AnonSec Shell

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . AnonSec Shell AnonSec Shell
Server IP : 20.75.53.88 / Your IP : 216.73.217.165 [ Reverse IP ]
Web Server : Apache
System : Linux VMRALP-3 4.4.0-256-generic #290~14.04.1-Ubuntu SMP Thu Jun 20 09:24:50 UTC 2024 x86_64
User : www-data ( 33)
PHP Version : 5.5.9-1ubuntu4.29+esm15
Disable Function : pcntl_alarm,pcntl_fork,pcntl_waitpid,pcntl_wait,pcntl_wifexited,pcntl_wifstopped,pcntl_wifsignaled,pcntl_wexitstatus,pcntl_wtermsig,pcntl_wstopsig,pcntl_signal,pcntl_signal_dispatch,pcntl_get_last_error,pcntl_strerror,pcntl_sigprocmask,pcntl_sigwaitinfo,pcntl_sigtimedwait,pcntl_exec,pcntl_getpriority,pcntl_setpriority,
Domains : 3 Domains
MySQL : ON | cURL : OFF | WGET : ON | Perl : ON | Python : ON | Sudo : ON | Pkexec : ON
Directory : /var/www/html/siac/cnsmaracanau/admin/components/htmldoc-master/htmldoc/
Upload File :
[ HOME ] [ BACKUP SHELL ] [ JUMPING ] [ MASS DEFACE ] [ SCAN ROOT ] [ SYMLINK ]
Current File : /var/www/html/siac/cnsmaracanau/admin/components/htmldoc-master/htmldoc/htmllib.cxx
/*
 * HTML parsing routines for HTMLDOC, a HTML document processing program.
 *
 * Copyright 2011-2021 by Michael R Sweet.
 * Copyright 1997-2010 by Easy Software Products.  All rights reserved.
 *
 * This program is free software.  Distribution and use rights are outlined in
 * the file "COPYING".
 */

/*
 * Include necessary headers.
 */

#include "htmldoc.h"
#include "http.h"
#include <ctype.h>


/*
 * Markup strings...
 */

const char	*_htmlMarkups[] =
		{
		  "",		/* MARKUP_NONE */
		  "!--",	/* MARKUP_COMMENT */
		  "!DOCTYPE",
		  "a",
		  "acronym",
		  "address",
		  "applet",
		  "area",
		  "b",
		  "base",
		  "basefont",
		  "big",
		  "blink",
		  "blockquote",
		  "body",
		  "br",
		  "caption",
		  "center",
		  "cite",
		  "code",
		  "col",
		  "colgroup",
		  "dd",
		  "del",
		  "dfn",
		  "dir",
		  "div",
		  "dl",
		  "dt",
		  "em",
		  "embed",
		  "font",
		  "form",
		  "frame",
		  "frameset",
		  "h1",
		  "h2",
		  "h3",
		  "h4",
		  "h5",
		  "h6",
		  "h7",
		  "h8",
		  "h9",
		  "h10",
		  "h11",
		  "h12",
		  "h13",
		  "h14",
		  "h15",
		  "head",
		  "hr",
		  "html",
		  "i",
		  "img",
		  "input",
		  "ins",
		  "isindex",
		  "kbd",
		  "li",
		  "link",
		  "map",
		  "menu",
		  "meta",
		  "multicol",
		  "nobr",
		  "noframes",
		  "ol",
		  "option",
		  "p",
		  "pre",
		  "s",
		  "samp",
		  "script",
		  "select",
		  "small",
		  "spacer",
		  "span",
		  "strike",
		  "strong",
		  "style",
		  "sub",
		  "sup",
		  "table",
		  "tbody",
		  "td",
		  "textarea",
		  "tfoot",
		  "th",
		  "thead",
		  "title",
		  "tr",
		  "tt",
		  "u",
		  "ul",
		  "var",
		  "wbr"
		};

const char	*_htmlCurrentFile = "UNKNOWN";
					/* Current file */
const char	*_htmlData = HTML_DATA;	/* Data directory */
float		_htmlPPI = 80.0f;	/* Image resolution */
int		_htmlGrayscale = 0;	/* Grayscale output? */
uchar		_htmlTextColor[255] =	/* Default text color */
		{ 0 };
float		_htmlBrowserWidth = 680.0f;
					/* Browser width for pixel scaling */
float		_htmlSizes[8] =		/* Point size for each HTML size */
		{ 6.0f, 8.0f, 9.0f, 11.0f, 14.0f, 17.0f, 20.0f, 24.0f };
float		_htmlSpacings[8] =	/* Line height for each HTML size */
		{ 7.2f, 9.6f, 10.8f, 13.2f, 16.8f, 20.4f, 24.0f, 28.8f };
typeface_t	_htmlBodyFont = TYPE_TIMES,
		_htmlHeadingFont = TYPE_HELVETICA;

int		_htmlInitialized = 0;	/* Initialized glyphs yet? */
char		_htmlCharSet[256] = "iso-8859-1";
					/* Character set name */
int		_htmlWidthsLoaded[TYPE_MAX][STYLE_MAX] =
		{			/* Have the widths been loaded? */
		  { 0, 0, 0, 0 },
		  { 0, 0, 0, 0 },
		  { 0, 0, 0, 0 },
		  { 0, 0, 0, 0 },
		  { 0, 0, 0, 0 },
		  { 0, 0, 0, 0 },
		  { 0, 0, 0, 0 },
		  { 0, 0, 0, 0 }
		};
short		_htmlWidths[TYPE_MAX][STYLE_MAX][256];
					/* Character widths of fonts */
short		_htmlWidthsAll[TYPE_MAX][STYLE_MAX][65536];
                                        /* Unicode widths of fonts */
int		_htmlUnicode[256];	/* Character to Unicode mapping */
uchar           _htmlCharacters[65536]; /* Unicode to character mapping */
int             _htmlUTF8 = 0;          /* Doing UTF-8? */
const char	*_htmlGlyphsAll[65536];	/* Character glyphs for Unicode */
const char	*_htmlGlyphs[256];	/* Character glyphs for charset */
int		_htmlNumSorted = 0;	/* Number of sorted glyphs */
const char	*_htmlSorted[256];	/* Sorted character glyphs for charset */
uchar		_htmlSortedChars[256];	/* Sorted character indices */
const char	*_htmlFonts[TYPE_MAX][STYLE_MAX] =
		{
		  {
		    "Courier",
		    "Courier-Bold",
		    "Courier-Oblique",
		    "Courier-BoldOblique"
		  },
		  {
		    "Times-Roman",
		    "Times-Bold",
		    "Times-Italic",
		    "Times-BoldItalic"
		  },
		  {
		    "Helvetica",
		    "Helvetica-Bold",
		    "Helvetica-Oblique",
		    "Helvetica-BoldOblique"
		  },
		  {
		    "Monospace",
		    "Monospace-Bold",
		    "Monospace-Oblique",
		    "Monospace-BoldOblique"
		  },
		  {
		    "Serif-Roman",
		    "Serif-Bold",
		    "Serif-Oblique",
		    "Serif-BoldOblique"
		  },
		  {
		    "Sans",
		    "Sans-Bold",
		    "Sans-Oblique",
		    "Sans-BoldOblique"
		  },
		  {
		    "Symbol",
		    "Symbol",
		    "Symbol",
		    "Symbol"
		  },
		  {
		    "Dingbats",
		    "Dingbats",
		    "Dingbats",
		    "Dingbats"
		  }
		};
int		_htmlStandardFonts[TYPE_MAX] =
		{
		  1,	// Courier
		  1,	// Times
		  1,	// Helvetica
		  0,	// Monospace
		  0,	// Sans
		  0,	// Serif
		  1,	// Symbol
		  1	// Dingbats
		};


