# -*- mode: Perl -*-
# /=====================================================================\ #
# |  listings                                                           | #
# | Implementation for LaTeXML                                          | #
# |=====================================================================| #
# | Part of LaTeXML:                                                    | #
# |  Public domain software, produced as part of work done by the       | #
# |  United States Government & not subject to copyright in the US.     | #
# |---------------------------------------------------------------------| #
# | Bruce Miller <bruce.miller@nist.gov>                        #_#     | #
# | http://dlmf.nist.gov/LaTeXML/                              (o o)    | #
# \=========================================================ooo==U==ooo=/ #
package LaTeXML::Package::Pool;
use strict;
use warnings;
use LaTeXML::Package;
use MIME::Base64;
use Encode qw(is_utf8 encode);

#======================================================================
# To the extent we succeed in doing all the pretty-printing...
# It rather seems that preserving a raw, unformatted, copy of the code
# would be a Useful thing, and in keeping with XML.
# Wouldn't you want to see the pretty print, but cut&paste the plain code?
# This may eventually need some schema support...

# NOTE: The MoreSelectCharTable parameter for languages styles is
# NOT yet implemented; this has an impact on
#  HTML, XML comments & CDATA sections not recognized.
#  Perl, Visual Basic, Java, tcl; something (comments?) is not recognized, but what?

RequireResource('ltx-listings.css');
RequirePackage('textcomp');
#======================================================================
# Top-level listings commands
#======================================================================

# Set various Listings keys
DefPrimitive('\lstset RequiredKeyVals:LST', sub { lstActivate($_[1]); return; });
AssignValue(LISTINGS_PREAMBLE        => []);
AssignValue(LISTINGS_PREAMBLE_BEFORE => []);
AssignValue(LISTINGS_POSTAMBLE       => []);
our $EMPTY_CATTABLE = LaTeXML::Core::State->new(catcodes => 'none');

DefMacro('\lstinline OptionalKeyVals:LST', sub {
    my ($gullet, $keyvals) = @_;
    $STATE->getStomach->bgroup;    # To localize activation
    lstActivate($keyvals);         # But do BEFORE reading arg, since some options screw things up.
                                   # get opening delim from gullet, before cattable swap,
                                   # in case token was already peeked for the optional args!
    my $init = $gullet->readToken;
    my $body = listingsReadRawString($gullet, (Equals($init, T_BEGIN) ? T_END : $init));
    return (
      @{ LookupValue('LISTINGS_PREAMBLE_BEFORE') },
      lstProcessInline($body),
      @{ LookupValue('LISTINGS_POSTAMBLE') },
      T_END); });                  # to balance ->bgroup

# But it can also be used as an environment!
DefMacroI(T_CS('\begin{lstinline}'), 'OptionalKeyVals:LST', sub {
    my ($gullet, $keyvals) = @_;
    $STATE->getStomach->bgroup;
    AssignValue(current_environment => 'lstlisting');
    DefMacroI('@currenvir', undef, 'lstlisting');
    my $text = listingsReadRawLines($gullet, 'lstinline');
    lstActivate($keyvals);
    return (
      @{ LookupValue('LISTINGS_PREAMBLE_BEFORE') },
      lstProcessInline($text),
      @{ LookupValue('LISTINGS_POSTAMBLE') },
      T_END); });    # to balance ->bgroup

sub lstProcessInline {
  my ($text) = @_;
  return Invocation(T_CS('\@listings@inline'),
    lstProcess('inline', $text)); }

DefConstructor('\@listings@inline {}',
  "<ltx:text class='ltx_lstlisting' _noautoclose='1'>#1</ltx:text>",
  reversion => '\lstinline{#1}');

# Not a regular environment, since we're going to read the body verbatim!
DefMacroI(T_CS('\begin{lstlisting}'), 'OptionalKeyVals:LST', sub {
    my ($gullet, $keyvals) = @_;
    $STATE->getStomach->bgroup;
    AssignValue(current_environment => 'lstlisting');
    DefMacroI('@currenvir', undef, 'lstlisting');
    my $text = listingsReadRawLines($gullet, 'lstlisting');
    lstActivate($keyvals);
    return lstProcessDisplay(lstGetTokens('name'), $text); });

DefMacro('\lstinputlisting OptionalKeyVals:LST Semiverbatim', sub {
    my ($gullet, $keyvals, $file) = @_;
    my $text = listingsReadRawFile($gullet, $file);
    $STATE->getStomach->bgroup;
    lstActivate($keyvals);
    AssignValue('LST@toctitle', $file);    # so it shows up in list of..
    return lstProcessDisplay($file, $text); });

NewCounter('lstlisting', 'document', idprefix => 'LST');
DefMacro('\ext@lstlisting', 'lol');
AssignValue(LISTINGS_DATA_COUNTER => 0);

# Defining new listing environments
DefPrimitive('\lstnewenvironment {}[Number][]{}{}', sub {
    my ($stomach, $name, $n, $opt, $start, $end) = @_;
    $name = ToString($name);
    DefMacroI(T_CS("\\begin{$name}"), LaTeXML::Package::convertLaTeXArgs($n, $opt), sub {
        my ($gullet, @args) = @_;
        $STATE->getStomach->bgroup;
        PushValue(LISTINGS_POSTAMBLE => $end->substituteParameters(@args));
        # This will typically have \lstset, equivalent of lstActivate
        Digest($start->substituteParameters(@args));
        my $text = listingsReadRawLines($gullet, $name);
        return lstProcessDisplay(lstGetTokens('name'), $text); });
});

# Return 2 lists: body, trailer
sub lstProcessBlock {
  my ($name, $text) = @_;
  # store the data to be placed in the listing
  my $c = LookupValue('LISTINGS_DATA_COUNTER') + 1;
  AssignValue('LISTINGS_DATA_COUNTER' => $c,    'global');
  AssignValue('LISTINGS_DATA_' . $c   => $text, 'global');
  return (
    [@{ LookupValue('LISTINGS_PREAMBLE_BEFORE') },
      Invocation(T_CS('\@@listings@block'), $c, lstProcess('block', $text))],
    [@{ LookupValue('LISTINGS_POSTAMBLE') },
      T_END]); }

sub lstProcessDisplay {
  my ($name, $text) = @_;
  # Hmm.. should locally define \lstname to be either name or the file...
  my ($body, $trailer) = lstProcessBlock($name, $text);
  my @body = @$body;
  # Figure out whether the display is numbered, or has a caption or titles.
  my @caption = ();
  my ($numbered, $labelled, $caption, $x);
  if (($x = lstGetTokens('caption')) && scalar($x->unlist)) {
    my @t  = $x->unlist;
    my @tc = ();
    if (Equals($t[0], T_OTHER('['))) {
      while (!Equals($t[0], T_OTHER(']'))) { push(@tc, shift(@t)); } }
    $numbered = 1;
    $caption = Invocation(T_CS('\lstlisting@makecaption'), (@tc ? Tokens(@tc) : undef), Tokens(@t)); }
  elsif (($x = lstGetTokens('title')) && scalar($x->unlist)) {
    $caption = Invocation(T_CS('\lstlisting@maketitle'), $x); }
  elsif (($x = lstGetTokens('toctitle')) && scalar($x->unlist)) {
    $caption = Invocation(T_CS('\lstlisting@maketoctitle'), $x); }
  if (($x = lstGetTokens('label')) && scalar($x->unlist)) {
    $labelled = 1;
    unshift(@body, Invocation(T_CS('\label'), $x)); }
  if ($caption) {
    if (lstGetLiteral('captionpos') eq 't') {
      unshift(@body, $caption); }
    else {
      push(@body, $caption); } }

  push(@body, @$trailer);
  # We go a bit (a bit too far?) to try to treat this as a separate Para level object
  # (if with captions or titled),
  # or as an in-block item (within a logical paragraph)

  return (
    ($numbered || $caption ? (T_CS('\par')) : ()),
    T_BEGIN,
    ($name ? (T_CS('\def'), T_CS('\lstname'), T_BEGIN, $name->unlist, T_END) : ()),
    ($numbered
      ? Invocation(T_CS('\@listings'), Tokens(@body))
      : ($caption
        ? Invocation(T_CS('\@@listings'), Tokens(@body))
        : @body)),
    T_END); }

DefMacro('\lstlisting@makecaption[]{}',
  '\def\@captype{lstlisting}'
    . '\@@add@caption@counters'
    . '\@@toccaption{\lx@format@toctitle@@{lstlisting}{\ifx.#1.#2\else#1\fi}}'
    . '\@@caption{\lx@format@title@@{lstlisting}{#2}}');
DefMacroI('\fnum@lstlisting', undef, '\lstlistingname\nobreakspace\thelstlisting');
DefMacro('\format@title@lstlisting{}', '\lx@tag[][: ]{\fnum@lstlisting}#1');

DefMacro('\lstlisting@maketitle{}',
  '\@@toccaption{#1}'
    . '\@@caption{#1}');

DefMacro('\lstlisting@maketoctitle{}',
  '\@@toccaption{#1}');

# Numbered form,  with caption
# \@listings <classes> <formatted>
DefConstructor('\@listings {}',
  "<ltx:float inlist='lol' xml:id='#id' class='ltx_lstlisting'>"
    . "#tags"
    . "#1"
    . "</ltx:float>",
  beforeDigest => sub { DefMacroI('\@captype', undef, 'lstlisting'); },
  afterDigest => sub {
    my ($stomach, $whatsit) = @_;
    RescueCaptionCounters('lstlisting', $whatsit); });

