view lib-src/grep-changelog @ 72917:17942cb3949e

(allout-regexp, allout-line-boundary-regexp) (allout-bob-regexp): Correct grouping and boundaries to fix backwards traversal. (allout-depth-specific-regexp, allout-depth-one-regexp): New versions that exploit \\{M\\} regexp syntax, to avoid geometric or worse time in allout-ascend. (allout-doublecheck-at-and-shallower): Identify depth threshold below which topics are checked for and disqualified by containment discontinuities. (allout-hotspot-key-handler): Correctly handle multiple-key strokes. Remove some unused variables. (allout-mode-leaders): Clarify that mode-specific comment-start will be used (set-allout-regexp): Correctly regexp-quote allout regexps to properly accept alternative header-leads and primary bullets with regexp-specific characters (eg, C "/*", mathematica "(*"). Include new regular expressions among those configured. (allout-infer-header-lead-and-primary-bullet): Rename allout-infer-header-lead. (allout-recent-depth): Manifest as a variable as well as a function. (allout-prefix-data): Simplify into an inline instead of a macro, assuming current match data rather than being explicitly passed it. Establish allout-recent-depth value as well as allout-recent-prefix-beginning and allout-recent-prefix-end. (allout-aberrant-container-p): True when an item's immediate offspring discontinuously contained. Useful for disqualifying unintended topic prefixes, likely at low depths. (allout-goto-prefix-doublechecked): Elaborated version of allout-goto-prefix which disqualifies aberrant pseudo-items. (allout-pre-next-prefix): Layer on top of lower-level routines, to get disqualification of aberrant containers. (allout-end-of-prefix, allout-end-of-subtree): Disqualify aberrant containers. (allout-beginning-of-current-entry): Position at start of buffer when in container (depth 0) entry. (nullify-allout-prefix-data): Invalidate allout-recent-* prefix data. (allout-current-bullet): Strip text properties. (allout-get-prefix-bullet): Use right match groups. (allout-beginning-of-line, allout-next-heading): Disqualify aberrant containers. (allout-previous-heading): Disqualify aberrant containers, and change to regular (rather than inline) function, to allow self-recursion. (allout-get-invisibility-overlay): Increment so progress is made when the first overlay is not the sought one. (allout-end-of-prefix): Disqualify aberrant containers. (allout-end-of-line): Cycle something like allout-beginning-of-line. (allout-mode): Make allout-old-style-prefixes (ie, enabling use with outline.el outlines) functional again. Change the primary bullet along with the header-lead - level 1 new-style bullets now work. Engage allout-before-change-handler in mainline emacs, not just xemacs, to do undo handling. (allout-before-change-handler): Expose undo changes occurring in hidden regions. Use allout-get-invisibility-overlay instead of reimplementing it inline. (allout-chart-subtree): Use start rather than end of prefix in charts. Use allout-recent-depth variable. (allout-chart-siblings): Disqualify aberrant topics. (allout-beginning-of-current-entry): Position correctly. (allout-ascend): Use new allout-depth-specific-regexp and allout-depth-one-regexp for linear instead of O(N^2) or worse behavior. (allout-ascend-to-depth): Depend on allout-ascend, rather than reimplementing an algorithm. (allout-up-current-level): Depend on allout-ascend, rather than reimplementing an algorithm. Return to start-point if we fail. (allout-descend-to-depth): Use allout-recent-depth variable instead of function. (allout-next-sibling): On traversal of numerous intervening topics, resort to economical allout-next-sibling-leap. (allout-next-sibling-leap): Specialized version of allout-next-sibling that uses allout-ascend cleverly, to depend on a regexp search to leap large numbers of contained topics, rather than arbitrarily many one-by-one traversals. (allout-next-visible-heading): Disqualify aberrant topics. (allout-previous-visible-heading): Position consistently when interactive. (allout-forward-current-level): Base on allout-previous-sibling rather than (differently) reimplmenting the algorithm. Remove some unused variables. (allout-solicit-alternate-bullet): Present default choice stripped of text properties. (allout-rebullet-heading): Use bullet stripped of text properties. Register changes using allout-exposure-change-hook. Disregard aberrant topics. (allout-shift-in): With universal-argument, make topic a peer of it's former offspring. Simplify the code by separating out allout-shift-out functionality. (allout-shift-out): With universal-argument, make offspring peers of their former container, and its siblings. Implement the functionality here, rather than inappropriately muddling the implementation of allout-shift-in. (allout-rebullet-topic): Respect additional argument for new parent-child separation function. (allout-yank-processing): Use allout-ascend directly. (allout-show-entry): Disqualify aberrant topics. (allout-show-children): Handle discontinuous children gracefully, extending the depth being revealed to expose them and posting a message indicating the situation. (allout-show-to-offshoot): Remove obsolete and incorrect comment. Leave cursor in correct position. (allout-hide-current-subtree): Use allout-ascend directly. Disqualify aberrant topics. (allout-kill-line, allout-kill-topic): Preserve exposure layout in a way that the yanks can restore it, as used to happen. (allout-yank-processing): Restore exposure layout as recorded by allout-kill-*, as used to happen. (allout-annotate-hidden, allout-hide-by-annotation): New routines for preseving and restoring exposure layout across kills. (allout-toggle-subtree-encryption): Run allout-exposure-change-hook. (allout-encrypt-string): Strip text properties. Rearranged order and outline-headings for some of the miscellaneous functions. (allout-resolve-xref): No need to quote the error name in the condition-case handler section. (allout-flatten): Classic recursive (and recursively intensive, without tail-recursion) list-flattener, needed by allout-shift-out when confronted with discontinuous children.
author Eli Zaretskii <eliz@gnu.org>
date Sat, 16 Sep 2006 10:24:24 +0000
parents 3661e9b3c48f
children 3e566ec9ef5d c5406394f567
line wrap: on
line source