/*
 * Local functions.
 */

extern "C" {
typedef int	(*compare_func_t)(const void *, const void *);
}

static int	write_file(tree_t *t, FILE *fp, int col);
static int	compare_variables(var_t *v0, var_t *v1);
static int	compare_markups(uchar **m0, uchar **m1);
static void	delete_node(tree_t *t);
static void	insert_space(tree_t *parent, tree_t *t);
static int	parse_markup(tree_t *t, FILE *fp, int *linenum);
static int	parse_variable(tree_t *t, FILE *fp, int *linenum);
static int	compute_size(tree_t *t);
static int	compute_color(tree_t *t, uchar *color);
static int	get_alignment(tree_t *t);
static const char *fix_filename(char *path, char *base);
static int      utf8_getc(int ch, FILE *fp);

#define issuper(x)	((x) == MARKUP_CENTER || (x) == MARKUP_DIV ||\
			 (x) == MARKUP_BLOCKQUOTE)
#define isblock(x)	((x) == MARKUP_ADDRESS || \
			 (x) == MARKUP_P || (x) == MARKUP_PRE ||\
			 ((x) >= MARKUP_H1 && (x) <= MARKUP_H6) ||\
			 (x) == MARKUP_HR || (x) == MARKUP_TABLE)
#define islist(x)	((x) == MARKUP_DL || (x) == MARKUP_OL ||\
			 (x) == MARKUP_UL || (x) == MARKUP_DIR ||\
			 (x) == MARKUP_MENU)
#define islentry(x)	((x) == MARKUP_LI || (x) == MARKUP_DD ||\
			 (x) == MARKUP_DT)
#define istable(x)	((x) == MARKUP_TBODY || (x) == MARKUP_THEAD ||\
			 (x) == MARKUP_TFOOT || (x) == MARKUP_TR)
#define istentry(x)	((x) == MARKUP_TD || (x) == MARKUP_TH)

#ifdef DEBUG
static uchar	indent[255] = "";
#endif /* DEBUG */


/*
 * 'htmlReadFile()' - Read a file for HTML markup codes.
 */

tree_t *				// O - Pointer to top of file tree
htmlReadFile(tree_t     *parent,	// I - Parent tree entry
             FILE       *fp,		// I - File pointer
	     const char *base)		// I - Base directory for file
{
  int		ch;			// Character from file
  uchar		*ptr,			// Pointer in string
		entity[16],		// Character entity name (&#nnn; or &name;)
		*eptr;			// Pointer in entity string
  tree_t	*tree,			// "top" of this tree
		*t,			// New tree entry
		*prev,			// Previous tree entry
		*temp;			// Temporary looping var
  int		descend;		// Descend into node?
  FILE		*embed;			// File pointer for EMBED
  char		newbase[1024];		// New base directory for EMBED
  uchar		*filename,		// Filename for EMBED tag
		*face,			// Typeface for FONT tag
		*color,			// Color for FONT tag
		*size,			// Size for FONT tag
		*type,			// Type for EMBED tag
		*span;			// Value for SPAN tag
  int		sizeval;		// Size value from FONT tag
  int		linenum;		// Line number in file
  static uchar	s[10240];		// String from file
  static int	have_whitespace = 0;	// Non-zero if there was leading whitespace


  DEBUG_printf(("htmlReadFile(parent=%p, fp=%p, base=\"%s\")\n",
                (void *)parent, (void *)fp, base ? base : "(null)"));

#ifdef DEBUG
  indent[0] = '\0';
#endif // DEBUG

 /*
  * Start off with no previous tree entry...
  */

  prev = NULL;
  tree = NULL;

 /*
  * Parse data until we hit end-of-file...
  */

  linenum = 1;

  while ((ch = getc(fp)) != EOF)
  {
   /*
    * Ignore leading whitespace...
    */

    if (parent == NULL || !parent->preformatted)
    {
      while (isspace(ch))
      {
	if (ch == '\n')
	  linenum ++;

        have_whitespace = 1;
        ch              = getc(fp);
      }

      if (ch == EOF)
        break;
    }

   /*
    * Allocate a new tree entry - use calloc() to get zeroed data...
    */

    t = (tree_t *)calloc(sizeof(tree_t), 1);
    if (t == NULL)
    {
#ifndef DEBUG
      progress_error(HD_ERROR_OUT_OF_MEMORY,
                     "Unable to allocate memory for HTML tree node!");
#endif /* !DEBUG */
      break;
    }

   /*
    * Set/copy font characteristics...
    */

    if (parent == NULL)
    {
      t->halignment   = ALIGN_LEFT;
      t->valignment   = ALIGN_BOTTOM;
      t->typeface     = _htmlBodyFont;
      t->size         = SIZE_P;

      compute_color(t, _htmlTextColor);
    }
    else
    {
      t->link          = parent->link;
      t->halignment    = parent->halignment;
      t->valignment    = parent->valignment;
      t->typeface      = parent->typeface;
      t->size          = parent->size;
      t->style         = parent->style;
      t->superscript   = parent->superscript;
      t->subscript     = parent->subscript;
      t->preformatted  = parent->preformatted;
      t->indent        = parent->indent;
      t->red           = parent->red;
      t->green         = parent->green;
      t->blue          = parent->blue;
      t->underline     = parent->underline;
      t->strikethrough = parent->strikethrough;
    }

