Re: change bars

[Costa Sapuntzakis <csapuntz@cisco.com>]

My original script had bugs in it. Here is a newer version with better
revision tracking and hopefully fewer bugs. :-)

-Costa

On Wed, 5 Jul 2000, Matt Wakeley wrote:

> Hi Costa,
> 
> Thanks for the perl script.  It really helps.
> 
> But I noticed that it did not catch the "difference" between 0628 and 0706 of the
> changing of the "ping" and the new "map" commands in section 3.1.1 (these are "new"
> lines that it did not flag).
> 
> -Matt
> 
> Costa Sapuntzakis wrote:
> 
> > I have included the Perl script I used to generate the change bars.
> > The draft was generated as follows:
> >
> > perl create-change-bar.pl iSCSI000628.txt iscsi00615.txt > iSCSI000628.2.txt
> >
> > Hope it helps!
> >
> > -Costa
> 
> 
> 
#!

#
# The script operates in the following fashion. First, it
# strips all blank lines, headers, and footers from both
# document. 
#
# It then runs diff on the stripped files to find the
# differences.
#
# Finally, it takes the output of diff and generates the
# change bars at the destination.
#
# Note: the change bars only mark additions and changes
#
# -Costa (csapuntz@alum.mit.edu) 06/28/00

use strict;

use Getopt::Long;

my $lines_flag = 0;

sub usage {
    print STDERR "Usage:\nperl create-change-bar.pl [--lines] <file1> <file2>\n";
    print STDERR "file1 will be printed with change bars\n\n";
    print STDERR "--lines is a less exact match that may work better for\n";
    print STDERR "        large diffs\n";
    exit 1;
}

&GetOptions(lines => \$lines_flag);

&usage if ($#ARGV != 1);

my $src = $ARGV[0];
my $dst = $ARGV[1];

&usage if (!defined $src || !defined $dst);

open(SRC, $src) or die "Unabled to open $src: $!";

$src .= ".tmp";

if (-r $src) {
    print STDERR "Error: temporary file $src already exists\n";
    exit 1;
}

open(SRC2, ">$src")  or die "Unabled to open $src: $!";;

open(DST, $dst)  or die "Unabled to open $dst: $!";;

$dst .= ".tmp";

if (-r $dst) {
    print STDERR "Error: temporary file $dst already exists\n";
    exit 1;
}

open(DST2, ">$dst")  or die "Unabled to open $dst: $!";;


sub process_file {
    my ($linemap, $fh1, $fh2) = @_;

    # Strip leading change bars
    s/^|//;
    

    my $oldlineno = -1;
    my $newlineno = 0;
    my $state = 0;

    my $saveword ;
    my $savewordlineno = 0;
    
    while(<$fh1>) {
	chomp;
	
	$oldlineno++;

	if ($state == 1) {
	    $state = 0;
	    next;
	}
	
	if (/^\014/) {
	    $state = 1;
	    next;
	}
	next if (/\[Page/);
	next if (/^\s*$/);
	
	if ( $lines_flag ) {
	    $$linemap[$newlineno] = $oldlineno;
	    $newlineno++;
	    print $fh2 $_ . "\n";
	} else {
	    s/[\t\n\r]//g;
	    my @words = split;
	    
	    if ($saveword) {
		print $fh2 $saveword;
		undef $saveword;
	    }
	    
	    if ($#words >= 0 &&
		$words[$#words] =~ /-$/) {
		$saveword = substr(pop @words, 0, -1);
		$savewordlineno = $oldlineno;
	    }
	    
	    foreach my $word (@words) {
#	    print $word . "\n";
		print $fh2 $word . "\n";
		
		$$linemap[$newlineno] = $oldlineno;
		$newlineno++;
	    }
	}
    }

}

#
# Store mapping between stripped line numbers and original
# line numbers
#
my @linemap = ();
my @changes = ();

&process_file(\@linemap, \*SRC, \*SRC2);
&process_file([], \*DST, \*DST2);

close DST2;
close DST;
close SRC2;

#
# diff the two documents
#
open (DIFFS, "diff -w $dst $src |") or die "Unabled to run diff: $!";

while(<DIFFS>) {
#    print ;

    next if (!/^\d/);

    /[acd](\d+)(,(\d+))?/;
    
    next if (!defined $1);

    my $start = $1;
    my $end = $3;

    if (!defined $end) {
	$end = $start;
    }

    for (my $idx = $start - 1; $idx <= $end - 1; $idx++) {
	$changes[$linemap[$idx]] = 1;
    }
}

#
# Print out document, marking changed lines
#
seek SRC, 0, 0;

my $oldlineno = 0;

while (<SRC>) {
    # Strip leading change bars
    s/^|//;

    if ($changes[$oldlineno]) {
	print "|";
    } else {
	print " ";
    }

    print;

    $oldlineno++;
}

#
# Cleanup
#

unlink $src;
unlink $dst;