#! /usr/bin/perl

# Copyright (C) 1999, 2000, 2001, 2002, 2003, 2004,
#               2005, 2006 Free Software Foundation, Inc.
#
# This file is part of GNU Emacs.
#
# GNU Emacs 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 2, or (at your option)
# any later version.
#
# GNU Emacs is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with GNU Emacs; see the file COPYING.  If not, write to the
# Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
# Boston, MA 02110-1301, USA.


# Extract entries from ChangeLogs matching specified criteria.
# Optionally format the resulting output to a form suitable for RCS
# logs, like they are used in Emacs, for example.  In this format,
# author lines, leading spaces, and file names are removed.

require 5;
use strict;

# Parse command line options.

use vars qw($author $regexp $exclude $from_date $to_date
            $rcs_log $with_date $version $help $reverse
            @entries);

use Getopt::Long;
my $result = GetOptions ("author=s" => \$author,
                         "text=s"  => \$regexp,
                         "exclude=s"  => \$exclude,
                         "from-date=s" => \$from_date,
                         "to-date=s" => \$to_date,
                         "rcs-log" => \$rcs_log,
                         "with-date" => \$with_date,
                         "reverse!" => \$reverse,
                         "version" => \$version,
                         "help"    => \$help);

# If date options are specified, check that they have the format
# YYYY-MM-DD.

$result = 0 if $from_date && $from_date !~ /^\d\d\d\d-\d\d-\d\d$/;
$result = 0 if $to_date && $to_date !~ /^\d\d\d\d-\d\d-\d\d$/;

# Print usage information and exit when necessary.

if ($result == 0 || $help) {
    print <<USAGE;

Usage: $0 [options] [CHANGELOG...]

Print entries in ChangeLogs matching various criteria.
Valid options are:

  --author=AUTHOR         Match entries whose author line matches
			  regular expression AUTHOR
  --text=TEXT             Match entries whose text matches regular
			  expression TEXT
  --exclude=TEXT	  Exclude entries matching TEXT
  --from-date=YYYY-MM-DD  Match entries not older than given date
  --to-date=YYYY-MM-DD    Match entries not younger than given date
  --rcs-log		  Format output suitable for RCS log entries
  --with-date		  Print short date line in RCS log
  --reverse               Show entries in reverse (chronological) order
  --version		  Print version info
  --help		  Print this help

If no CHANGELOG is specified scan the files "ChangeLog" and
"ChangeLog.1+" in the current directory.  Old-style dates in ChangeLogs
are not recognized.
USAGE
    exit !$help;
}

# Print version info and exit if `--version' was specified.

if ($version) {
    print "0.2\n";
    exit 0;
}


# Value is non-zero if HEADER matches according to command line
# options specified, i.e. it matches $author, and its date is in
# the range $from_date <= date <= $to_date.