   /*
    * See what the character was...
    */

    if (ch == '<')
    {
     /*
      * Markup char; grab the next char to see if this is a /...
      */

      ch = getc(fp);

      if (isspace(ch) || ch == '=' || ch == '<')
      {
       /*
        * Sigh...  "<" followed by anything but an element name is
	* invalid HTML, but so many people have asked for this to
	* be supported that we have added this hack...
	*/

	progress_error(HD_ERROR_HTML_ERROR, "Unquoted < on line %d of %s.",
	               linenum, _htmlCurrentFile);

	if (ch == '\n')
	  linenum ++;

	ptr = s;
	if (have_whitespace)
	{
          *ptr++ = ' ';
	  have_whitespace = 0;
	}

        *ptr++ = '<';
	if (ch == '=')
	  *ptr++ = '=';
	else if (ch == '<')
	  ungetc(ch, fp);
	else
	  have_whitespace = 1;

	*ptr++ = '\0';

	t->markup = MARKUP_NONE;
	t->data   = (uchar *)strdup((char *)s);
      }
      else
      {
       /*
        * Start of a markup...
	*/

	if (ch != '/')
          ungetc(ch, fp);

	if (parse_markup(t, fp, &linenum) == MARKUP_ERROR)
	{
#ifndef DEBUG
          progress_error(HD_ERROR_READ_ERROR,
                         "Unable to parse HTML element on line %d of %s!",
			 linenum, _htmlCurrentFile);
#endif /* !DEBUG */

          delete_node(t);
          break;
	}

       /*
	* Eliminate extra whitespace...
	*/

	if (issuper(t->markup) || isblock(t->markup) ||
            islist(t->markup) || islentry(t->markup) ||
            istable(t->markup) || istentry(t->markup) ||
	    t->markup == MARKUP_TITLE)
          have_whitespace = 0;

       /*
	* If this is the matching close mark, or if we are starting the same
	* markup, or if we've completed a list, we're done!
	*/

	if (ch == '/')
	{
	 /*
          * Close markup; find matching markup...
          */

          for (temp = parent; temp != NULL; temp = temp->parent)
            if (temp->markup == t->markup)
              break;
	    else if (temp->markup == MARKUP_EMBED)
	    {
	      temp = NULL;
              break;
	    }
	}
	else if (t->markup == MARKUP_BODY || t->markup == MARKUP_HEAD)
	{
	 /*
          * Make sure we aren't inside an existing HEAD or BODY...
	  */

          for (temp = parent; temp != NULL; temp = temp->parent)
            if (temp->markup == MARKUP_BODY || temp->markup == MARKUP_HEAD)
              break;
	    else if (temp->markup == MARKUP_EMBED)
	    {
	      temp = NULL;
	      break;
	    }
	}
	else if (t->markup == MARKUP_EMBED)
	{
	 /*
          * Close any text blocks...
	  */

          for (temp = parent; temp != NULL; temp = temp->parent)
            if (isblock(temp->markup) || islentry(temp->markup))
              break;
	    else if (istentry(temp->markup) || islist(temp->markup) ||
	             issuper(temp->markup) || temp->markup == MARKUP_EMBED)
	    {
	      temp = NULL;
	      break;
	    }
	}
	else if (issuper(t->markup))
	{
          for (temp = parent; temp != NULL; temp = temp->parent)
	    if (istentry(temp->markup) || temp->markup == MARKUP_EMBED)
	    {
	      temp = NULL;
              break;
	    }
	}
	else if (islist(t->markup))
	{
          for (temp = parent; temp != NULL; temp = temp->parent)
            if (isblock(temp->markup))
	      break;
	    else if (islentry(temp->markup) || istentry(temp->markup) ||
	             issuper(temp->markup) || temp->markup == MARKUP_EMBED)
	    {
	      temp = NULL;
              break;
	    }
	}
	else if (islentry(t->markup))
	{
          for (temp = parent; temp != NULL; temp = temp->parent)
            if (islentry(temp->markup))
              break;
	    else if (islist(temp->markup) || issuper(temp->markup) ||
	             istentry(temp->markup) || temp->markup == MARKUP_EMBED)
            {
	      temp = NULL;
	      break;
	    }
	}
	else if (isblock(t->markup))
	{
          for (temp = parent; temp != NULL; temp = temp->parent)
            if (isblock(temp->markup))
              break;
	    else if (istentry(temp->markup) || islist(temp->markup) ||
	             islentry(temp->markup) ||
	             issuper(temp->markup) || temp->markup == MARKUP_EMBED)
	    {
	      temp = NULL;
	      break;
	    }
	}
	else if (t->markup == MARKUP_THEAD || t->markup == MARKUP_TBODY || t->markup == MARKUP_TFOOT)
	{
          for (temp = parent; temp != NULL; temp = temp->parent)
	    if (temp->markup == MARKUP_TABLE || temp->markup == MARKUP_EMBED)
	    {
	      temp = NULL;
              break;
	    }
	}
	else if (istentry(t->markup))
	{
          for (temp = parent; temp != NULL; temp = temp->parent)
            if (istentry(temp->markup))
              break;
	    else if (temp->markup == MARKUP_TABLE || istable(temp->markup) ||
	             temp->markup == MARKUP_EMBED)
	    {
	      if (temp->markup != MARKUP_TR)
	      {
	        // Strictly speaking, this is an error - TD/TH can only
		// be found under TR, but web browsers automatically
		// inject a TR...
		progress_error(HD_ERROR_HTML_ERROR,
		               "No TR element before %s element on line %d of %s.",
			       _htmlMarkups[t->markup], linenum,
			       _htmlCurrentFile);

                parent = htmlAddTree(temp, MARKUP_TR, NULL);
		prev   = NULL;
		DEBUG_printf(("%str (inserted) under %s, line %d\n", indent,
		              _htmlMarkups[temp->markup], linenum));
	      }

	      temp = NULL;
              break;
	    }
	}
	else
          temp = NULL;

	if (temp != NULL)
	{
          DEBUG_printf(("%s>>>> Auto-ascend <<<\n", indent));

          if (ch != '/' &&
	      temp->markup != MARKUP_BODY &&
	      temp->markup != MARKUP_DD &&
	      temp->markup != MARKUP_DT &&
	      temp->markup != MARKUP_HEAD &&
	      temp->markup != MARKUP_HTML &&
	      temp->markup != MARKUP_LI &&
	      temp->markup != MARKUP_OPTION &&
	      temp->markup != MARKUP_P &&
	      temp->markup != MARKUP_TBODY &&
	      temp->markup != MARKUP_TD &&
	      temp->markup != MARKUP_TFOOT &&
	      temp->markup != MARKUP_TH &&
	      temp->markup != MARKUP_THEAD &&
	      temp->markup != MARKUP_TR)
	  {
	    // Log this condition as an error...
	    progress_error(HD_ERROR_HTML_ERROR,
	                   "No /%s element before %s element on line %d of %s.",
	                   _htmlMarkups[temp->markup],
			   _htmlMarkups[t->markup], linenum, _htmlCurrentFile);
	    DEBUG_printf(("%sNo /%s element before %s element on line %d.\n",
	                  indent, _htmlMarkups[temp->markup],
			  _htmlMarkups[t->markup], linenum));
	  }