# Unnumbered form, but with caption
# \@listings <classes> <formatted>
DefConstructor('\@@listings {}',
  "<ltx:float xml:id='#id' class='ltx_lstlisting'>"
    . "#tags"
    . "#1"
    . "</ltx:float>",
  properties => sub { RefStepID('lstlisting'); },
  beforeDigest => sub { DefMacroI('\@captype', undef, 'lstlisting'); },
  afterDigest => sub {
    my ($stomach, $whatsit) = @_;
    RescueCaptionCounters('lstlisting', $whatsit); });

DefConstructor('\@@listings@block {} {}',
"<ltx:listing class='ltx_lstlisting' data='#data' datamimetype='#datamimetype' dataencoding='#dataencoding'>#2</ltx:listing>",
  afterDigest => sub {
    my ($stomach, $whatsit) = @_;
    # Could have some options about encoding?
    my $data_key      = 'LISTINGS_DATA_' . ToString($whatsit->getArg(1));
    my $listings_data = LookupValue($data_key);
    if (is_utf8($listings_data)) {
      $listings_data = encode('UTF-8', $listings_data);
    }
    my $data = encode_base64($listings_data, '');    # NO linebreaking!
    $whatsit->setProperties(data => $data, datamimetype => 'text/plain', dataencoding => 'base64'); });

#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# Low Level String stuff
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# Read raw string until closing $until.
# Note that this does NOT balance groups, even if $until is '}'!
sub listingsReadRawString {
  my ($gullet, $until) = @_;
  # NOTE that this normally does NOT balance {, but DOES within mathescape'd $
  # Moreover, neither escapechar nor escapeinside have this effect!
  # I'd swear this is a bug that became a feature.
  my $SAVESTATE  = $STATE;
  my $mathescape = lstGetBoolean('mathescape');
  local $STATE = $EMPTY_CATTABLE;
  my $inmath = 0;
  my @tokens = ();
  while (defined(my $token = $gullet->readToken())) {
    last if $until and $token->getString eq $until->getString;
    if ($mathescape && ($token->getString eq '$')) {
      if ($inmath) { $inmath = 0; $STATE = $EMPTY_CATTABLE; }
      else { $inmath = 1; $STATE = $SAVESTATE; } }
    if ($inmath && $token->equals(T_BEGIN)) {
      push(@tokens, T_BEGIN, $gullet->readBalanced->unlist, T_END); }
    else {
      push(@tokens, $token); } }
  while (@tokens && $tokens[-1]->getCatcode == CC_SPACE) {    # Remove trailing space
    pop(@tokens); }
  return ToString(Tokens(@tokens)); }

# Read raw strings for environment, until matching \end{$environment}
sub listingsReadRawLines {
  my ($gullet, $environment) = @_;
  my @lines = ();
  my ($line);
  $gullet->readRawLine;                                       # Ignore 1st line (following \begin{...}
  while (defined($line = $gullet->readRawLine)) {
    if ($line =~ /^\s*\\end\{\Q$environment\E\}(.*?)$/) {
      $gullet->unread(Tokenize($1), T_CR);                    # put BACK what follows the \end{whatever}
      last; }
    push(@lines, $line); }
  return join("\n", @lines); }

sub listingsReadRawFile {
  my ($gullet, $file) = @_;
  my $filename = ToString(Expand($file));
  my $path     = FindFile($filename);
  my $text;
  my $LST_FH;
  if ($path && open($LST_FH, '<', $path)) {
    { local $/ = undef;
      $text = <$LST_FH>;
      close($LST_FH); } }
  else {
    Error('I/O', $filename, $gullet, "Can't read listings file '$filename'", $!); }
  return $text; }

#======================================================================
our $lst_charmapping = { '#' => T_CS('\#'), '$' => T_CS('\textdollar'), '&' => T_CS('\&'),
  "'" => T_CS('\textquoteright'),
  '*' => T_CS('\textasteriskcentered'),
  # ??  '-'=>$-$ ??
  '<' => T_CS('\textless'),        '>' => T_CS('\textgreater'), '\\' => T_CS('\textbackslash'),
  '^' => T_CS('\textasciicircum'), '_' => T_CS('\textunderscore'),
  '`' => T_CS('\textquoteleft'),
  '{' => T_CS('\textbraceleft'),   '}' => T_CS('\textbraceright'), '%' => T_CS('\%'),
  '|' => T_CS('\textbar'),
  '~' => T_CS('\textasciitilde'),
};

# Note that listings.sty uses a couple of textcomp characters
# (\textquotesingle, \textasciigrave)
# when the upquote option is used.
our $lst_charmapping_upquote = {
  %$lst_charmapping,
  "'" => T_CS('\textquotesingle'),
  '`' => T_CS('\textasciigrave') };

sub lstRescan {
  my ($tokens) = @_;
  my $mapping = (lstGetBoolean('upquote') ? $lst_charmapping_upquote : $lst_charmapping);
  return (defined $tokens
    ? Tokens(map { ($_->getCatcode == CC_OTHER ? $$mapping{ $_->getString } || $_ : $_) } $tokens->unlist)
    : undef); }

#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# Managing the sets of keyvals that compose a Listings Style or Language.
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# Assign (locally) all values or effects from a Listings keyvals
# Note that we operate on the Pairs form of keyvals to preserve order, repetition
#
# LST_CHARACTERS hash (letter|digit|other) => hash : charre=>1
# LST_CLASSES hash classname => hash : begin, end => Tokens
#    and some extra: index=>indexclassname, escape=>0|1, eval=>0|1, ...
# LST_WORDS hash word => hash : class=>classname, index=>indexclassname
# LST_DELIMTERS hash open => hash: regexp=>re, close => re, classname?

foreach my $table (qw(LST_CHARACTERS LST_CLASSES LST_WORDS LST_DELIMITERS)) {
  AssignValue($table => {}); }

sub lstActivate {
  my ($kv) = @_;
  if ($kv) {
    # We will construct distillations of the various keyword, delimiter, etc data
    # These tables will sit in the current binding, but we need to copy the data from previous bindings
    # to get the effect of grouping
    # Each table is a hash of hashes.
    foreach my $table (qw(LST_CHARACTERS LST_CLASSES LST_WORDS LST_DELIMITERS)) {
      my %data = ();
      if (my $prev = LookupValue($table)) {
        map { $data{$_} = { %{ $$prev{$_} } } } keys %$prev; }
      AssignValue($table => {%data}); }
    # Now start scanning the keywords, in order, and activate their effects.
    my @pairs = $kv->getPairs();
    while (@pairs) {
      my ($key, $val) = (shift(@pairs), shift(@pairs));
      $val = lstUnGroup($val);
      my $cs = T_CS('\lst@@' . $key);
      if (IsDefined($cs)) {
        $val = LookupValue('KEYVAL@LST@' . $key . '@default') unless $val;
        # Done for effect.
        Digest(Tokens($cs, ($val ? $val->unlist : Tokens()), T_CS('\end'))); }
      AssignValue('LST@' . $key => $val); } }
  return; }

#----------------------------------------------------------------------
# Various helpers for dealing with the arguments to options.
# Strip outer {} if there's only a single group
# [The need for this may be a sign of not-quite-correct keyval handling?]
sub lstUnGroup {
  my ($tokens) = @_;
  if ($tokens && ref $tokens) {
    my @t = $tokens->unlist;
    if (Equals($t[0], T_BEGIN) && Equals($t[-1], T_END) && (count_groups(@t) == 1)) {
      $tokens = Tokens(@t[1 .. $#t - 1]); } }
  return $tokens; }

sub count_groups {
  my (@tokens) = @_;
  my $groups   = 0;
  my $level    = 0;
  foreach my $t (@tokens) {
    if (Equals($t, T_END)) {
      $level--; }
    elsif (Equals($t, T_BEGIN)) {
      $groups++ if $level == 0;
      $level++; } }
  return $groups; }

sub lstSplit {
  my ($stuff) = @_;
  my $string = ToString(lstUnGroup($stuff));
  $string =~ s/%.*?\n\s*//sg;
  $string =~ s/\s+//sg;
  return split(/,/, $string); }

# Strip of TeX's quoting.
sub lstDeslash {
  my ($string) = @_;
  if ($string) {
    $string = ToString($string);
    $string =~ s/^\\(.)/$1/g;    # Strip off TeX's "quoting"
    return $string; } }

# Convert a string of TeX chars to a regexp to match it.
sub lstRegexp {
  my ($chars) = @_;
  if (my $string = lstDeslash($chars)) {
    $string =~ s/([\!\@\#\$\%\^\&\*\(\)\_\-\+\{\}\[\]\\\<\>\?\/\|])/\\$1/g;    # Put back for Perl.
    return $string; } }

#----------------------------------------------------------------------
# A rather bizarro set of keyword value parsing bits.
# Perhaps should be handled by the keyval types themselves?
sub lstGetLiteral {
  my ($value) = @_;
  my $v = ToString(LookupValue('LST@' . $value));
  if ($v =~ /^\{(.*?)\}$/) {
    $v = $1; }
  return $v; }

sub lstGetBoolean {
  my ($value) = @_;
  return lstGetLiteral($value) eq 'true'; }

sub lstGetNumber {
  my ($value) = @_;
  my $n = LookupValue('LST@' . $value);
  return ($n ? $n->valueOf : 0); }

sub lstGetTokens {
  my ($value) = @_;
  if (my $v = LookupValue('LST@' . $value)) {
    return lstUnGroup($v); }
  else {
    return Tokens(); } }

