#!/usr/bin/perl

# tplxml 
# by Troy Hanson   27 Feb 2006
# convert between tpl and XML

# Copyright (c) 2005-2013, Troy Hanson     http://troydhanson.github.com/tpl/
# All rights reserved.
# 
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
# 
#   * Redistributions of source code must retain the above copyright
#     notice, this list of conditions and the following disclaimer.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS
# IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
# TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
# PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
# OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
# LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
# NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
# SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

use strict;
use warnings;
use XML::Parser;
use FindBin;
use lib "$FindBin::Bin";  #locate Tpl.pm in same directory as tplxml
use Tpl;
use bytes;

sub quote_chars {
    my $str = shift;
    $$str =~ s/&/&/g;  #order matters
    $$str =~ s/</&lt;/g;
    $$str =~ s/>/&gt;/g;
}
sub unquote_chars {
    my $str = shift;
    $$str =~ s/&lt;/</g;
    $$str =~ s/&gt;/>/g;
    $$str =~ s/&amp;/&/g;
}
sub hex_chars {
    my $str = shift;
    my $hex;
    for(my $i=0; $i < length $$str; $i++) {
        my $byte = unpack("C",substr($$str,$i,1));
        $hex .= sprintf("%02x", $byte);
    }
    $$str = $hex;
}
sub unhex_chars {
    my $str = shift;
    my $bytes;
    for(my $i=0; $i < length $$str; $i+=2) {
        my $hexbyte = substr($$str,$i,2);
        $bytes .= pack("C", hex($hexbyte));
    }
    $$str= $bytes;
}

sub tpl2xml {
    my $src = shift;
    my (@out,@args);

    # build list of references to hold output of unpacking
    my ($fmt,@fxlens) = peek_fmt($src);
    for(my ($i,$j,$k)=(0,0,0);$i<length($fmt);$i++) {
        push @args, [] if substr($fmt,$i,2) =~ /^[iucfIU]\#$/; # octothorpic
        push @args, \$out[$j++] if substr($fmt,$i,2) =~ /^[iuBscfIU][^\#]*$/;
        push @args, $fxlens[$k++] if substr($fmt,$i,1) eq "#";
    }

    my $tpl = Tpl->tpl_map($fmt,@args);
    $tpl->tpl_load($src);
    $tpl->tpl_unpack(0);

    # construct xml preamble
    my $pre = qq{<?xml version="1.0" encoding="utf-8" ?>
      <!DOCTYPE tplxml [ 
      <!ELEMENT tplxml (A|i|u|I|U|B|s|c|f|fx)*>
      <!ATTLIST tplxml
         format CDATA #REQUIRED
         fxlens CDATA #REQUIRED
      >
      <!ELEMENT i (#PCDATA)>
      <!ELEMENT u (#PCDATA)>
      <!ELEMENT I (#PCDATA)>
      <!ELEMENT U (#PCDATA)>
      <!ELEMENT B (#PCDATA)>
      <!ELEMENT s (#PCDATA)>
      <!ELEMENT c (#PCDATA)>
      <!ELEMENT f (#PCDATA)>
      <!ELEMENT A (el)*>
      <!ELEMENT el (A|i|u|I|U|B|s|c|f|fx)+>
      <!ELEMENT fx (i|u|I|U|c|f)*>
      ]>\n};
    print $pre;
    my $fxattr = join ",", @fxlens;
    print qq{<tplxml format="$fmt" fxlens="$fxattr">\n};
    tpl2xml_node($tpl,"A0",1);
    print qq{</tplxml>\n};
}