#ifdef DEBUG
          for (tree_t *p = parent;
	       p && p != temp && indent[0];
	       p = p->parent)
	    indent[strlen((char *)indent) - 4] = '\0';

          if (indent[0])
            indent[strlen((char *)indent) - 4] = '\0';
#endif // DEBUG

          // Safety check; should never happen, since MARKUP_FILE is
	  // the root node created by the caller...
          if (temp->parent)
	  {
	    parent = temp->parent;
            prev   = parent->last_child;
	  }
	  else
	  {
	    for (prev = temp; prev->next; prev = prev->next);
	    parent = NULL;
	  }

          if (ch == '/')
	  {
	    // Closing element, so delete this node...
            delete_node(t);
	    continue;
	  }
	  else
	  {
	    // Reparent the node...
	    if (parent == NULL)
	    {
	      t->halignment   = ALIGN_LEFT;
	      t->valignment   = ALIGN_BOTTOM;
	      t->typeface     = _htmlBodyFont;
	      t->size         = SIZE_P;

	      compute_color(t, _htmlTextColor);
	    }
	    else
	    {
	      t->link          = parent->link;
	      t->halignment    = parent->halignment;
	      t->valignment    = parent->valignment;
	      t->typeface      = parent->typeface;
	      t->size          = parent->size;
	      t->style         = parent->style;
	      t->superscript   = parent->superscript;
	      t->subscript     = parent->subscript;
	      t->preformatted  = parent->preformatted;
	      t->indent        = parent->indent;
	      t->red           = parent->red;
	      t->green         = parent->green;
	      t->blue          = parent->blue;
	      t->underline     = parent->underline;
	      t->strikethrough = parent->strikethrough;
	    }

          }
	}
	else if (ch == '/')
	{
	  // Log this condition as an error...
	  if (t->markup != MARKUP_UNKNOWN &&
	      t->markup != MARKUP_COMMENT)
	  {
	    progress_error(HD_ERROR_HTML_ERROR,
	                   "Dangling /%s element on line %d of %s.",
			   _htmlMarkups[t->markup], linenum, _htmlCurrentFile);
	    DEBUG_printf(("%sDangling /%s element on line %d.\n",
			  indent, _htmlMarkups[t->markup], linenum));
          }

	  delete_node(t);
	  continue;
	}
      }
    }
    else if (t->preformatted)
    {
     /*
      * Read a pre-formatted string into the current tree entry...
      */

      ptr = s;
      while (ch != '<' && ch != EOF && ptr < (s + sizeof(s) - 1))
      {
        if (ch == '&')
        {
	  // Possibly a character entity...
	  eptr = entity;
	  while (eptr < (entity + sizeof(entity) - 1) &&
	         (ch = getc(fp)) != EOF)
	    if (!isalnum(ch) && ch != '#')
	      break;
	    else
	      *eptr++ = (uchar)ch;

          if (ch != ';')
	  {
	    ungetc(ch, fp);
	    ch = 0;
	  }

          *eptr = '\0';
          if (!ch)
	  {
	    progress_error(HD_ERROR_HTML_ERROR, "Unquoted & on line %d of %s.",
	                   linenum, _htmlCurrentFile);

            if (ptr < (s + sizeof(s) - 1))
	      *ptr++ = '&';
            strlcpy((char *)ptr, (char *)entity, sizeof(s) - (size_t)(ptr - s));
	    ptr += strlen((char *)ptr);
	  }
	  else if ((ch = iso8859(entity)) == 0)
	  {
	    progress_error(HD_ERROR_HTML_ERROR,
	                   "Unknown character entity \"&%s;\" on line %d of %s.",
	                   entity, linenum, _htmlCurrentFile);

            if (ptr < (s + sizeof(s) - 1))
	      *ptr++ = '&';
            strlcpy((char *)ptr, (char *)entity, sizeof(s) - (size_t)(ptr - s));
	    ptr += strlen((char *)ptr);
            if (ptr < (s + sizeof(s) - 1))
	      *ptr++ = ';';
	  }
	  else
	    *ptr++ = (uchar)ch;
        }
        else if ((ch & 0x80) && _htmlUTF8)
        {
          // Collect UTF-8 value...
          ch = utf8_getc(ch, fp);

          if (ch)
            *ptr++ = (uchar)ch;
        }
	else if (ch != 0 && ch != '\r')
	{
          *ptr++ = (uchar)ch;

          if (ch == '\n')
	  {
	    linenum ++;
            break;
	  }
	}

        ch = getc(fp);
      }

      *ptr = '\0';

      if (ch == '<')
        ungetc(ch, fp);

      t->markup = MARKUP_NONE;
      t->data   = (uchar *)strdup((char *)s);

      DEBUG_printf(("%sfragment \"%s\", line %d\n", indent, s, linenum));
    }
    else
    {
     /*
      * Read the next string fragment...
      */

      ptr = s;
      if (have_whitespace)
      {
        *ptr++ = ' ';
	have_whitespace = 0;
      }

      while (!isspace(ch) && ch != '<' && ch != EOF && ptr < (s + sizeof(s) - 1))
      {
        if (ch == '&')
        {
	  // Possibly a character entity...
	  eptr = entity;
	  while (eptr < (entity + sizeof(entity) - 1) &&
	         (ch = getc(fp)) != EOF)
	    if (!isalnum(ch) && ch != '#')
	      break;
	    else
	      *eptr++ = (uchar)ch;