#======================================================================
# Support for managing classes, delimiters and such.

sub lstClassName {
  my ($class, $n) = @_;
  $n = 1 unless $n;
  $n = $n->valueOf if ref $n;
  $n += lstGetNumber('classoffset');
  return $class . ($n <= 1 ? '' : $n); }

# Define properties of a Class (comments, strings, etc)
sub lstSetClassStyle {
  my ($class, $style, %props) = @_;
  my $classes = LookupValue('LST_CLASSES');
  if ($style) {
    my $stylestring = ToString($style);
    $stylestring =~ s/^\s+//s; $stylestring =~ s/\s+$//s;
    if ($stylestring =~ s/style(\d*)$/s$1/) {    # If names a style, convert it into the class name
      delete $$classes{$class}{begin};           # remove explicit styling
      $props{class} = $stylestring; }            # add indirect to class.
    else {
      delete $$classes{$class}{class};
      $props{begin} = $style; } }                # Otherwise, presumably TeX
  if ($class =~ /^(\w+?)s?$/) {
    $props{cssclass} = $1; }
  map { $$classes{$class}{$_} = $props{$_} } keys %props;
  return; }

# Specify a set of words belonging to a class
sub lstSetClassWords {
  my ($class, $words, $prefix) = @_;
  # First delete existing words
  my $wordslist = LookupValue('LST_WORDS');
  foreach my $word (keys %$wordslist) {
    delete $$wordslist{$word}{class} if ($$wordslist{$word}{class} || '') eq $class; }
  lstAddClassWords($class, $words, $prefix);
  return; }

sub lstAddClassWords {
  my ($class, $words, $prefix) = @_;
  my $wordslist = LookupValue('LST_WORDS');
  foreach my $word (lstSplit($words)) {
    $word = $prefix . $word if $prefix;
    $$wordslist{$word}{class} = $class unless $$wordslist{$word}{class}; }
  return; }

sub lstDeleteClassWords {
  my ($class, $words, $prefix) = @_;
  my $wordslist = LookupValue('LST_WORDS');
  foreach my $word (lstSplit($words)) {
    $word = $prefix . $word if $prefix;
    delete $$wordslist{$word}{class} if $$wordslist{$word}{class} eq $class; }
  return; }

# This probably needs a different way of decoding $type.
#   General: b,d,l,s,n  (+ i)
#   String: b,d,m,bd  (backslash, doubled, matlab-like(?) or backslash or doubled)
# Need to pull out the $delims decoding, to allow deleting delimiters.
# Recognized keys:
#   recursive   : allows keywords, comments & strings inside
#   cummulative : the effects are cummulative (?)
#   nested      : allows comments to be nested
sub lstAddDelimiter {
  my ($kind, $type, $style, $delims, %keys) = @_;
  $type = ToString($type);
  my $delimlist = LookupValue('LST_DELIMITERS');
  my $invisible = ($type =~ /^(?:bd|b|d|l|s|n)i$/) || ($type =~ /^i(?:bd|b|d|l|s|n)$/);
  $type =~ s/i// if $invisible;
  my $quoted;
  my ($open, $close, $openre, $closere, $cssclass);
  if ($type eq 'b') {    # Balanced; same delim open & close; but not when slashed
    $open = $close = $delims;
    $openre  = lstRegexp($open);
    $closere = "(?<!\\\\)$openre";
    $quoted  = "\\\\\\$openre"; }
  elsif ($type eq 'd') {    # Doubled: same delim open & close; but not when doubled.
    $open = $close = $delims;
    $openre  = lstRegexp($open);
    $closere = "(?<!$openre)$openre(?!$openre)";
    $quoted  = $openre . $openre; }
  elsif ($type eq 'bd') {    # Doubled: same delim open & close; not when doubled OR slashed
    $open = $close = $delims;
    $openre  = lstRegexp($open);
    $closere = "(?<!\\\\|$openre)$openre(?!$openre)";
    $quoted  = "\\\\\\$openre|$openre$openre"; }
  elsif ($type eq 'l') {     # Line: close is till end of line
    $open    = $delims;
    $openre  = lstRegexp($open);
    $close   = undef;
    $closere = "(?=\n)"; }
  elsif ($type eq 's') {     # String: different open & close
    ($open, $close) = lst_splitDelimiters($delims);
    $openre  = lstRegexp($open);
    $closere = lstRegexp($close); }
  elsif ($type eq 'n') {     # like String, but allows nesting!!!
    ($open, $close) = lst_splitDelimiters($delims);
    $openre       = lstRegexp($open);
    $closere      = lstRegexp($close);
    $keys{nested} = 1; }
  # Special case? Maybe we have to deal with lstmisc.sty and understand "aspects"???
  elsif ($type eq 'directive') {
    $kind    = $type . 's';
    $open    = $delims;
    $openre  = lstRegexp($open);
    $closere = "(?=\\W)"; }        # ? word boundary but appearing at beginning of string!
  else {                           # ??? What should be default? (same as 'directive'???)
    $open   = $delims;
    $openre = lstRegexp($open); }
  if (my $openstring = lstDeslash($open)) {
    # The styling can be a class name, or markup
    my $class       = $kind;
    my $stylestring = ToString($style);
    my $styleTeX;
    if ($stylestring =~ s/style(\d*)$/s$1/) {    # Names the style associated with a class.
      $class = $stylestring; }
    else {                                       # Otherwise, assume it is markup.
      $styleTeX = $style; }
    my $oldclass = $class;
    $class = $class . ToString($open) . ToString($close); # Create an artificial class for this delimiter.
    my $openTeX = ($styleTeX
      ? ($invisible ? $styleTeX : Tokens($styleTeX->unlist, $open))
      : ($invisible ? () : $open));
    my $closeTeX = ($invisible ? () : $close);
    lstSetClassStyle($class, undef, begin => $openTeX, end => $closeTeX,
      class => $oldclass, cssclass => $cssclass);
    # NOT DONE:
    #   invisibility of the whole delimited expression
    #   nestability.
    $$delimlist{$openstring} = { open => $openre, close => $closere, class => $class,
      quoted => $quoted, %keys }; }
  return; }

# Helper for lstAddDelimiter:
# Here's the goofy thing: there may or may be {} in delimiters;
# And, when there's 2 delimiters, it could even be is: open}{close
# we'll hope there're no extra braces!
# If type eq 'n', comments are allowed to nest!!!
sub lst_splitDelimiters {
  my ($delims) = @_;
  my @t = grep { !Equals($_, T_BEGIN) } $delims->unlist;    # Remove any T_BEGIN
  my @t1 = ();
  if (scalar(@t) == 2) {
    @t1 = ($t[0]); @t = ($t[1]); }
  else {
    while (@t && !Equals($t[0], T_END)) { push(@t1, shift(@t)); }
    @t = grep { !Equals($_, T_END) } @t; }                  # Remove any remaining T_END
  return (Tokens(@t1), Tokens(@t)); }                       # return open & close

# Set character classes
sub lstSetCharacterClass {
  my ($class, $chars) = @_;
  my $charslist = LookupValue('LST_CHARACTERS');
  foreach my $char ($chars->unlist) {
    $char = lstRegexp($char);
    delete $$charslist{letter}{$char};
    delete $$charslist{digit}{$char};
    delete $$charslist{other}{$char};
    $$charslist{$class}{$char} = 1; }
  return; }

#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# The various parameters
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

#======================================================================
# 4.3 Space and placement
#======================================================================
# Ignorable
DefKeyVal('LST', 'float',          '');            # [*] t,b,p,h  [or defaults?]
DefKeyVal('LST', 'floatplacement', '');            # t,b,p
DefKeyVal('LST', 'aboveskip',      'Dimension');
DefKeyVal('LST', 'belowskip',      'Dimension');
DefKeyVal('LST', 'lineskip',       'Dimension');
DefKeyVal('LST', 'boxpos',         '');            # b,c,t

#======================================================================
# 4.4 Printed range
#======================================================================
# Seemingly handled....
DefKeyVal('LST', 'print', '', 'true');
DefKeyVal('LST', 'firstline',  'Number');
DefKeyVal('LST', 'lastline',   'Number');
DefKeyVal('LST', 'showlines',  '', 'true');
DefKeyVal('LST', 'emptylines', '');           # NOTE: NOT YET HANDLED.
DefKeyVal('LST', 'gobble',     'Number');

#======================================================================
# 4.5 Language and styles
#======================================================================
# Define a Style being a shorthand for a set of Listings keyvals
# \lstdefinestyle{stylename}{keys}
DefPrimitive('\lstdefinestyle{} RequiredKeyVals:LST', sub {
    my ($stomach, $style, $keyvals) = @_;
    $style = uc(ToString(lstUnGroup($style)));
    $style =~ s/\s+//g;
    AssignValue('LST@STYLE@' . $style => $keyvals); });

