith ('.tmp'):
+            # Looks like something we tried to create before.
+            try:
+                os.remove (pathname)
+            except OSError:
+                pass
+            return
+
+        lines = []
+        changed = False
+        line_filter = filter.get_line_filter (dir, filename)
+        with open (pathname, 'r') as file:
+            prev = None
+            for line in file:
+                while line:
+                    next_line = None
+                    # Leave filtered-out lines alone.
+                    if not (line_filter and line_filter.match (line)):
+                        match = self.copyright_re.search (line)
+                        if match:
+                            res = self.update_copyright (dir, filename, filter,
+                                                         file, line, match)
+                            (this_changed, line, next_line) = res
+                            changed = changed or this_changed
+
+                        # Check for copyright lines that might have slipped by.
+                        elif self.other_copyright_re.search (line):
+                            self.errors.report (pathname,
+                                                'unrecognised copyright: %s'
+                                                % line.strip())
+                    lines.append (line)
+                    line = next_line
+
+        # If something changed, write the new file out.
+        if changed and self.errors.ok():
+            tmp_pathname = pathname + '.tmp'
+            with open (tmp_pathname, 'w') as file:
+                for line in lines:
+                    file.write (line)
+            if self.use_quilt:
+                subprocess.call (['quilt', 'add', pathname])
+            os.rename (tmp_pathname, pathname)
+
+    def process_tree (self, tree, filter):
+        for (dir, subdirs, filenames) in os.walk (tree):
+            # Don't recurse through directories that should be skipped.
+            for i in xrange (len (subdirs) - 1, -1, -1):
+                if filter.skip_dir (dir, subdirs[i]):
+                    del subdirs[i]
+
+            # Handle the files in this directory.
+            for filename in filenames:
+                if filter.skip_file (dir, filename):
+                    sys.stdout.write ('Skipping %s\n'
+                                      % os.path.join (dir, filename))
+                else:
+                    self.process_file (dir, filename, filter)
+
+class CmdLine:
+    def __init__ (self, copyright = Copyright):
+        self.errors = Errors()
+        self.copyright = copyright (self.errors)
+        self.dirs = []
+        self.default_dirs = []
+        self.chosen_dirs = []
+        self.option_handlers = dict()
+        self.option_help = []
+
+        self.add_option ('--help', 'Print this help', self.o_help)
+        self.add_option ('--quilt', '"quilt add" files before changing them',
+                         self.o_quilt)
+        self.add_option ('--this-year', 'Add the current year to every notice',
+                         self.o_this_year)
+
+    def add_option (self, name, help, handler):
+        self.option_help.append ((name, help))
+        self.option_handlers[name] = handler
+
+    def add_dir (self, dir, filter = GenericFilter()):
+        self.dirs.append ((dir, filter))
+
+    def o_help (self, option = None):
+        sys.stdout.write ('Usage: %s [options] dir1 dir2...\n\n'
+                          'Options:\n' % sys.argv[0])
+        format = '%-15s %s\n'
+        for (what, help) in self.option_help:
+            sys.stdout.write (format % (what, help))
+        sys.stdout.write ('\nDirectories:\n')
+
+        format = '%-25s'
+        i = 0
+        for (dir, filter) in self.dirs:
+            i += 1
+            if i % 3 == 0 or i == len (self.dirs):
+                sys.stdout.write (dir + '\n')
+            else:
+                sys.stdout.write (format % dir)
+        sys.exit (0)
+
+    def o_quilt (self, option):
+        self.copyright.set_use_quilt (True)
+
+    def o_this_year (self, option):
+        self.copyright.include_year (time.localtime().tm_year)
+
+    def main (self):
+        for arg in sys.argv[1:]:
+            if arg[:1] != '-':
+                self.chosen_dirs.append (arg)
+            elif arg in self.option_handlers:
+                self.option_handlers[arg] (arg)
+            else:
+                self.errors.report (None, 'unrecognised option: ' + arg)
+        if self.errors.ok():
+            if len (self.chosen_dirs) == 0:
+                self.chosen_dirs = self.default_dirs
+            if len (self.chosen_dirs) == 0:
+                self.o_help()
+            else:
+                for chosen_dir in self.chosen_dirs:
+                    canon_dir = os.path.join (chosen_dir, '')
+                    count = 0
+                    for (dir, filter) in self.dirs:
+                        if (dir + os.sep).startswith (canon_dir):
+                            count += 1
+                            self.copyright.process_tree (dir, filter)
+                    if count == 0:
+                        self.errors.report (None, 'unrecognised directory: '
+                                            + chosen_dir)
+        sys.exit (0 if self.errors.ok() else 1)
+
+#----------------------------------------------------------------------------
+
+class TopLevelFilter (GenericFilter):
+    def skip_dir (self, dir, subdir):
+        return True
+
+class ConfigFilter (GenericFilter):
+    def __init__ (self):
+        GenericFilter.__init__ (self)
+
+    def skip_file (self, dir, filename):
+        if filename.endswith ('.m4'):
+            pathname = os.path.join (dir, filename)
+            with open (pathname) as file:
+                # Skip files imported from gettext.