          *eptr = '\0';

          if (ch != ';')
	  {
	    ungetc(ch, fp);
	    ch = 0;
	  }

          if (!ch)
	  {
	    progress_error(HD_ERROR_HTML_ERROR, "Unquoted & on line %d of %s.",
	                   linenum, _htmlCurrentFile);

            if (ptr < (s + sizeof(s) - 1))
	      *ptr++ = '&';
            strlcpy((char *)ptr, (char *)entity, sizeof(s) - (size_t)(ptr - s));
	    ptr += strlen((char *)ptr);
	  }
	  else if ((ch = iso8859(entity)) == 0)
	  {
	    progress_error(HD_ERROR_HTML_ERROR,
	                   "Unknown character entity \"&%s;\" on line %d of %s.",
	                   entity, linenum, _htmlCurrentFile);

            if (ptr < (s + sizeof(s) - 1))
	      *ptr++ = '&';
            strlcpy((char *)ptr, (char *)entity, sizeof(s) - (size_t)(ptr - s));
	    ptr += strlen((char *)ptr);
            if (ptr < (s + sizeof(s) - 1))
	      *ptr++ = ';';
	  }
	  else
	    *ptr++ = (uchar)ch;
        }
        else
        {
          if ((ch & 0x80) && _htmlUTF8)
          {
            // Collect UTF-8 value...
            ch = utf8_getc(ch, fp);
          }

          if (ch)
            *ptr++ = (uchar)ch;
        }

	if ((_htmlUTF8 && ch == _htmlCharacters[173]) || (!_htmlUTF8 && ch == 173))
	  break;

        ch = getc(fp);
      }

      if (ch == '\n')
	linenum ++;

      if (isspace(ch))
        have_whitespace = 1;

      *ptr = '\0';

      if (ch == '<')
        ungetc(ch, fp);

      t->markup = MARKUP_NONE;
      t->data   = (uchar *)strdup((char *)s);

      DEBUG_printf(("%sfragment \"%s\" (len=%d), line %d\n", indent, s,
                    (int)(ptr - s), linenum));
    }

   /*
    * If the parent tree pointer is not null and this is the first
    * entry we've read, set the child pointer...
    */

    DEBUG_printf(("%sADDING %s node to %s parent!\n", indent,
                  _htmlMarkups[t->markup],
		  parent ? _htmlMarkups[parent->markup] : "ROOT"));
    if (parent != NULL && prev == NULL)
      parent->child = t;

    if (parent != NULL)
      parent->last_child = t;

   /*
    * Do the prev/next links...
    */

    t->parent = parent;
    t->prev   = prev;
    if (prev != NULL)
      prev->next = t;

    if (tree == NULL)
      tree = t;

    prev = t;

   /*
    * Do markup-specific stuff...
    */

    descend = 0;

    switch (t->markup)
    {
      case MARKUP_BODY :
         /*
	  * Update the text color as necessary...
	  */

          if ((color = htmlGetVariable(t, (uchar *)"TEXT")) != NULL)
            compute_color(t, color);
	  else
            compute_color(t, _htmlTextColor);

          if ((color = htmlGetVariable(t, (uchar *)"BGCOLOR")) != NULL &&
	      !BodyColor[0])
	    strlcpy(BodyColor, (char *)color, sizeof(BodyColor));

          // Update the background image as necessary...
          if ((filename = htmlGetVariable(t, (uchar *)"BACKGROUND")) != NULL)
	    htmlSetVariable(t, (uchar *)"BACKGROUND",
	                    (uchar *)fix_filename((char *)filename,
			                          (char *)base));

          descend = 1;
          break;

      case MARKUP_IMG :
          if (have_whitespace)
	  {
	    // Insert a space before this image...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          // Get the image alignment...
          t->valignment = ALIGN_BOTTOM;
          get_alignment(t);

          // Update the image source as necessary...
          if ((filename = htmlGetVariable(t, (uchar *)"SRC")) != NULL)
	    htmlSetVariable(t, (uchar *)"REALSRC",
	                    (uchar *)fix_filename((char *)filename,
			                          (char *)base));

      case MARKUP_BR :
      case MARKUP_NONE :
      case MARKUP_SPACER :
	 /*
	  * Figure out the width & height of this markup...
	  */

          compute_size(t);
	  break;

      case MARKUP_H1 :
      case MARKUP_H2 :
      case MARKUP_H3 :
      case MARKUP_H4 :
      case MARKUP_H5 :
      case MARKUP_H6 :
      case MARKUP_H7 :
      case MARKUP_H8 :
      case MARKUP_H9 :
      case MARKUP_H10 :
      case MARKUP_H11 :
      case MARKUP_H12 :
      case MARKUP_H13 :
      case MARKUP_H14 :
      case MARKUP_H15 :
          get_alignment(t);

          t->typeface      = _htmlHeadingFont;
          t->subscript     = 0;
          t->superscript   = 0;
          t->strikethrough = 0;
          t->preformatted  = 0;

	  if (t->markup > MARKUP_H6)
          {
	    t->size  = SIZE_H7;
            t->style = STYLE_ITALIC;
	  }
	  else
	  {
            t->size  = (unsigned)(SIZE_H1 - t->markup + MARKUP_H1);
            t->style = STYLE_BOLD;
	  }

          descend = 1;
          break;

      case MARKUP_P :
          get_alignment(t);

          t->typeface      = _htmlBodyFont;
          t->size          = SIZE_P;
          t->style         = STYLE_NORMAL;
          t->subscript     = 0;
          t->superscript   = 0;
          t->strikethrough = 0;
          t->preformatted  = 0;

          descend = 1;
          break;

      case MARKUP_PRE :
          t->typeface      = _htmlBodyFont >= TYPE_MONOSPACE ? TYPE_MONOSPACE
	                                                     : TYPE_COURIER;
          t->size          = SIZE_PRE;
          t->style         = STYLE_NORMAL;
          t->subscript     = 0;
          t->superscript   = 0;
          t->strikethrough = 0;
          t->preformatted  = 1;

          descend = 1;
          break;

      case MARKUP_BLOCKQUOTE :
      case MARKUP_DIR :
      case MARKUP_MENU :
      case MARKUP_UL :
      case MARKUP_OL :
      case MARKUP_DL :
          t->indent ++;

          descend = 1;
          break;

      case MARKUP_DIV :
          get_alignment(t);

          descend = 1;
          break;

      case MARKUP_HR :
          t->halignment = ALIGN_CENTER;
          get_alignment(t);
          break;

      case MARKUP_DOCTYPE :
      case MARKUP_AREA :
      case MARKUP_COMMENT :
      case MARKUP_INPUT :
      case MARKUP_ISINDEX :
      case MARKUP_LINK :
      case MARKUP_META :
      case MARKUP_WBR :
      case MARKUP_COL :
          break;

      case MARKUP_EMBED :
          if ((type = htmlGetVariable(t, (uchar *)"TYPE")) != NULL &&
	      strncasecmp((const char *)type, "text/html", 9) != 0)
	    break;

          if ((filename = htmlGetVariable(t, (uchar *)"SRC")) != NULL)
	  {
	    const char *save_name = _htmlCurrentFile;

	    filename = (uchar *)fix_filename((char *)filename,
	                                     (char *)base);

            if ((embed = fopen((char *)filename, "r")) != NULL)
            {
	      strlcpy(newbase, file_directory((char *)filename), sizeof(newbase));