DefKeyVal('LST', 'style', '');
DefMacro('\lst@@style Until:\end', sub {
    my ($gullet, $style) = @_;
    if ($style = uc(ToString(lstUnGroup($style)))) {
      $style =~ s/\s+//g;
      if (my $values = LookupValue('LST@STYLE@' . $style)) {
        lstActivate($values); }
      else {
        Warn('unexpected', $style, $gullet, "No listings style '$style' found"); } }
    return; });

sub lstActivateLanguage {
  my ($language, $dialect) = @_;
  $language = uc(ToString($language)); $language =~ s/\s+//g;
  my ($values, $name);
  if ($language) {
    while (1) {
      # Construct the language$dialect that we're trying to find.
      my $d = ($dialect && $dialect->unlist ? $dialect : LookupValue('LSTDD@' . $language));
      $name = 'LST@LANGUAGE@' . $language;
      if ($d && $d->unlist) {
        $d = uc(ToString($d)); $d =~ s/\s+//g;
        $name .= '$' . $d; }
      # language definition is loaded!
      last if $values = LookupValue($name);
      # try to load next configuration file; fail if no more configs.
      if (my $file = ShiftValue('LST_LANGUAGE_FILES')) {
        InputDefinitions($file, noerror => 1); }    # Note: OK if some definition files missing!
      else { last; } }
    # Found a definition, activate it.
    if ($values) {
      lstActivate($values); }
    # Else failed to find one.
    else {
      Warn('unexpected', $name, $STATE->getStomach, "No listings language '$language' found"); } }
  return; }

DefKeyVal('LST', 'language', '');
DefMacro('\lst@@language [] Until:\end', sub {
    lstActivateLanguage($_[2], $_[1]);
    PushValue(LISTINGS_PREAMBLE => T_CS('\lst@@@set@language'));
    return; });
DefConstructor('\lst@@@set@language', sub {
    my ($document, %props) = @_;
    if (my $lang = $props{language}) {
      $lang = "$2_$1" if $lang =~ /^\[([^\]]*)\](.*)$/;
      $_[0]->addClass($_[0]->getElement, 'ltx_lst_language_' . $lang); } },
  properties => { language => sub { lstGetLiteral('language'); } });

DefKeyVal('LST', 'alsolanguage', '');
DefMacro('\lst@@alsolanguage [] Until:\end', sub {
    lstActivateLanguage($_[2], $_[1]); return; });

DefKeyVal('LST', 'defaultdialect', '');
DefMacro('\lst@@defaultdialect[] Until:\end', sub {
    my ($gullet, $dialect, $language) = @_;
    $language = uc(ToString($language)); $language =~ s/\s+//g;
    AssignValue('LSTDD@' . $language => $dialect); });

DefKeyVal('LST', 'printpod', '', 'true');    # NOTE: NOT YET HANDLED

DefKeyVal('LST', 'usekeywordsintag', '', 'true'); # NOTE: NOT YET HANDLED; I don't even understand it
DefKeyVal('LST', 'tagstyle', '');
DefMacro('\lst@@tagstyle Until:\end', sub {
    lstSetClassStyle('tags', $_[1]); });
DefKeyVal('LST', 'markfirstintag', '');    # NOTE: NOT YET HANDLED; I don't even understand it

DefKeyVal('LST', 'makemacrouse', '', 'true');    # NOTE: NOT YET HANDLED

#======================================================================
# 4.6 Appearance
#======================================================================
DefKeyVal('LST', 'basicstyle', '');

DefKeyVal('LST', 'identifierstyle', '');
DefMacro('\lst@@identifierstyle Until:\end', sub {
    lstSetClassStyle('identifiers', $_[1]); });

DefKeyVal('LST', 'commentstyle', '');
DefMacro('\lst@@commentstyle Until:\end', sub {
    lstSetClassStyle('comments', $_[1]); });

DefKeyVal('LST', 'stringstyle', '');
DefMacro('\lst@@stringstyle Until:\end', sub {
    lstSetClassStyle('strings', $_[1]); });

DefKeyVal('LST', 'keywordstyle', '');
DefMacro('\lst@@keywordstyle [Number] OptionalMatch:* Until:\end', sub {
    lstSetClassStyle(lstClassName('keywords', $_[1]), $_[3], uppercase => $_[2]); });
DefKeyVal('LST', 'ndkeywordstyle', '');
DefMacro('\lst@@ndkeywordstyle Until:\end', sub {
    lstSetClassStyle('keywords2', $_[1]); });

DefKeyVal('LST', 'classoffset', 'Number');

DefKeyVal('LST', 'texcsstyle', '');
DefMacro('\lst@@texcsstyle  OptionalMatch:* [Number] Until:\end', sub {
    lstSetClassStyle(lstClassName('texcss', $_[2]), $_[3], slash => $_[1]); });
DefKeyVal('LST', 'directivestyle', '');
DefMacro('\lst@@directivestyle Until:\end', sub {
    lstSetClassStyle('directives', $_[1]); });

DefKeyVal('LST', 'emph', '');
DefMacro('\lst@@emph [Number] Until:\end', sub {
    lstSetClassWords(lstClassName('emph', $_[1]), $_[2]); });
DefKeyVal('LST', 'moreemph', '');
DefMacro('\lst@@moreemph [Number] Until:\end', sub {
    lstAddClassWords(lstClassName('emph', $_[1]), $_[2]); });
DefKeyVal('LST', 'deleteemph', '');
DefMacro('\lst@@deleteemph [Number] Until:\end', sub {
    lstDeleteClassWords(lstClassName('emph', $_[1]), $_[2]); });
DefKeyVal('LST', 'emphstyle', '');
DefMacro('\lst@@emphstyle [Number] Until:\end', sub {
    lstSetClassStyle(lstClassName('emph', $_[1]), $_[2]); });

DefKeyVal('LST', 'delim', '');
# \lst@delim=**[type][style]{delim}{delim2_if_needed}
# *  allow keywords, comments & strings inside
# * effects are cummulative
DefMacro('\lst@@delim OptionalMatch:* OptionalMatch:* [] [] Until:\end', sub {
    # clear delimiters, first ???
    lstAddDelimiter('delimiter', $_[3], $_[4], $_[5],
      ($_[1] ? (recursive   => 1) : ()),
      ($_[2] ? (cummulative => 1) : ())); });
DefKeyVal('LST', 'moredelim', '');
DefMacro('\lst@@moredelim OptionalMatch:* OptionalMatch:* [] [] Until:\end', sub {
    lstAddDelimiter('delimiter', $_[3], $_[4], $_[5],
      ($_[1] ? (recursive   => 1) : ()),
      ($_[2] ? (cummulative => 1) : ())); });

#======================================================================
# 4.7 Getting characters right.
#======================================================================
DefKeyVal('LST', 'extendedchars', '', 'true');
DefMacro('\lst@@extendedchars Until:\end', sub {
    my @chars = map { UTF($_) } 128 .. 255;
    my $charslist = LookupValue('LST_CHARACTERS');
    if (ToString($_[1]) eq 'true') {
      foreach my $char (@chars) {
        $$charslist{letter}{$char} = 1; } }
    else {
      foreach my $char (@chars) {
        delete $$charslist{letter}{$char}; } }
    return; });
DefKeyVal('LST', 'inputencoding',    '');            # Ignorable?
DefKeyVal('LST', 'upquote',          '', 'true');    # Ignorable?
DefKeyVal('LST', 'tabsize',          'Number');
DefKeyVal('LST', 'showtabs',         '', 'true');    # NOTE: Not yet handled
DefKeyVal('LST', 'tab',              '');            # NOTE: Not yet handled
DefKeyVal('LST', 'showspaces',       '', 'true');
DefKeyVal('LST', 'showstringspaces', '', 'true');
DefKeyVal('LST', 'formfeed',         '');

#======================================================================
# 4.8 Line numbers
#======================================================================
# Done...
DefKeyVal('LST', 'numbers', '');    # none | left | right
DefPrimitive('\lst@@numbers Until:\end', sub {
    PushValue(LISTINGS_PREAMBLE => T_CS('\lst@@@set@numbers')); });

DefConstructor('\lst@@@set@numbers', sub {
    my ($document, %props) = @_;
    if (($props{position} || 'none') ne 'none') {
      $_[0]->addClass($_[0]->getElement, 'ltx_lst_numbers_' . $props{position}); } },
  properties => { position => sub { lstGetLiteral('numbers'); } });

DefKeyVal('LST', 'stepnumber',       'Number');
DefKeyVal('LST', 'numberfirstline',  '', 'true');
DefKeyVal('LST', 'numberstyle',      '');
DefKeyVal('LST', 'numbersep',        'Dimension');
DefKeyVal('LST', 'numberblanklines', '', 'true');
DefKeyVal('LST', 'firstnumber',      '');
DefKeyVal('LST', 'name',             '');
NewCounter('lstnumber');
DefMacro('\thelstnumber', '\arabic{lstnumber}');

#======================================================================
# 4.9 Captions
#======================================================================
# Done.
DefKeyVal('LST', 'title',   '');
DefKeyVal('LST', 'caption', '');
DefKeyVal('LST', 'label',   'Semiverbatim');
DefKeyVal('LST', 'nolol',   '', 'true');       # Ignorable

DefMacroI('\lstlistlistingname', undef, 'Listings');
DefConstructorI('\lstlistoflistings', undef,
  "<ltx:TOC lists='lol' scope='global'><ltx:title>#name</ltx:title></ltx:TOC>",
  properties => sub { (
      name => DigestIf('\lstlistlistingname')); });