sub tpl2xml_node {
    my $tpl = shift;
    my $node = shift;
    my $indent = shift;
    my $i = "  " x $indent;
    for my $c (@{ $tpl->{$node} }) {
        if (ref($c)) { 
            my ($type,$addr,$fxlen) = @$c;
            quote_chars $addr if $type eq 's';
            hex_chars $addr if $type eq 'B';
            if (not defined $fxlen) {
                print qq{$i<$type>$$addr</$type>\n}; # singleton
            } else {
                # all elements of octothorpic fixed-len array
                print qq{$i<fx>\n};
                print qq{$i  <$type>$addr->[$_]</$type>\n} for (0..$fxlen-1); 
                print qq{$i</fx>\n};
            }
        } else { 
            # A node
            print qq{$i<A>\n};
            my $idx = $1 if $c =~ /^A(\d+)$/;
            while($tpl->tpl_unpack($idx) > 0) {
                print qq{$i<el>\n};
                tpl2xml_node($tpl,$c,$indent+1);  
                print qq{$i</el>\n};
            }
            print qq{$i</A>\n};
        }
    }
}

sub xml2tpl {
    my $src = shift;
    my $p = new XML::Parser( Style => 'Tree' );
    my $tree = $p->parse($$src); 
    die "not a tpl xml document" unless $tree->[0] eq 'tplxml';
    die "no format attribute" unless defined $tree->[1][0]->{format};
    my $fmt = $tree->[1][0]->{format};
    die "no fxlens attribute" unless defined $tree->[1][0]->{fxlens};
    my @fxlens = split /,/, $tree->[1][0]->{fxlens};

    # build list of references to variables for use in packing
    my (@args,@out);
    for(my ($i,$j,$k)=(0,0,0);$i<length($fmt);$i++) {
        push @args, [] if substr($fmt,$i,2) =~ /^[iucfIU]\#$/; # octothorpic
        push @args, \$out[$j++] if substr($fmt,$i,2) =~ /^[iuBscfIU][^\#]*$/;
        push @args, $fxlens[$k++] if substr($fmt,$i,1) eq "#";
    }

    my $tpl = Tpl->tpl_map($fmt,@args);
    xml2tpl_dfs($tpl,$tree->[1]);
    $tpl->tpl_pack(0);
    print $tpl->tpl_dump;
}

sub xml2tpl_dfs {
    my $tpl = shift;
    my $xml = shift;

    my @next = @$xml;  # ($attr,@tagvals) = $$xml;
    shift @next;  # discard <tplxml> attributes
    my @tpltoks = @{ $tpl->{"A0"} }; #expected tokens when parsing
    
    TAG: while (@next) {
        my $xmltag = shift @next;
        my $xmlval = shift @next;

        # skip whitespace/newlines embedded between tags
        next TAG if ($xmltag eq "0" and $xmlval =~ /^\s+$/); 

        # pack if necessary. consume tokens by look-ahead until non-pack token.
        while (@tpltoks > 0 and $tpltoks[0] =~ /^P(\d+)$/) {
            shift @tpltoks;
            $tpl->tpl_pack($1);
        }

        # If tpl format specifies a non-array type should appear at this point 
        # in the XML tree, then validate the type matches the format and assign 
        # the value from the XML to the variable from which it'll be packed
        my $tpltoken = shift @tpltoks;
        my $octothorpic=0;
        if (ref $tpltoken) {
            my ($tpltype,$tpladdr,$fxlen) = @$tpltoken;

            #  This block is how we handle octothorpic (fixed length) arrays.
            #  If $fxlen is defined then an octothorpic <fx> node is expected.
            #  After finding the <fx> node we put its subnodes (the array elements)
            #  onto the @next array for immediate parsing and we use $fxlen:$remaining
            #  as a signet version of the $fxlen to induce the element-processing loop.
            if (defined $fxlen) {
                if ($fxlen =~ /^(\d+):(\d+)$/) { # $1==orig $fxlen, $2==remain $fxlen
                    $octothorpic=1;
                    unshift @tpltoks, [$tpltype, $tpladdr, $1.":".($2-1)] if $2 > 1;
                } else { # octothorpic array expected; look for <fx> parent node
                    die "expected '<fx>' but got '<$xmltag>'" unless $xmltag eq 'fx';
                    @{ $tpladdr } = (); # Empty accumulator array for octothorpic values
                    unshift @tpltoks, [$tpltype, $tpladdr, "$fxlen:$fxlen"]; # x:x signet
                    shift @$xmlval; # discard 'A' attributes
                    unshift @next, @$xmlval;  #parse xml subtree now (dfs)
                    next TAG; # proceed to children of <fx> node
                }
            }