              _htmlCurrentFile = (char *)filename;
              htmlReadFile(t, embed, newbase);
              fclose(embed);
	      _htmlCurrentFile = save_name;
            }
#ifndef DEBUG
	    else
	      progress_error(HD_ERROR_FILE_NOT_FOUND,
                             "Unable to embed \"%s\" - %s", filename,
	                     strerror(errno));
#endif /* !DEBUG */
	  }
          break;

      case MARKUP_TH :
          if (htmlGetVariable(t->parent, (uchar *)"ALIGN") != NULL)
	    t->halignment = t->parent->halignment;
	  else
            t->halignment = ALIGN_CENTER;

          if (htmlGetVariable(t->parent, (uchar *)"VALIGN") != NULL)
	    t->valignment = t->parent->valignment;
	  else
            t->valignment = ALIGN_MIDDLE;

          get_alignment(t);

          t->style = STYLE_BOLD;

          descend = 1;
          break;

      case MARKUP_TD :
          if (htmlGetVariable(t->parent, (uchar *)"ALIGN") != NULL)
	    t->halignment = t->parent->halignment;
	  else
            t->halignment = ALIGN_LEFT;

          if (htmlGetVariable(t->parent, (uchar *)"VALIGN") != NULL)
	    t->valignment = t->parent->valignment;
	  else
            t->valignment = ALIGN_MIDDLE;

	  get_alignment(t);

          t->style = STYLE_NORMAL;

          descend = 1;
          break;

      case MARKUP_SPAN :
          // Pull style data, if present...
          if (have_whitespace)
	  {
	    // Insert a space before this element...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          get_alignment(t);

          if ((color = htmlGetStyle(t, (uchar *)"color:")) != NULL)
            compute_color(t, color);

          if ((face = htmlGetStyle(t, (uchar *)"font-family:")) != NULL)
          {
	    char	font[255],	// Font name
			*fontptr;	// Pointer into font name

            for (ptr = face; *ptr;)
	    {
	      while (isspace(*ptr) || *ptr == ',')
	        ptr ++;

              if (!*ptr)
	        break;

	      for (fontptr = font; *ptr && *ptr != ',' && !isspace(*ptr); ptr ++)
	        if (fontptr < (font + sizeof(font) - 1))
		  *fontptr++ = (char)*ptr;

              *fontptr = '\0';

              if (!strcasecmp(font, "serif"))
	      {
        	t->typeface = TYPE_SERIF;
		break;
	      }
              else if (!strcasecmp(font, "sans-serif") ||
	               !strcasecmp(font, "sans"))
	      {
        	t->typeface = TYPE_SANS_SERIF;
		break;
	      }
              else if (!strcasecmp(font, "monospace"))
	      {
        	t->typeface = TYPE_MONOSPACE;
		break;
	      }
              else if (!strcasecmp(font, "arial") ||
	               !strcasecmp(font, "helvetica"))
              {
        	t->typeface = TYPE_HELVETICA;
		break;
	      }
              else if (!strcasecmp(font, "times"))
	      {
        	t->typeface = TYPE_TIMES;
		break;
	      }
              else if (!strcasecmp(font, "courier"))
	      {
        	t->typeface = TYPE_COURIER;
		break;
	      }
	      else if (!strcasecmp(font, "symbol"))
	      {
        	t->typeface = TYPE_SYMBOL;
		break;
	      }
	      else if (!strcasecmp(font, "dingbat"))
	      {
        	t->typeface = TYPE_DINGBATS;
		break;
	      }
	    }
          }

          if ((size = htmlGetStyle(t, (uchar *)"font-size:")) != NULL)
          {
            // Find the closest size to the fixed sizes...
            unsigned i;
            double fontsize = atof((char *)size);

            for (i = 0; i < 7; i ++)
              if (fontsize <= _htmlSizes[i])
                break;

	    t->size = i;
          }

          if ((span = htmlGetStyle(t, (uchar *)"font-style:")) != NULL)
          {
            if (!strcmp((char *)span, "normal"))
              t->style &= ~STYLE_ITALIC;
            else if (!strcmp((char *)span, "italic") || !strcmp((char *)span, "oblique"))
              t->style |= STYLE_ITALIC;
          }

          if ((span = htmlGetStyle(t, (uchar *)"font-weight:")) != NULL)
          {
            if (!strcmp((char *)span, "bold") || !strcmp((char *)span, "bolder") || !strcmp((char *)span, "700") || !strcmp((char *)span, "800") || !strcmp((char *)span, "900"))
              t->style |= STYLE_BOLD;
            else if (strcmp((char *)span, "inherit"))
              t->style &= ~STYLE_BOLD;
          }

          if ((span = htmlGetStyle(t, (uchar *)"text-decoration:")) != NULL)
          {
            if (!strcmp((char *)span, "underline"))
              t->underline = 1;
            else if (!strcmp((char *)span, "line-through"))
              t->strikethrough = 1;
            else if (strcmp((char *)span, "inherit"))
              t->underline = t->strikethrough = 0;
          }

          if ((span = htmlGetStyle(t, (uchar *)"vertical-align:")) != NULL)
          {
            if (!strcmp((char *)span, "sub"))
              t->subscript = 1;
            else if (!strcmp((char *)span, "super"))
              t->superscript = 1;
            else if (strcmp((char *)span, "inherit"))
              t->subscript = t->superscript = 0;
          }

          descend = 1;
          break;

      case MARKUP_FONT :
          if (have_whitespace)
	  {
	    // Insert a space before this element...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          if ((face = htmlGetVariable(t, (uchar *)"FACE")) != NULL)
          {
	    char	font[255],	// Font name
			*fontptr;	// Pointer into font name


            for (ptr = face; *ptr;)
	    {
	      while (isspace(*ptr) || *ptr == ',')
	        ptr ++;

              if (!*ptr)
	        break;

	      for (fontptr = font; *ptr && *ptr != ',' && !isspace(*ptr); ptr ++)
	        if (fontptr < (font + sizeof(font) - 1))
		  *fontptr++ = (char)*ptr;