DefMacroI('\lstlistingname', undef, 'Listing');
DefMacro('\thelstlisting', '\arabic{lstlisting}');
DefMacro('\thename',       '');

DefKeyVal('LST', 'captionpos',       '');             #  t,b  # done
DefKeyVal('LST', 'abovecaptionskip', 'Dimension');    # Ignorable
DefKeyVal('LST', 'belowcaptionskip', 'Dimension');    # Ignorable

#======================================================================
# 4.10 Margins and line shape
#======================================================================
# Ignorable
DefKeyVal('LST', 'linewidth',       'Dimension');
DefKeyVal('LST', 'xleftmargin',     'Dimension');
DefKeyVal('LST', 'xrightmargin',    'Dimension');
DefKeyVal('LST', 'resetmargins',    '');
DefKeyVal('LST', 'breaklines',      '', 'true');
DefKeyVal('LST', 'prebreak',        '');
DefKeyVal('LST', 'postbreak',       '');
DefKeyVal('LST', 'breakindent',     'Dimension');
DefKeyVal('LST', 'breakautoindent', '', 'true');

#======================================================================
# 4.11 Frames
#======================================================================
# Mosly ignorable, but some could be used
DefKeyVal('LST', 'frame', ''); # none | leftline | topline | bottomline | lines | single | shadowbox
our %frames = (none => undef, leftline => 'left', topline => 'top', bottomline => 'bottom',
  lines => 'topbottom', single => 'rectangle', shadowbox => 'rectangle');
DefPrimitive('\lst@@frame Until:\end', sub {
    my $name = ToString(Digest($_[1]));
    AssignValue(LISTINGS_FRAME => $frames{$name});
    PushValue(LISTINGS_PREAMBLE => T_CS('\lst@@@set@frame')); });
DefConstructor('\lst@@@set@frame', "^framed='#frame'",
  properties => { frame => sub { LookupValue('LISTINGS_FRAME'); } });

DefKeyVal('LST', 'framearound',        '');            # t|f * 4
DefKeyVal('LST', 'framesep',           'Dimension');
DefKeyVal('LST', 'rulesep',            'Dimension');
DefKeyVal('LST', 'framerule',          'Dimension');
DefKeyVal('LST', 'framexleftmargin',   'Dimension');
DefKeyVal('LST', 'framexrightmargin',  'Dimension');
DefKeyVal('LST', 'framextopmargin',    'Dimension');
DefKeyVal('LST', 'framexbottommargin', 'Dimension');
DefKeyVal('LST', 'backgroundcolor',    '');

sub lstExtractColor {
  my ($stomach, $cmd) = @_;
  my $color;
  $stomach->bgroup;
  if ($cmd->unlist) {
    Digest($cmd);
    $color = LookupValue('font')->getColor; }
  $stomach->egroup;
  return $color; }

DefPrimitive('\lst@@backgroundcolor Until:\end', sub {
    my ($stomach, $cmd) = @_;
    AssignValue(LISTINGS_BACKGROUND => lstExtractColor($stomach, $cmd));
    PushValue(LISTINGS_PREAMBLE_BEFORE => T_CS('\lst@@@set@background')); });

DefPrimitive('\lst@@@set@background', sub {
    MergeFont(background => LookupValue('LISTINGS_BACKGROUND')); });

DefKeyVal('LST', 'rulecolor', '');
DefPrimitive('\lst@@rulecolor Until:\end', sub {
    my ($stomach, $cmd) = @_;
    AssignValue(LISTINGS_RULECOLOR => lstExtractColor($stomach, $cmd));
    PushValue(LISTINGS_PREAMBLE => T_CS('\lst@@@set@rulecolor')); });

DefConstructor('\lst@@@set@rulecolor', "^framecolor='#color'",
  properties => { color => sub { LookupValue('LISTINGS_RULECOLOR'); } });

DefKeyVal('LST', 'fillcolor',    '');
DefKeyVal('LST', 'rulesepcolor', '');

#======================================================================
# 4.12 Indexing
#======================================================================
DefKeyVal('LST', 'index', '');
# HACK: The 2nd optional arg is a list of other classes that should also be indexed!!
DefMacro('\lst@@index [Number] [] Until:\end', sub {
    my ($gullet, $n, $c, $words) = @_;
    my $indexname = lstClassName('index', $n);
    if ($c) {
      my $classes = LookupValue('LST_CLASSES');
      my @classes = lstSplit($c);
      map { $$classes{$_}{index} = $indexname } @classes; }
    my $wordslist = LookupValue('LST_WORDS');
    foreach my $word (keys %$wordslist) {
      delete $$wordslist{$word}{index} if ($$wordslist{$word}{index} || '') eq $indexname; }

    my @words = lstSplit($words);
    foreach my $word (@words) {
      $$wordslist{$word}{index} = $indexname; }
    return; });

DefKeyVal('LST', 'moreindex', '');
DefMacro('\lst@@moreindex [Number] [] Until:\end', sub {
    my ($gullet, $n, $c, $words) = @_;
    my $indexname = lstClassName('index', $n);
    if ($c) {
      my $classes = LookupValue('LST_CLASSES');
      my @classes = lstSplit($c);
      map { $$classes{$_}{index} = $indexname } @classes; }
    my $wordslist = LookupValue('LST_WORDS');
    my @words     = lstSplit($words);
    foreach my $word (@words) {
      $$wordslist{$word}{index} = $indexname; }
    return; });

DefKeyVal('LST', 'deleteindex', '');
DefMacro('\lst@@deleteindex [Number] [] Until:\end', sub {
    my ($gullet, $n, $c, $words) = @_;
    my $indexname = lstClassName('index', $n);
    if ($c) {
      my $classes = LookupValue('LST_CLASSES');
      my @classes = lstSplit($c);
      foreach my $cl (@classes) {
        delete $$classes{$cl}{index} if ($$classes{$cl}{index} || '') eq $indexname; } }
    my $wordslist = LookupValue('LST_WORDS');
    foreach my $word (keys %$wordslist) {
      delete $$wordslist{$word}{index} if ($$wordslist{$word}{index} || '') eq $indexname; }
    return; });

DefKeyVal('LST', 'indexstyle', '');
DefMacro('\lst@@indexstyle [Number] Until:\end', sub {
    lstSetClassStyle(lstClassName('index', $_[1]), $_[2]); });

DefMacro('\lstindexmacro{}', '\index{{\ttfamily #1}}');

#======================================================================
# 4.13 Column alignment
#======================================================================
# Ignorable (?)
DefKeyVal('LST', 'columns',         '');
DefKeyVal('LST', 'flexiblecolumns', '', 'true');
DefKeyVal('LST', 'keepspaces',      '', 'true');
#DefKeyVal('LST','basewidth','Dimension'); #  or 2 Dimensions!!!!
DefKeyVal('LST', 'basewidth', '');            #  or 2 Dimensions!!!!
DefKeyVal('LST', 'fontadjust', '', 'true');

#======================================================================
# 4.14 Escaping to LaTeX
#======================================================================

DefKeyVal('LST', 'texcl', '', 'true');
DefMacro('\lst@@texcl Until:\end', sub {
    my ($gullet, $boole) = @_;
    my $classes = LookupValue('LST_CLASSES');
    # This only gets comments classes already defined!! Is that correct?
    my @commentclasses = grep { /^comment/ } keys %$classes;
    if (ToString($boole) eq 'true') {
      map { $$classes{$_}{eval} = 1 } @commentclasses; }
    else {
      map { delete $$classes{$_}{eval} } @commentclasses; }
    return; });

DefKeyVal('LST', 'mathescape', '', 'true');
DefMacro('\lst@@mathescape Until:\end', sub {
    my ($gullet, $boole) = @_;
    if (ToString($boole) eq 'true') {
      LookupValue('LST_DELIMITERS')->{'$'} = { open => '\$', close => '\$', class => 'mathescape', escape => 1 };
      LookupValue('LST_CLASSES')->{mathescape} = { begin => T_MATH, end => T_MATH, eval => 1 }; }
    else {
      delete(LookupValue('LST_DELIMITERS')->{'$'}); }
    return; });
DefKeyVal('LST', 'escapechar', '');
DefMacro('\lst@@escapechar Until:\end', sub {
    my ($gullet, $escape) = @_;
    $escape = lstDeslash($escape);
    if ($escape) {
      my $escapere = lstRegexp($escape);
      LookupValue('LST_DELIMITERS')->{$escape} = { open => $escapere, close => $escapere, class => 'evaluate', escape => 1 };
      LookupValue('LST_CLASSES')->{evaluate}{eval} = 1;
      delete LookupValue('LST_CHARACTERS')->{letter}{$escapere}; }
    return; });
DefKeyVal('LST', 'escapeinside', '');
DefMacro('\lst@@escapeinside Until:\end', '\ifx.#1.\else\lst@@escapeinside@#1\end\fi');
DefMacro('\lst@@escapeinside@ {} {} Until:\end', sub {
    my ($gullet, $escape1, $escape2) = @_;
    if ($escape1 && $escape2) {
      $escape1 = lstDeslash($escape1);
      $escape2 = lstDeslash($escape2);
      LookupValue('LST_DELIMITERS')->{$escape1} = {
        open => lstRegexp($escape1), close => lstRegexp($escape2),
        class => 'evaluate', escape => 1 };
      LookupValue('LST_CLASSES')->{evaluate}{eval} = 1; }
    return; });
DefKeyVal('LST', 'escapebegin', '');
DefMacro('\lst@@escapebegin Until:\end', sub {
    LookupValue('LST_CLASSES')->{evaluate}{begin} = $_[1];
    return; });
DefKeyVal('LST', 'escapeend', '');
DefMacro('\lst@@escapeend Until:\end', sub {
    LookupValue('LST_CLASSES')->{evaluate}{end} = $_[1];
    return; });