+                if file.readline().find ('gettext-') >= 0:
+                    return True
+        return GenericFilter.skip_file (self, dir, filename)
+
+class LdFilter (GenericFilter):
+    def __init__ (self):
+        GenericFilter.__init__ (self)
+
+        self.skip_extensions |= set ([
+                # ld testsuite output match files.
+                '.ro',
+                ])
+
+class BinutilsCopyright (Copyright):
+    def __init__ (self, errors):
+        Copyright.__init__ (self, errors)
+
+        canon_fsf = 'Free Software Foundation, Inc.'
+        self.add_package_author ('Free Software Foundation', canon_fsf)
+        self.add_package_author ('Free Software Foundation.', canon_fsf)
+        self.add_package_author ('Free Software Foundation Inc.', canon_fsf)
+        self.add_package_author ('Free Software Foundation, Inc', canon_fsf)
+        self.add_package_author ('Free Software Foundation, Inc.', canon_fsf)
+        self.add_package_author ('The Free Software Foundation', canon_fsf)
+        self.add_package_author ('The Free Software Foundation, Inc.', canon_fsf)
+        self.add_package_author ('Software Foundation, Inc.', canon_fsf)
+
+        self.add_external_author ('Carnegie Mellon University')
+        self.add_external_author ('John D. Polstra.')
+        self.add_external_author ('Linaro Ltd.')
+        self.add_external_author ('MIPS Computer Systems, Inc.')
+        self.add_external_author ('Red Hat Inc.')
+        self.add_external_author ('Regents of the University of California.')
+        self.add_external_author ('The Regents of the University of California.')
+        self.add_external_author ('Third Eye Software, Inc.')
+        self.add_external_author ('Ulrich Drepper')
+        self.add_external_author ('Synopsys Inc.')
+
+class BinutilsCmdLine (CmdLine):
+    def __init__ (self):
+        CmdLine.__init__ (self, BinutilsCopyright)
+
+        self.add_dir ('.', TopLevelFilter())
+        self.add_dir ('bfd')
+        self.add_dir ('binutils')
+        self.add_dir ('config', ConfigFilter())
+        self.add_dir ('cpu')
+        self.add_dir ('elfcpp')
+        self.add_dir ('etc')
+        self.add_dir ('gas')
+        self.add_dir ('gdb')
+        self.add_dir ('gold')
+        self.add_dir ('gprof')
+        self.add_dir ('include')
+        self.add_dir ('ld', LdFilter())
+        self.add_dir ('libdecnumber')
+        self.add_dir ('libiberty')
+        self.add_dir ('opcodes')
+        self.add_dir ('readline')
+        self.add_dir ('sim')
+
+        self.default_dirs = [
+            'bfd',
+            'binutils',
+            'elfcpp',
+            'etc',
+            'gas',
+            'gold',
+            'gprof',
+            'include',
+            'ld',
+            'libiberty',
+            'opcodes',
+            ]
+
+BinutilsCmdLine().main()