              *fontptr = '\0';

              if (!strcasecmp(font, "serif"))
	      {
        	t->typeface = TYPE_SERIF;
		break;
	      }
              else if (!strcasecmp(font, "sans-serif") ||
	               !strcasecmp(font, "sans"))
	      {
        	t->typeface = TYPE_SANS_SERIF;
		break;
	      }
              else if (!strcasecmp(font, "mono") || !strcasecmp(font, "monospace"))
	      {
        	t->typeface = TYPE_MONOSPACE;
		break;
	      }
              else if (!strcasecmp(font, "arial") ||
	               !strcasecmp(font, "helvetica"))
              {
        	t->typeface = TYPE_HELVETICA;
		break;
	      }
              else if (!strcasecmp(font, "times"))
	      {
        	t->typeface = TYPE_TIMES;
		break;
	      }
              else if (!strcasecmp(font, "courier"))
	      {
        	t->typeface = TYPE_COURIER;
		break;
	      }
	      else if (!strcasecmp(font, "symbol"))
	      {
        	t->typeface = TYPE_SYMBOL;
		break;
	      }
	      else if (!strcasecmp(font, "dingbat"))
	      {
        	t->typeface = TYPE_DINGBATS;
		break;
	      }
	    }
          }

          if ((color = htmlGetVariable(t, (uchar *)"COLOR")) != NULL)
            compute_color(t, color);

          if ((size = htmlGetVariable(t, (uchar *)"SIZE")) != NULL)
          {
            if (have_whitespace)
	    {
	      // Insert a space before sized text...
	      insert_space(parent, t);

	      have_whitespace = 0;
	    }

	    if (isdigit(size[0]))
	      sizeval = atoi((char *)size);
	    else
              sizeval = t->size + atoi((char *)size);

            if (sizeval < 0)
              t->size = 0;
            else if (sizeval > 7)
              t->size = 7;
            else
              t->size = (unsigned)sizeval;
          }

          descend = 1;
          break;

      case MARKUP_BIG :
          if (have_whitespace)
	  {
	    // Insert a space before big text...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          if (t->size < 6)
            t->size += 2;
          else
            t->size = 7;

          descend = 1;
          break;

      case MARKUP_SMALL :
          if (have_whitespace)
	  {
	    // Insert a space before small text...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          if (t->size > 2)
            t->size -= 2;
          else
            t->size = 0;

          descend = 1;
          break;

      case MARKUP_SUP :
          if (have_whitespace)
	  {
	    // Insert a space before superscript text...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          t->superscript = 1;

          if ((sizeval = t->size + SIZE_SUP) < 0)
	    t->size = 0;
	  else
	    t->size = (unsigned)sizeval;

          descend = 1;
          break;

      case MARKUP_SUB :
          if (have_whitespace)
	  {
	    // Insert a space before subscript text...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          t->subscript = 1;

          if ((sizeval = t->size + SIZE_SUB) < 0)
	    t->size = 0;
	  else
	    t->size = (unsigned)sizeval;

          descend = 1;
          break;

      case MARKUP_KBD :
          t->style    = STYLE_BOLD;

      case MARKUP_TT :
      case MARKUP_CODE :
      case MARKUP_SAMP :
          if (isspace(ch = getc(fp)))
	    have_whitespace = 1;
	  else
	    ungetc(ch, fp);

          if (have_whitespace)
	  {
	    // Insert a space before monospaced text...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          t->typeface = _htmlBodyFont >= TYPE_MONOSPACE ? TYPE_MONOSPACE
	                                                : TYPE_COURIER;

          descend = 1;
          break;

      case MARKUP_STRONG :
      case MARKUP_B :
          t->style = (style_t)(t->style | STYLE_BOLD);

          descend = 1;
          break;

      case MARKUP_DD :
          t->indent ++;

          descend = 1;
          break;

      case MARKUP_VAR :
          t->style = (style_t)(t->style | STYLE_ITALIC);
      case MARKUP_DFN :
          t->typeface = _htmlBodyFont >= TYPE_MONOSPACE ? TYPE_SANS_SERIF
	                                                : TYPE_HELVETICA;

          descend = 1;
          break;

      case MARKUP_CITE :
      case MARKUP_EM :
      case MARKUP_I :
          t->style = (style_t)(t->style | STYLE_ITALIC);

          descend = 1;
          break;

      case MARKUP_U :
      case MARKUP_INS :
          if (have_whitespace)
	  {
	    // Insert a space before underlined text...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          t->underline = 1;

          descend = 1;
          break;

      case MARKUP_STRIKE :
      case MARKUP_S :
      case MARKUP_DEL :
          if (have_whitespace)
	  {
	    // Insert a space before struck-through text...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          t->strikethrough = 1;

          descend = 1;
          break;

      case MARKUP_CENTER :
          t->halignment = ALIGN_CENTER;

          descend = 1;
          break;

      case MARKUP_A :
          if (have_whitespace)
	  {
	    // Insert a space before this link...
	    insert_space(parent, t);

	    have_whitespace = 0;
	  }

          descend = 1;
	  break;

      default :
         /*
          * All other markup types should be using <MARK>...</MARK>
          */

          get_alignment(t);

          descend = 1;
          break;
    }

    if (descend)
    {
#ifdef DEBUG
      strlcat((char *)indent, "    ", sizeof(indent));
#endif // DEBUG

      parent = t;
      prev   = NULL;
    }
  }

  return (tree);
}


/*
 * 'write_file()' - Write a tree entry to a file...
 */

static int			/* I - New column */
write_file(tree_t *t,		/* I - Tree entry */
           FILE   *fp,		/* I - File to write to */
           int    col)		/* I - Current column */
{
  int	i;			/* Looping var */
  uchar	*ptr;			/* Character pointer */


  while (t != NULL)
  {
    if (t->markup == MARKUP_NONE)
    {
      if (t->preformatted)
      {
        for (ptr = t->data; *ptr != '\0'; ptr ++)
          fputs((char *)iso8859(*ptr), fp);

	if (t->data[0] && t->data[strlen((char *)t->data) - 1] == '\n')
          col = 0;
	else
          col += strlen((char *)t->data);
      }
      else
      {
	if ((col + (int)strlen((char *)t->data)) > 72 && col > 0)
	{
          putc('\n', fp);
          col = 0;
	}

        for (ptr = t->data; *ptr != '\0'; ptr ++)
          fputs((char *)iso8859(*ptr), fp);

	col += strlen((char *)t->data);