#======================================================================
# 4.15 Interface to fancyvrb
#======================================================================
# NOTE: fancyvrb Not yet handled, probably won't be
DefKeyVal('LST', 'fancyvrb', '', 'true');
DefKeyVal('LST', 'fvcmdparams',     '');
DefKeyVal('LST', 'morefvcmdparams', '');

#======================================================================
# 4.17 Language definitions
#======================================================================

# \lstdefinelanguage[dialect]{language}[base_dialect]{base_language_if_base_dialect}{keys}[required_aspects]
DefMacro('\lstdefinelanguage []{}',
  '\@ifnextchar[{\@lstdefinelanguage[#1]{#2}}{\@lstdefinelanguage[#1]{#2}[]{}}');
Let('\lst@definelanguage', '\lstdefinelanguage');

use Data::Dumper;

DefPrimitive('\@lstdefinelanguage []{}[]{} SkipSpaces RequiredKeyVals:LST []', sub {
    my ($stomach, $dialect, $language, $base_dialect, $base_language, $keyvals, $aspects) = @_;
    my @base = ();
    if ($base_language->unlist) {
      push(@base, T_OTHER('['), $base_dialect->unlist, T_OTHER(']')) if $base_dialect;
      push(@base, $base_language->unlist); }
    $language = uc(ToString($language)); $language =~ s/\s+//g;
    my $name = 'LST@LANGUAGE@' . $language;
    if ($dialect && $dialect->unlist) {
      $dialect = uc(ToString($dialect)); $dialect =~ s/\s+//g;
      $name .= '$' . $dialect; }
    $keyvals->setValue('language', Tokens(@base)) if @base;    # Probably don't need to clone, first?
    AssignValue($name => $keyvals, 'global'); });

# Seems to use <language>$<dialect> as the naming scheme.
DefPrimitive('\lstalias []{} []{}', sub {
    my ($stomach, $aliasdialect, $alias, $language, $dialect) = @_;
    # NOTE! Figure out how aliasing is supposed to work...?
    return; });

# keywords (keywordstyle in section 4.6)
DefKeyVal('LST', 'keywordprefix', '');                         #  ???
DefKeyVal('LST', 'keywords',      'Semiverbatim');
DefMacro('\lst@@keywords [Number] Until:\end', sub {
    lstSetClassWords(lstClassName('keywords', $_[1]), $_[2]); });
DefKeyVal('LST', 'morekeywords', 'Semiverbatim');
DefMacro('\lst@@morekeywords [Number] Until:\end', sub {
    lstAddClassWords(lstClassName('keywords', $_[1]), $_[2]); });
DefKeyVal('LST', 'deletekeywords', 'Semiverbatim');
DefMacro('\lst@@deletekeywords [Number] Until:\end', sub {
    lstDeleteClassWords(lstClassName('keywords', $_[1]), $_[2]); });

DefKeyVal('LST', 'ndkeywords', 'Semiverbatim');
DefMacro('\lst@@ndkeywords Until:\end', sub {
    lstSetClassWords('keywords2', $_[1]); });
DefKeyVal('LST', 'morendkeywords', 'Semiverbatim');
DefMacro('\lst@@morendkeywords Until:\end', sub {
    lstAddClassWords('keywords2', $_[1]); });
DefKeyVal('LST', 'deletendkeywords', 'Semiverbatim');
DefMacro('\lst@@deletendkeywords Until:\end', sub {
    lstDeleteClassWords('keywords2', $_[1]); });

DefKeyVal('LST', 'texcs', '');
DefMacro('\lst@@texcs [Number] Until:\end', sub {
    AssignValue('LST@TEXCS' => 1);
    lstSetClassWords(lstClassName('texcss', $_[1]), $_[2], "\\"); });
DefKeyVal('LST', 'moretexcs', '');
DefMacro('\lst@@moretexcs [Number] Until:\end', sub {
    AssignValue('LST@TEXCS' => 1);
    lstAddClassWords(lstClassName('texcss', $_[1]), $_[2], "\\"); });
DefKeyVal('LST', 'deletetexcs', '');
DefMacro('\lst@@deletetexcs [Number] Until:\end', sub {
    lstDeleteClassWords(lstClassName('texcss', $_[1]), $_[2], "\\"); });

# directives (directivestyle in section 4.6)
DefKeyVal('LST', 'directives', 'Semiverbatim');
DefMacro('\lst@@directives Until:\end', sub {
    lstSetClassWords('directives', $_[1]); });
DefKeyVal('LST', 'moredirectives', 'Semiverbatim');
DefMacro('\lst@@moredirectives Until:\end', sub {
    lstAddClassWords('directives', $_[1]); });
DefKeyVal('LST', 'deletedirectives', 'Semiverbatim');
DefMacro('\lst@@deletedirectives Until:\end', sub {
    lstDeleteClassWords('directives', $_[1]); });

DefKeyVal('LST', 'sensitive', '', 'true');
DefKeyVal('LST', 'alsoletter', '');
DefMacro('\lst@@alsoletter Until:\end', sub {
    lstSetCharacterClass('letter', $_[1]); });
DefKeyVal('LST', 'alsodigit', '');
DefMacro('\lst@@alsodigit Until:\end', sub {
    lstSetCharacterClass('digit', $_[1]); });
DefKeyVal('LST', 'alsoother', '');
DefMacro('\lst@@alsoother Until:\end', sub {
    lstSetCharacterClass('other', $_[1]); });
DefKeyVal('LST', 'otherkeywords', '');    # NOTE: Not yet handled

DefKeyVal('LST', 'tag', '');
DefMacro('\lst@@tag OptionalMatch:* OptionalMatch:* [] Until:\end', sub {
    lstAddDelimiter('delimiter', $_[3], 'tagstyle', $_[4],
      ($_[1] ? (recursive   => 1) : ()),
      ($_[2] ? (cummulative => 1) : ())); });

# Strings
DefKeyVal('LST', 'string', '');
DefMacro('\lst@@string [] Until:\end', sub {
    lstAddDelimiter('string', $_[1], 'stringstyle', $_[2]); });
DefKeyVal('LST', 'morestring', '');
DefMacro('\lst@@morestring [] Until:\end', sub {
    lstAddDelimiter('string', $_[1], 'stringstyle', $_[2]); });
DefKeyVal('LST', 'deletestring', '');
# How to handle???

# Comments
DefKeyVal('LST', 'comment', '');
DefMacro('\lst@@comment [] [] Until:\end', sub {
    lstAddDelimiter('comment', $_[1], 'commentstyle', $_[3]); });
DefKeyVal('LST', 'morecomment', '');
DefMacro('\lst@@morecomment [] [] Until:\end', sub {
    lstAddDelimiter('comment', $_[1], 'commentstyle', $_[3]); });
DefKeyVal('LST', 'deletecomment', '');
# How to handle???

DefKeyVal('LST', 'keywordcomment',          '');
DefKeyVal('LST', 'morekeywordcomment',      '');
DefKeyVal('LST', 'deletekeywordcomment',    '');
DefKeyVal('LST', 'keywordcommentsemicolon', '');
DefKeyVal('LST', 'podcomment',              '', 'true');

DefPrimitive('\lstloadlanguages Semiverbatim', undef);

#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# The listing parser
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# Process the listing
#   The listing is supplied as a list of strings
#   The result is a Tokens containing the formatted results
sub lstProcess {
  my ($mode, $text) = @_;

  # === Return nothing if print is false
  return Tokens() unless $text && lstGetBoolean('print');

# === Possibly strip trailing blank lines.
# NOTE: Not sure if this is supposed to trim from the whole listing, or the requested subset(s) of lines!
  if (!lstGetBoolean('showlines')) {    # trim empty lines from end.
    $text =~ s/\s*$//s; }

  # === Establish line numbering parameters
  my $name        = lstGetLiteral('name');
  my $firstnumber = lstGetLiteral('firstnumber');
  my $line0       = (($firstnumber eq 'last')
    ? (LookupValue('LISTINGS_LAST_NUMBER') || 1)
    : ($firstnumber eq 'auto'
      ? (($name && LookupValue('LISTINGS_LAST_NUMBER_' . $name)) || 1)
      : $firstnumber));
  my $numpos = ((lstGetNumber('stepnumber') == 0) ? 'none' : lstGetLiteral('numbers'));
  AssignValue('LISTINGS_NEEDS_NUMBER' => (($numpos ne 'none') && lstGetBoolean('numberfirstline')));

