#!/usr/bin/perl

###############################################################################
# 
#  When run without any arguments, this script reads the CCEL files:
#  Bible.*.xml.
#  
#  Otherwise, supply the script with a list (or wildcard) of VPL files and it
#  will attempt to create a v11n system definition on that basis.
#  
#  If the --nosis switch is added as the first argument, the script will
#  interpret book abbreviations rather than simply assume that they are OSIS 
#  abbreviations.
#
###############################################################################


sub buildBooks {
    if ($lastBook ne "" && $osis{lc($lastBook)} eq "") {
	print "ERROR: Unknown book abbreviation: $lastBook in $v11n\n";
    }
    if ($osis{lc($lastBook)} eq "Matt") {
	$otnt = 1;
    }

    $osisBook = $osis{lc($lastBook)};
    if ($otnt == 0) {
	$otbooks .= "\t{\"$idmap{$osisBook}\", \"$osisBook\", \"$osisBook\", $cCount},\n";
    }
    else {
	$ntbooks .= "\t{\"$idmap{$osisBook}\", \"$osisBook\", \"$osisBook\", $cCount},\n";
    }
}

# @canons will contain this list of files, these are in a basic XML format.
# Each file lists osisIDs along with the English names associated with the
# osisID. These aren't exhaustive, and may or may not overlap (but hopefully
# don't). We are only using these to load mappings from osisIDs.
@canons = (
    "canon.bible.xml",      # the Bible, broadly defined
    "canon.af.xml",         # Apostolic Fathers
    "canon.otp.xml",        # OT pseudepigrapha
    "canon.nta.xml",        # NT apocrypha
    "canon.lds.xml",        # Mormon books
    "canon.naghammadi.xml", # Nag Hammadi Library
    "canon.qumran.xml",     # Qumran mss
    "canon.classical.xml",  # intended for classical works, currently just Josephus
);

foreach $mapfile (@canons) {
    open MAP, "$mapfile";
    while (<MAP>) {
	$line = $_;
	
	$line =~ s/<!\-\-.+?\-\->//g;
	$line =~ s/\&amp;/\&/g;

	if ($line =~ /<id>(.+?)<\/id>/) {
	    $id = $1;
	    $osis{lc($id)} = $id;
	}
	elsif ($line =~ /<abbr>(.+?)<\/abbr>/) {
	    $name = $1;
	    if ($osis{lc($name)} eq "") {
		$osis{lc($name)} = $id;
	    }
	    else {
		print "ERROR: Duplicate mapping from $id found in $mapfile (<abbr>).\n";
	    }
	}
	elsif ($line =~ /<name>(.+?)<\/name>/) {
	    $name = $1;
	    if ($osis{lc($name)} eq "") {
		$osis{lc($name)} = $id;
	    }
	    else {
		print "ERROR: Duplicate mapping from $id found in $mapfile (<name>).\n";
	    }

	    if ($idmap{$id} eq "") {
		$idmap{$id} = $name;
	    }
	    else {
		# Duplicates most likely indicate alternate names, so ignore them.
#		print "ERROR: Duplicate mapping from $id found in $mapfile.\n";
	    }
	}
    }
    close (MAP);
}
die();

if (@ARGV[0] ne "") {
    $n = 0;
    $mode = "xml";
    while (@ARGV[$n] ne "") {
	if ($n == 0 && @ARGV[0] eq "--vpl") {
	    $mode = "vpl";
	}
	elsif ($n == 0 && @ARGV[0] eq "--imp") {
	    $mode = "imp";
	}
	elsif ($n == 0 && @ARGV[0] eq "--xml") {
	    $mode = "xml";
	}
	else {
	    push @srcfiles, @ARGV[$n];
	}
	$n++;
    }
}
else {
    opendir (DIR, ".");
    @srcfiles = grep /Bible\.[^\.]+\.xml$/, readdir DIR;
    closedir(DIR);
    $mode = "xml";
}