sub header_match_p {
    my $header = shift;

    return 0 unless $header;

    # No match if AUTHOR-regexp specified and doesn't match.
    return 0 if $author && $header !~ /$author/;

    # Check that the date of the entry matches if date options
    # `--from-date' and/or `--to-date' were specified .  Old-style
    # dates in ChangeLogs are not recognized, and never match.
    if ($from_date || $to_date) {
	if ($header =~ /^(\d\d\d\d-\d\d-\d\d)/) {
	    my $date = $1;
	    return 0 if $from_date && $date lt $from_date;
	    return 0 if $to_date && $date gt $to_date;
	} else {
	    # Don't bother recognizing old-style dates.
	    return 0;
	}
    }

    return 1;
}


# Value is non-zero if ENTRY matches the criteria specified on the
# command line, i.e. it matches $regexp, and it doesn't match
# $exclude.

sub entry_match_p {
    my $entry = shift;

    return 0 unless $entry;

    if ($regexp) {
	return 1 if ($entry =~ /$regexp/
		     && (!$exclude || $entry !~ $exclude));
    } else {
	return 1 if !$exclude || $entry !~ $exclude;
    }

    return 0;
}


# Print HEADER and/or ENTRY in a format suitable for what was
# specified on the command line.  If $rcs_log is specified, author
# lines are not printed, and leading spaces and file names are removed
# from ChangeLog entries.

sub print_log {
    my ($header, $entry) = @_;
    my $output = '';

    if ($rcs_log) {
	# Remove leading whitespace from entry.
	$entry =~ s/^\s+//mg;
	# Remove file name parts.
	$entry =~ s/^\*.*\(/(/mg;
	# Remove file name parts, 2.
	$entry =~ s/^\*.*://mg;
        if ($with_date) {
	    $header =~ /(\d\d\d\d-\d\d-\d\d)/;
	    $output = "!changelog-date $1\n";
	}
	$output .= $entry;
    } else {
	$output .= $header . $entry;
    }

    if ($reverse) {
        push @entries, $output;
    } else {
        print $output;
    }
}

# Scan LOG for matching entries, and print them to standard output.

sub parse_changelog {
    my $log = shift;
    my $entry = undef;
    my $header = undef;

    @entries = () if $reverse;

    # Open the ChangeLog.
    open (IN, "< $log") || die "Cannot open $log: $!";

    while (defined(my $line = <IN>)) {
	if ($line =~ /^\S/) {
	    # Line is an author-line.  Print previous entry if
	    # it matches.
	    print_log ($header, $entry)
		if header_match_p ($header) && entry_match_p ($entry);

	    $entry = "";
  	    $header = $line;

	    # Add empty lines below the header.
	    while (defined($line = <IN>) && $line =~ /^\s*$/) {
	        $header = "$header$line";
	    }
        }

        last unless defined $line;

	if ($line =~ /^\s*\*/) {
	    # LINE is the first line of a ChangeLog entry.  Print
	    # previous entry if it matches.
	    print_log ($header, $entry)
		if header_match_p ($header) && entry_match_p ($entry);
	    $entry = $line;
	} else {
	    # Add LINE to the current entry.
	    $entry = "$entry$line";
	}
    }

    # Print last entry if it matches.
    print_log ($header, $entry)
	if header_match_p ($header) && entry_match_p ($entry);

    close IN;

    if ($reverse) {
        for (my $entry = @entries; $entry; $entry--) {
            print $entries[$entry-1];
        }
    }
}


# Main program.  Process ChangeLogs.

# If files were specified on the command line, parse those files in the
# order supplied by the user; otherwise parse default files ChangeLog and
# ChangeLog.1+ according to $reverse.
unless (@ARGV > 0) {
    @ARGV = ("ChangeLog");

    push @ARGV,
      map {"ChangeLog.$_"}
        sort {$b <=> $a}
          map {/\.(\d+)$/; $1}
            do {
                opendir D, '.';
                grep /^ChangeLog\.\d+$/, readdir D;
            };

    @ARGV = reverse @ARGV if $reverse;
}

while (defined (my $log = shift @ARGV)) {
    parse_changelog ($log) if -f $log;
}


# arch-tag: 9e4f6749-e053-4bb7-b3ad-11947318418e
# grep-changelog ends here.