  # === Create a line test based on linerange, or firstline & lastline
  my $linetest = sub { 1; };
  my ($l1, $l2);
  if (my $lr = lstGetLiteral('linerange')) {
    my @lr = map { [split(/-/, $_)] } lstSplit($lr);
    $linetest = sub { grep { ($$_[0] <= $_[0]) && ($_[0] <= $$_[1]) } @lr; }; }
  elsif (($l1 = lstGetNumber('firstline'))
    && ($l2 = lstGetNumber('lastline'))) {
    $linetest = sub { ($l1 <= $_[0]) && ($_[0] <= $l2); }; }

  local $LaTeXML::linetest = $linetest;
  # === These hashes have been set up by "activating" the various keywords.
  my $words      = LookupValue('LST_WORDS');
  my $delimiters = LookupValue('LST_DELIMITERS');
  my $classes    = LookupValue('LST_CLASSES');
  my $characters = LookupValue('LST_CHARACTERS');
  # === Extract some regexps to match various important things
  my $letter_re = join('', sort keys %{ $$characters{letter} });
  my $digit_re  = join('', sort keys %{ $$characters{digit} });
  local $LaTeXML::ID_RE = (LookupValue('LST@TEXCS') ? "\\\\?" : '') . "[$letter_re][$letter_re$digit_re]*";
  local $LaTeXML::DELIM_RE = join('|', map { $$delimiters{$_}{open} } sort keys %$delimiters);
  local $LaTeXML::ESCAPE_RE = join('|', map { $$delimiters{$_}{open} }
      grep { $$delimiters{$_}{escape} } sort keys %$delimiters);
  local $LaTeXML::QUOTED_RE = "\\\\\\\\";    # start w/ backslashed backslash?
  local $LaTeXML::SPACE = (lstGetBoolean('showspaces') ? T_CS('\@lst@visible@space') : T_CS("~"));
  local $LaTeXML::CASE_SENSITIVE = lstGetBoolean('sensitive');
  if (!$LaTeXML::CASE_SENSITIVE) {           # Clunky, but until know, we don't know
    foreach my $word (keys %$words) {
      $$words{ uc($word) } = $$words{$word}; } }

# === Start processing
# This whole set of vars probably needs to be adjusted,
# since we'll need to recognize constructs inside strings that we've already pulled out (strings,comments)
# Better would be to treat the whole string.
# then gobble lines etc, can probably work...
  local $LaTeXML::linenum   = $line0;
  local $LaTeXML::colnum    = 0;
  local $LaTeXML::listing   = $text;
  local $LaTeXML::mode      = $mode;
  local $LaTeXML::linestart = undef;
  local $LaTeXML::emptyfrom = undef;
  local @LaTeXML::lsttokens = (T_BEGIN);

  lstProcessPush(@{ LookupValue('LISTINGS_PREAMBLE') });
  lstProcessPush(lstGetTokens('basicstyle')->unlist);

  while ($LaTeXML::listing && !&$linetest($LaTeXML::linenum)) {    # Ignore initial lines?
    $LaTeXML::listing =~ s/^.*?\n//s;
    $LaTeXML::linenum++; }
  if ($mode ne 'inline') {
    lstProcessPush(Invocation(T_CS('\setcounter'), T_OTHER('lstnumber'), Number($LaTeXML::linenum)));
    lstProcessStartLine(); }
  lstProcess_internal();
  if ($mode ne 'inline') {
    lstProcessEndLine(); }

  # === Save line number for possible later use.
  AssignValue('LISTINGS_LAST_NUMBER' => CounterValue('lstnumber')->valueOf, 'global');
  AssignValue('LISTINGS_LAST_NUMBER_' . $name => CounterValue('lstnumber')->valueOf, 'global') if $name;
  # Remove empty trailing lines, if any (GACK!)
  @LaTeXML::lsttokens = @LaTeXML::lsttokens[0 .. $LaTeXML::emptyfrom - 1] if $LaTeXML::emptyfrom;
  # === And finally, return the tokens we've constructed.
  return Tokens(@LaTeXML::lsttokens, T_END); }

sub lstProcessPush {
  my (@stuff) = @_;
  push(@LaTeXML::lsttokens, @stuff);
  return; }

sub lstProcessStartLine {
  my $numpos = ((lstGetNumber('stepnumber') == 0) ? 'none' : lstGetLiteral('numbers'));
  $LaTeXML::linestart = scalar(@LaTeXML::lsttokens); # Remember where line started, for potential truncation
  lstProcessPush(T_CS('\@lst@startline'),
    T_BEGIN, ($numpos ne 'none') ? lstDoNumber($LaTeXML::listing =~ /^\s*?\n/s) : (), T_END);
  return; }

sub lstProcessEndLine {
  if ($LaTeXML::colnum == 0) {    # Line was empty; remember where emptyness started...
    $LaTeXML::emptyfrom = $LaTeXML::linestart unless $LaTeXML::emptyfrom; }
  else {
    $LaTeXML::emptyfrom = undef; }
  lstProcessPush(T_CS('\@lst@endline'));
  return; }

sub lstProcess_internal {
  my ($end_re, $outerclass) = @_;
  my $numpos     = ((lstGetNumber('stepnumber') == 0) ? 'none' : lstGetLiteral('numbers'));
  my $words      = LookupValue('LST_WORDS');
  my $delimiters = LookupValue('LST_DELIMITERS');
  my $classes    = LookupValue('LST_CLASSES');
  while ($LaTeXML::listing) {
    # Matched the ending regular expression? (typically a close delimiter)
    if ($end_re && $LaTeXML::listing =~ s/^($end_re)//s) {
      $LaTeXML::colnum += length($1);
      last; }
    # Various kinds of delimited expressions: escapes, strings, comments, general delimiters.
    elsif ($LaTeXML::DELIM_RE && $LaTeXML::listing =~ s/^($LaTeXML::DELIM_RE)//s) {
      my $open = $1;
      $LaTeXML::colnum += length($1);
      my $delim     = $$delimiters{$1};
      my $classname = $$delim{class};
      lstProcessPush(lstClassBegin($classname));
# With escapes or texcl, some might be evaluated as TeX; those we match the close delim and simply tokenize.
      if (lstClassProperty($classname, 'eval')) { # If this is a comment with texcl applied, just match & expand
        if ($LaTeXML::listing =~ s/^(.*?)($$delim{close})//s) {    # Simply match until closing regexp
          my ($string, $close) = ($1, $2);
          my @l = split("\n", $string . $close);    # This is the only(?) potentially multiline block
          $LaTeXML::linenum += scalar(@l) - 1 if @l > 2;    # So adjust line & column
          lstProcessPush(TokenizeBalanced($string)); } }
      # Others become tricky because the contents of the string, comment etc may need to be processed
      # including matching _some_ delimited expressions!
      #   escaped constructs are always matched.
      #   nested : allows comments to be nested (ie the SAME delimiter pair)
      #   recursive: allows any(?) "comments, strings & keywords" to be matched inside.
      else {
        local $LaTeXML::DELIM_RE = ($$delim{recursive}
          ? $LaTeXML::DELIM_RE
          : join('|', grep { $_ } $LaTeXML::ESCAPE_RE, $$delim{nested} && $$delim{open}));
        local $LaTeXML::ID_RE = ($$delim{recursive} ? $LaTeXML::ID_RE : undef);
        local $LaTeXML::QUOTED_RE = join('|', grep { $_ } $LaTeXML::QUOTED_RE, $$delim{quoted});
        local $LaTeXML::SPACE = ($classname && ($classname =~ /^string/) && lstGetBoolean('showstringspaces')
          ? T_CS('\@lst@visible@space') : $LaTeXML::SPACE);
        # Recurse [note that eval should make the individual tokens tokenize as usual!]
        lstProcess_internal($$delim{close}, $classname); }
      lstProcessPush(lstClassEnd($classname)); }
    # Identifiers (possibly keywords, or other classes)
    elsif ($LaTeXML::ID_RE && $LaTeXML::listing =~ s/^($LaTeXML::ID_RE)//) {
      $LaTeXML::colnum += length($1);    # ?
      my $word      = $1;
      my $lookup    = ($LaTeXML::CASE_SENSITIVE ? $word : uc($word));
      my $classname = ($outerclass ? undef : $$words{$lookup}{class} || 'identifiers');
      my @tokens    = map { lstRescan($_) } Explode($word);                               # rescan??
      if (my $indexname = $$words{$lookup}{index} || lstClassProperty($classname, 'index')) { # Should be indexed?
        if (my $index = $indexname && $$classes{$indexname}) {
          lstProcessPush(lstRescan($$index{begin})->unlist, T_BEGIN, @tokens, T_END); } }
      lstProcessPush(($classname ? (lstClassBegin($classname), @tokens, lstClassEnd($classname))
          : @tokens)); }

    # NOTE: keywordprefix & otherkeywords probably need a specific regexp
    # Perhaps a special keywords_re : otherkeywords | keywordprefix$LaTeXML::ID_RE => keyword