            if ($tpltype ne $xmltag) {
                die "mismatch: xml has '$xmltag' where format specifies '$tpltype'";
            } 
            # expect @$xmlval to be ({},0,'value') i.e. a single, terminal text node
            if (@$xmlval > 3 || $xmlval->[1] ne '0') {
                die "error: xml tag '$xmltag' cannot enclose sub-tags";
            }
            if ($octothorpic) {
                push @{ $tpladdr }, $xmlval->[2]; 
            } else {
                $$tpladdr = $xmlval->[2]; 
            }
            unquote_chars $tpladdr if $tpltype eq 's';
            unhex_chars $tpladdr if $tpltype eq 'B';
        } elsif ($tpltoken =~ /^A(\d+)$/) {
            # tpl format specifies an array should appear at this point in the XML
            if ($xmltag ne 'A') {
                die "mismatch: xml has '$xmltag' where format specifies 'A'";
            }
            shift @$xmlval; # discard 'A' attributes

            # form token that means "replace me with tokens from A(n), x times"
            # (where x is the number of elements contained by this array).
            my $array_count=0;
            for(my $i=0; $i < @$xmlval; $i+=2) {
                $array_count++ if $xmlval->[$i] eq 'el';
            }

            unshift @tpltoks, "N$1:$array_count" if $array_count > 0; 
            unshift @next, @$xmlval;  #parse xml subtree now (dfs)
        } elsif ($tpltoken =~ /^N(\d+):(\d+)$/) {
            if ($xmltag ne "el") {
                die "mismatch: xml has '$xmltag' where array 'el' is expected";
            }
            # prepend A$1's tokens (and decremented N:count) to expected tokens 
            my ($n,$elsleft) = ($1, ($2 - 1));
            unshift @tpltoks, "N$n:$elsleft" if $elsleft > 0;  
            unshift @tpltoks, "P$n";  # "pack me now" token
            unshift @tpltoks, @{ $tpl->{"A$1"} };
            
            shift @$xmlval; # discard 'el' attributes
            unshift @next, @$xmlval;  # proceed to parse el subtree (dfs)
        } else {
            die "internal error, unexpected token $tpltoken";
        }
    }

    # pack if necessary. consume tokens by look-ahead until non-pack token.
    while (@tpltoks > 0 and $tpltoks[0] =~ /^P(\d+)$/) {
        shift @tpltoks;
        $tpl->tpl_pack($1);
    }

    if (@tpltoks > 0) {
        die "error: end of xml document reached but format requires more data";
    }
}

sub peek_fmt {
    my $buf = shift;
    die "invalid tpl file" unless ($$buf =~ /^tpl/);
    my $flags = CORE::unpack("C", substr($$buf,3,1));
    my $UF = ($flags & 1) ? "N" : "V";  # big or little endian fxlens
    my $fmt = (CORE::unpack("Z*", substr($$buf,8)));
    my $num_octothorpes = scalar (my @o = ($fmt =~ /#/g));
    my @fxlens;
    my $fx = 8 + length($fmt) + 1;
    for(my $i=0; $i < $num_octothorpes; $i++) {
        my $fxlen_bytes = substr($$buf,$fx,4);
        my $fxlen = unpack($UF, $fxlen_bytes);
        push @fxlens, $fxlen;
        $fx += 4;
    }
    return ($fmt,@fxlens);
}

##########################################################################
# Slurp input file, auto-detect if conversion is to tpl or XML, and run.
##########################################################################

undef $/; 
my $src = <>;
our $to = (substr($src,0,3) eq "tpl") ? "xml" : "tpl";
xml2tpl(\$src) if $to eq "tpl";
tpl2xml(\$src) if $to eq "xml";