#!/usr/bin/env perl =pod =head1 NAME bbl2bib.pl - convert thebibliography environment to a bib file =head1 SYNOPSIS bbl2bib.pl [B<-o> I] I =head1 OPTIONS =over 4 =item B<-o> I Output file. If this option is not used, the name for the output file is formed by changing the extension to C<.bib> =back =head1 DESCRIPTION The script tries to reconstruct a C file from the corresponding C environment. One can argue that this operation is akin to reconstructing a cow from the steak. The way the script does it is searching for the entry in the MR database, and creating the corresponding BibTeX fields. The script reads a TeX or Bbl file and extracts from it the C environment. For each bibitem it creates a plain text bibliography entry, and then tries to match it in the database. =head1 INPUT FILE We assume some structure of the input file: =over 4 =item 1. The bibliography is contained between the lines \begin{thebibliography}... and \end{thebibliography} =item 2. Each bibliography item starts from the line \bibitem[...]{....} =back =head1 EXAMPLES bbl2bib -o - file.tex > result.bib bbl2bib -o result.bib file.bbl bbl2bib file.tex =head1 AUTHOR Boris Veytsman =head1 COPYRIGHT AND LICENSE Copyright (C) 2014-2017 Boris Veytsman This is free software. You may redistribute copies of it under the terms of the GNU General Public License L. There is NO WARRANTY, to the extent permitted by law. =cut use strict; BEGIN { # find files relative to our installed location within TeX Live chomp(my $TLMaster = `kpsewhich -var-value=SELFAUTOPARENT`); # TL root if (length($TLMaster)) { unshift @INC, "$TLMaster/texmf-dist/scripts/bibtexperllibs"; } } use IO::File; use BibTeX::Parser; use FileHandle; use LaTeX::ToUnicode qw (convert); use Getopt::Std; use URI::Escape; use LWP::Simple; my $USAGE="USAGE: $0 [-o output] file\n"; my $VERSION = <new($inputfile) or die "Cannot find Bbl or TeX file $inputfile\n$USAGE\n"; my $output = IO::File->new("> $outputfile") or die "Cannot write to $outputfile\n$USAGE\n"; my $userAgent = LWP::UserAgent->new; # Bibitem is a hash with the entries 'key', 'text', 'mr', # 'zbl' my $bibitem; while (<$input>) { if (!(/\\begin\{thebibliography\}/../\\end\{thebibliography\}/) || /\\begin\{thebibliography\}/ || /\\end\{thebibliography\}/) { next; } if (/\\bibitem(\[[^\]]*\])?\{([^\}]*)\}/) { ProcessBibitem($bibitem); $bibitem = undef; $bibitem->{key}=$2; $bibitem->{text}=""; } if (!/^\s*$/) { $bibitem -> {text} .= $_; } } ProcessBibitem($bibitem); exit 0; sub ProcessBibitem { my $bibitem = shift; my $key = $bibitem->{key}; my $text=$bibitem->{text}; if (!length($text) || $text =~ /^\s+$/s) { return; } my $printtext = $text; $printtext =~ s/^(.)/% $1/mg; print $output "$printtext"; $text =~ s/\n/ /mg; $text =~ s/\\bibitem(\[[^\]]*\])?\{[^\}]*\}//; # Arxiv entry? if ($text =~ s/\\arxiv\{([^\}]+)\}\.?//) { $bibitem->{arxiv}=$1; } # Mr number exists? if ($text =~ s/\\mr\{([^\}]+)\}\.?//) { $bibitem->{mr}=$1; } # zbl number exists? if ($text =~ s/\\zbl\{([^\}]+)\}\.?//) { $bibitem->{zbl}=$1; } # doi number exists? if ($text =~ s/\\doi\{([^\}]+)\}\.?//) { $bibitem->{doi}=$1; } $bibitem->{bib} = SearchMref($bibitem); PrintBibitem($bibitem); return; } sub SearchMref { my $bibitem = shift; my $mirror = "http://www.ams.org/mathscinet-mref"; my $string=uri_escape_utf8($bibitem->{text}); my $response = $userAgent->get("$mirror?ref=$string&dataType=bibtex") -> decoded_content(); if ($response =~ /
(.*)<\/pre>/s) {
	my $bib= $1;
	my $fh = new FileHandle;
	open $fh, "<", \$bib;
	my $parser = new BibTeX::Parser($fh);
	my $entry = $parser->next;
	if ($entry->parse_ok()) {
	    $entry->key($bibitem->{key});
	    return ($entry);
	}
    }
}




sub PrintBibitem {
    print "\n";
    my $bibitem = shift;
    if (!ref($bibitem->{bib})) {
	return;
    }
    my $entry=$bibitem->{bib};
    if ($bibitem->{mr} && ! $entry->field('mrnumber')) {
	$entry->field('mrnumber', $bibitem->{mr});
    }
    if ($entry->field('mrnumber')) {
	my $mr=$entry->field('mrnumber');
	while (length($mr)<7) {
	    $mr = "0$mr";
	}
	$mr=$entry->field('mrnumber', $mr);
    }
    if ($bibitem->{zbl} && ! $entry->field('zblnumber')) {
	$entry->field('zblnumber', $bibitem->{zbl});
    }
    if ($bibitem->{doi} && ! $entry->field('doi')) {
	$entry->field('doi', $bibitem->{doi});
    }
    if ($bibitem->{arxiv} && ! $entry->field('arxiv')) {
	$entry->field('arxiv', $bibitem->{arxiv});
    }
	
    print $entry->to_string(), "\n\n";
}