    # Various kinds of whitespace, newlines, etc.
    elsif ($LaTeXML::listing =~ s/^\s*?\n//s) {    # Newline
      if ($LaTeXML::mode ne 'inline') {
        lstProcessEndLine();
        lstProcessPush(Invocation(T_CS('\stepcounter'), T_OTHER('lstnumber')));
        $LaTeXML::linenum++;                       # Increment line number
        $LaTeXML::colnum = 0;                      # Reset column number
          # NOTE: should ignore blank lines at end of listing, even if they aren't the last line of the code!
          # NOTE: should handle showlines, emptylines keywords
        while ($LaTeXML::listing && !&$LaTeXML::linetest($LaTeXML::linenum)) {    # Ignore next line?
          $LaTeXML::listing =~ s/^.*?(\n|$)//s;
          lstProcessPush(Invocation(T_CS('\stepcounter'), T_OTHER('lstnumber')));
          $LaTeXML::linenum++; }
        lstProcessStartLine(); }
      # === Possibly remove $gobble chars from line
      my $gobble = lstGetNumber('gobble');
      (map { $LaTeXML::listing =~ s/^.// } 1 .. $gobble) if $gobble;
    }
    elsif ($LaTeXML::listing =~ s/^\f//s) {                                       # Formfeed
      lstProcessPush(lstGetTokens('formfeed')->unlist);
      $LaTeXML::colnum++; }
    elsif ($LaTeXML::listing =~ s/^([\t\s]+)//s) {                                # Tab expansion
      my $s       = $1;
      my $n       = 0;
      my $tabsize = lstGetNumber('tabsize') || 1;
      foreach my $c (split(//, $s)) {
        $n += ($c eq ' ' ? 1 : ($tabsize - (($LaTeXML::colnum + $n) % $tabsize))); }
      lstProcessPush(lstClassBegin('spaces'), (map { $LaTeXML::SPACE } 1 .. $n), lstClassEnd('spaces'));
      $LaTeXML::colnum += $n; }
    # Quoted are typically quoted delimiters.
    elsif ($LaTeXML::QUOTED_RE && $LaTeXML::listing =~ s/^($LaTeXML::QUOTED_RE)//) { # Something quoted.
          # Don't just past together, and watch for leading \ (a common quoter)
      lstProcessPush(map { ($_ eq '\\' ? T_CS('\textbackslash') : T_OTHER($_)) } split('', $1));

      $LaTeXML::colnum += length($1); }
    else {
      if ($LaTeXML::listing =~ s/^(.)//s) {    # Anything else, just pass through.
        lstProcessPush(lstRescan(T_OTHER($1))); }
      $LaTeXML::colnum++; }
  }
  return; }

# Perversely guarantee that the tokenization is balanced to avoid peculiar bugs in bad TeX
sub TokenizeBalanced {
  my ($string) = @_;
  my $tokens = Tokenize($string);
  my @toks =
    my $level = 0;
  foreach my $t ($tokens->unlist) {
    if    (T_BEGIN->equals($t)) { $level++; }
    elsif (T_END->equals($t))   { $level--; } }
  if ($level != 0) {    # Probably doesn't work all the time, but let's try
    my @toks = $tokens->unlist;
    while ($level > 0) { push(@toks, T_END); $level--; }
    while ($level < 0) { unshift(@toks, T_BEGIN); $level--; }
    $tokens = Tokens(@toks); }
  return $tokens; }

DefConstructor('\@listingKeyword Semiverbatim {}',
  "?#class(<ltx:text class='ltx_lst_#class' _noautoclose='1'>#2</ltx:text>)(#2)",
  properties => { class => sub {
      my $classname = ToString($_[1]);
      my $class     = $classname && LookupValue('LST_CLASSES')->{$classname};
      my $cssclass  = $class && $$class{cssclass};
      $cssclass; } });

# It's conceivable that a group is INTENDED to carry over across lines?
# If so, we'd have to maintain a stack..
# But we have to be careful about a group closing ltx:text that may have autoclosed on prev.line!
DefConstructor('\@listingGroup Semiverbatim {}',
  #   "<ltx:text class='#1'>#2</ltx:text>");
  "<ltx:text class='#1'>#2",
  # sorta like maybeCloseElement, except only if CURRENT
  afterConstruct => sub { my ($doc) = @_;
    if ($doc->getNodeQName($doc->getElement) eq 'ltx:text') {
      $doc->closeElement('ltx:text'); } });

sub lstClassBegin {
  my ($classname) = @_;
  my @open        = ();
  my %classes     = ();
  if (($classname || '') eq 'spaces') { $classes{space} = 1; }
  while (my $class = $classname && LookupValue('LST_CLASSES')->{$classname}) {
    if (my $css = $$class{cssclass}) {
      $classes{$css} = 1; }
    if (my $begin = $$class{begin}) {
      unshift(@open, lstRescan($begin)->unlist); }
    $classname = $$class{class}; }
  return (T_BEGIN, T_CS('\@listingGroup'),
    T_BEGIN, T_OTHER(join(' ', map { 'ltx_lst_' . $_ } sort keys %classes)), T_END,
    T_BEGIN, @open); }

sub lstClassEnd {
  my ($classname) = @_;
  my @close = ();
  while (my $class = $classname && LookupValue('LST_CLASSES')->{$classname}) {
    if (my $end = $$class{end}) {
      push(@close, lstRescan($end)->unlist); }
    $classname = $$class{class}; }
  return (@close, T_END, T_END); }

sub lstClassProperty {
  my ($classname, $property) = @_;
  my $class = $classname && LookupValue('LST_CLASSES')->{$classname};
  return ($class && ($$class{$property}
      ? $$class{$property}
      : lstClassProperty($$class{class}, $property))); }

DefConstructor('\@lst@startline{}', "<ltx:listingline xml:id='#id'>#1",
  properties => sub { RefStepID('lstnumber'); });
DefConstructor('\@lst@endline',      "</ltx:listingline>");
DefConstructor('\@lst@linenumber{}', "<ltx:tags><ltx:tag>#1</ltx:tag></ltx:tags>");

DefConstructor('\@lst@visible@space', "\x{2423}");

sub lstDoNumber {
  my ($isempty) = @_;
  if ((LookupValue('LISTINGS_NEEDS_NUMBER')
      || ((($LaTeXML::linenum - 1) % lstGetNumber('stepnumber')) == 0))
    && (lstGetBoolean('numberblanklines') || !$isempty)) {
    AssignValue('LISTINGS_NEEDS_NUMBER' => 0);
    return Invocation(T_CS('\@lst@linenumber'),
      Tokens(T_BEGIN, lstGetTokens('numberstyle')->unlist, T_CS('\thelstnumber'), T_END)); }
  else {
    return Invocation(T_CS('\@lst@linenumber'), Tokens()); } }

#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# Configuration
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
# Initialize the various parameters...

RawTeX(<<'EoTeX');
\lstset{
 alsoletter={abcdefghiklmnopqrstuvwxyzABCDEFGHIKLMNOPQRSTUVWXYZ@$\_},
 alsodigit={0123456789},
 alsoother={!"#\%&'()*+,-./:;<=>?[\\]^\{|\}~},
 float=tbp,floatplacement=tbp,aboveskip=\medskipamount,belowskip=\medskipamount,
 lineskip=0pt,boxpos=c,
 print=true,firstline=1,lastline=9999999,showlines=false,emptylines=9999999,gobble=0,
 style={},language={},printpod=false,usekeywordsintag=true,tagstyle={},
 markfirstintag=false,makemacrouse=true,
 basicstyle={},identifierstyle={},commentstyle=\itshape,stringstyle={},
 keywordstyle=\bfseries,classoffset=0,
 emph={},delim={},
 extendedchars=false,inputencoding={},upquote=false,tabsize=8,showtabs=false,
 tabs={},showspaces=false,showstringspaces=true,formfeed=\bigbreak,
 numbers=none,stepnumber=1,numberfirstline=false,numberstyle={},numbersep=10pt,
 numberblanklines=true,firstnumber=auto,name={},
 title={},caption={},label={},nolol=false,
 captionpos=t,abovecaptionskip=\smallskipamount,belowcaptionskip=\smallskipamount,
 linewidth=\linewidth,xleftmargin=0pt,xrightmargin=0pt,resetmargins=false,breaklines=false,
 prebreak={},postbreak={},breakindent=20pt,breakautoindent=true,
 frame=none,frameround=ffff,framesep=3pt,rulesep=2pt,framerule=0.4pt,
 framexleftmargin=0pt,framexrightmargin=0pt,framextopmargin=0pt,framexbottommargin=0pt,
 backgroundcolor={},rulecolor={},fillcolor={},rulesepcolor={},
 frameshape={},
 index={},indexstyle=\lstindexmacro,
 columns=[c]fixed,flexiblecolumns=false,keepspaces=false,basewidth={0.6em,0.45em},
 fontadjust=false,texcl=false,mathescape=false,escapechar={},escapeinside={},
 escapebegin={},escapeend={},
 fancyvrb=false,fvcmdparams=\overlay1,morefvcmdparams={},
 ndkeywordstyle=keywordstyle,texcsstyle=keywordstyle,directivestyle=keywordstyle
}
EoTeX

#======================================================================
# Finally, we want to load the definitions from the configurations...
# Actually, we should just load .cfg
# and the extra files should be loaded as needed, but...
sub lstLoadConfiguration {
  InputDefinitions("listings", type => 'cfg');
  AssignValue(LST_LANGUAGE_FILES => [lstSplit(Digest(T_CS('\lstlanguagefiles')))], 'global');

  # Now, if you want to read in all definitions immediately, you could do this (on preload)
  # otherwise, they'll be read in whenever missing languages are used.
  if (LookupValue('InitialPreloads')) {
    while (my $file = ShiftValue('LST_LANGUAGE_FILES')) {
      InputDefinitions($file, noerror => 1); } }
  return; }

lstLoadConfiguration();

#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1;