bibmerge

#! /usr/bin/perl

# bibmerge - merge new BibTeX entries into a BibTeX database
#
# Usage: bibmerge <newfile.bib> <database.bib>
#
# newfile.bib is merged into database.bib; the merged entries
# are resorted by alphabetic order of keys; the merged entries
# are saved back into database.bib.
#
# Guaranteed not to clobber anything in <database.bib>; it can
# only grow.
#
# DIAGNOSTICS:
#   Warnings are printed about collisions; assign these new keys
#   in newfile.bib and remerge.
#

$usage  = "bibmerge <newfile.bib> <database.bib>";
if ($#ARGV != 1) { die("Usage: $usage\n"); }

$newfile  = shift;
$database = shift;



				# keep a backup of the database
open(BACKUP,">$database.bck") || die("Can't create backup database $database.bck");
open(DATABASE,$database) || die("Can't open database $database");
while (<DATABASE>)
{
    print BACKUP $_;
}
close(BACKUP);
close(DATABASE);

$/ = "";
open(DATABASE,$database) || die("Can't open database $database");
while (<DATABASE>)
{
    if (/\@string/)              # Journal abbrevs - add to preamble
    { push(@preamble,$_); }
    
    elsif (/\@Article\{TitlesOn,/) # TDS's hack - add to preamble
    { push(@preamble,$_); }

    elsif (/\@.+\{\s*(\S+),/)	# Entries look like @Article{foo99, on  first line
    {				# Save 'em in an associative array.
	$key = $1;		
	if ($entlist{$key} ne "") 
	{ print "COLLISION in database on key $1:\n$_"; } # if already is such a key
	else                     
	{ $entlist{$key} = $_; }                          # if unique key
    }

    else			# if we don't recognize it, just add to preamble
    {  push(@preamble,$_); }
} 
close(DATABASE);

open(NEWDATA,$newfile) || die("Can't open new BibTeX data $newfile");
while (<NEWDATA>)
{
    if (/\@string/)              # Journal abbrevs - ignore them and only keep master
    { next; }	
    
    elsif (/\@Article\{TitlesOn,/) # ignore TDS's hack in its place, only keep master
    { next; }			

    elsif (/\@.+\{\s*(\S+),/)	# Entries look like @Article{foo99, on their first line
    {				# Save 'em in an associative array.
	$key = $1;		
	if ($entlist{$key} ne "") 
	{ print "COLLISION on key $key\n(New database will not include this entry)\n$_"; }
	else
	{ $entlist{$key} = $_; } # if unique key
    }				
}

				# now sort by the keys, and save the sorted entries
				# under the old database name

open(DATABASE,">$database");
print DATABASE @preamble, "\n";
foreach $key (sort keys(%entlist))
{
    print DATABASE $entlist{$key};
}
close(DATABASE);