foreach $infile (@srcfiles) {
    
    if ($infile =~ /^Bible.+xml$/) {
	$infile =~ /^Bible\.([^\.]+)\.xml/;
	$v11n = $1;
	$outfile = lc("v11n$1.h");
    }
    else {
	$v11n = $infile;
	$v11n =~ s/\.[^\.]*$//;
	$v11n =~ s/^.+\///;
	$outfile = lc("v11n$v11n.h");
    }

    open INF, $infile;
    open OUTF, ">$outfile";

    print OUTF "/******************************************************************************\n * $outfile.h - Versification data for the $v11n system\n *\n * Copyright 1998-2005 CrossWire Bible Society (http://www.crosswire.org)\n *	CrossWire Bible Society\n *	P. O. Box 2528\n *	Tempe, AZ  85280-2528\n *\n * This program is free software; you can redistribute it and/or modify it\n * under the terms of the GNU General Public License as published by the\n * Free Software Foundation version 2.\n *\n * This program is distributed in the hope that it will be useful, but\n * WITHOUT ANY WARRANTY; without even the implied warranty of\n * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU\n * General Public License for more details.\n *\n";
    print OUTF "\n#ifndef V11N" . uc($v11n) . "_H\n#define V11N" . uc($v11n) . "_H\n\nSWORD_NAMESPACE_START\n\n";

    $thisBook = "";
    $thisChap = "";
    $thisVers = "";

    $bCount = 0;
    $cCount = 0;
    $vCount = 0;
    
    $vm = "/******************************************************************************\n *	Maximum verses per chapter\n */\n\nint vm$v11n\[\] = {";
    $otbooks = "/******************************************************************************\n * [on]tbooks$v11n - initialize static instance for all canonical text names\n *		and chapmax\n */\nstruct sbook otbooks$v11n\[\] = {\n";
    $ntbooks = "struct sbook ntbooks$v11n\[\] = {\n";
    $booksCloser = "\t{\"\", \"\", \"\", 0}\n};\n\n";
    
    $otnt = 0; # 0 = ot, 1 = nt
    
    while (<INF>) {
	$line = $_;
		
	$osisID = "";

	if ($line =~ /<osisID.+?code=\"(.+?)\"\/>/) {
	    $osisID = $1;
	}
	elsif ($mode eq "vpl" && $line =~ /^(.+?) ([0-9A-Za-z]+):([0-9A-Za-z]+)/) {
	    $osisID = "$1.$2.$3";
	}
	elsif ($mode eq "imp" && $line =~ /^\$\$\$(.+?) ([0-9A-Za-z]+):([0-9A-Za-z]+)/) {
	    $osisID = "$1.$2.$3";
	}

	if ($osisID ne "") {
	    $lastBook = $thisBook;
	    $lastChap = $thisChap;
	    $lastVers = $thisVers;
	    
	    $osisID =~ /([^\.]+)\.([a-zA-Z\d]+)\.([a-zA-Z\d]+)/;
	    
	    $thisBook = $1;
	    $thisChap = $2;
	    $thisVers = $3;
	    
	    if ((($thisBook ne $lastBook) ||($thisChap ne $lastChap)) && $lastVers ne "") {
		$vm .= "$vCount, ";
	    }

	    if ($thisBook ne $lastBook) {
		$bCount++;
		$vm .= "\n\t// $idmap{$osis{lc($thisBook)}}\n\t";
		if ($lastBook ne "") {
		    buildBooks{};
		}
		$cCount = 1;
		$vCount = 1;
	    }
	    elsif ($thisChap ne $lastChap) {
		$cCount++;
		$vCount = 1;
	    }
	    elsif ($thisVers ne $lastVers) {
		$vCount++;
	    }
	}
    }
    $vm .= "$vCount\n};\n";
    buildBooks();
    $otbooks .= $booksCloser;
    $ntbooks .= $booksCloser;
    
    print OUTF $otbooks;
    print OUTF $ntbooks;
    print OUTF $vm;
    
    print OUTF "\nSWORD_NAMESPACE_END\n\n#endif\n";
    
    close (INF);
    close (OUTF);
}