	if (col > 72)
	{
          putc('\n', fp);
          col = 0;
	}
      }
    }
    else if (t->markup == MARKUP_COMMENT)
      fprintf(fp, "\n<!--%s-->\n", t->data);
    else if (t->markup > 0)
    {
      switch (t->markup)
      {
        case MARKUP_AREA :
        case MARKUP_BR :
        case MARKUP_CENTER :
        case MARKUP_COMMENT :
        case MARKUP_DD :
        case MARKUP_DL :
        case MARKUP_DT :
        case MARKUP_H1 :
        case MARKUP_H2 :
        case MARKUP_H3 :
        case MARKUP_H4 :
        case MARKUP_H5 :
        case MARKUP_H6 :
        case MARKUP_HEAD :
        case MARKUP_HR :
        case MARKUP_LI :
        case MARKUP_MAP :
        case MARKUP_OL :
        case MARKUP_P :
        case MARKUP_PRE :
        case MARKUP_TABLE :
        case MARKUP_TITLE :
        case MARKUP_TR :
        case MARKUP_UL :
	case MARKUP_DIR :
	case MARKUP_MENU :
            if (col > 0)
            {
              putc('\n', fp);
              col = 0;
            }
        default :
            break;
      }

      col += fprintf(fp, "<%s", _htmlMarkups[t->markup]);
      for (i = 0; i < t->nvars; i ++)
      {
	if (col > 72 && !t->preformatted)
	{
          putc('\n', fp);
          col = 0;
	}

        if (col > 0)
        {
          putc(' ', fp);
          col ++;
        }

	if (t->vars[i].value == NULL)
          col += fprintf(fp, "%s", t->vars[i].name);
	else if (strchr((char *)t->vars[i].value, '\"') != NULL)
          col += fprintf(fp, "%s=\'%s\'", t->vars[i].name, t->vars[i].value);
	else
          col += fprintf(fp, "%s=\"%s\"", t->vars[i].name, t->vars[i].value);
      }

      putc('>', fp);
      col ++;

      if (col > 72 && !t->preformatted)
      {
	putc('\n', fp);
	col = 0;
      }

      if (t->child != NULL)
      {
	col = write_file(t->child, fp, col);

	if (col > 72 && !t->preformatted)
	{
	  putc('\n', fp);
	  col = 0;
	}

        col += fprintf(fp, "</%s>", _htmlMarkups[t->markup]);
        switch (t->markup)
        {
          case MARKUP_AREA :
          case MARKUP_BR :
          case MARKUP_CENTER :
          case MARKUP_COMMENT :
          case MARKUP_DD :
          case MARKUP_DL :
          case MARKUP_DT :
          case MARKUP_H1 :
          case MARKUP_H2 :
          case MARKUP_H3 :
          case MARKUP_H4 :
          case MARKUP_H5 :
          case MARKUP_H6 :
          case MARKUP_HEAD :
          case MARKUP_HR :
          case MARKUP_LI :
          case MARKUP_MAP :
          case MARKUP_OL :
          case MARKUP_P :
          case MARKUP_PRE :
          case MARKUP_TABLE :
          case MARKUP_TITLE :
          case MARKUP_TR :
          case MARKUP_UL :
          case MARKUP_DIR :
          case MARKUP_MENU :
              putc('\n', fp);
              col = 0;
          default :
	      break;
        }
      }
    }

    t = t->next;
  }

  return (col);
}


/*
 * 'htmlWriteFile()' - Write an HTML markup tree to a file.
 */

int				/* O - Write status: 0 = success, -1 = fail */
htmlWriteFile(tree_t *parent,	/* I - Parent tree entry */
              FILE   *fp)	/* I - File to write to */
{
  if (write_file(parent, fp, 0) < 0)
    return (-1);
  else
    return (0);
}


/*
 * 'htmlAddTree()' - Add a tree node to the parent.
 */

tree_t *			/* O - New entry */
htmlAddTree(tree_t   *parent,	/* I - Parent entry */
            markup_t markup,	/* I - Markup code */
            uchar    *data)	/* I - Data/text */
{
  tree_t	*t;		/* New tree entry */


  if ((t = htmlNewTree(parent, markup, data)) == NULL)
    return (NULL);

 /*
  * Add the tree entry to the end of the chain of children...
  */

  if (parent != NULL)
  {
    if (parent->last_child != NULL)
    {
      parent->last_child->next = t;
      t->prev                  = parent->last_child;
    }
    else
      parent->child = t;

    parent->last_child = t;
  }

  return (t);
}


/*
 * 'htmlDeleteTree()' - Free all memory associated with a tree...
 */

int				/* O - 0 for success, -1 for failure */
htmlDeleteTree(tree_t *parent)	/* I - Parent to delete */
{
  tree_t	*next;		/* Next tree entry */


  if (parent == NULL)
    return (-1);

  while (parent != NULL)
  {
    next = parent->next;

    if (parent->child != NULL)
      if (htmlDeleteTree(parent->child))
        return (-1);

    delete_node(parent);

    parent = next;
  }

  return (0);
}


/*
 * 'htmlInsertTree()' - Insert a tree node under the parent.
 */

tree_t *			/* O - New entry */
htmlInsertTree(tree_t   *parent,/* I - Parent entry */
               markup_t markup,	/* I - Markup code */
               uchar    *data)	/* I - Data/text */
{
  tree_t	*t;		/* New tree entry */


  if ((t = htmlNewTree(parent, markup, data)) == NULL)
    return (NULL);

 /*
  * Insert the tree entry at the beginning of the chain of children...
  */

  if (parent != NULL)
  {
    if (parent->child != NULL)
    {
      parent->child->prev = t;
      t->next             = parent->child;
    }
    else
      parent->last_child = t;

    parent->child = t;
  }

  return (t);
}


/*
 * 'htmlMapUnicode()' - Map a Unicode character to the custom character set.
 */

uchar				/* O - Charset character */
htmlMapUnicode(int ch)		/* I - Unicode character */
{
  uchar	newch;			/* New charset character */


  // If we already have a mapping for this character, return it...
  if (_htmlCharacters[ch])
    return (_htmlCharacters[ch]);

  if (_htmlUTF8 >= 0x100)
  {
    progress_error(HD_ERROR_READ_ERROR, "Too many Unicode code points.");
    return (0);
  }

  newch = _htmlUTF8++;

  _htmlCharacters[ch] = (uchar)newch;
  _htmlUnicode[newch] = ch;
  _htmlGlyphs[newch]  = _htmlGlyphsAll[ch];

  for (int i = 0; i < TYPE_MAX; i ++)
    for (int j = 0; j < STYLE_MAX; j ++)
      _htmlWidths[i][j][newch] = _htmlWidthsAll[i][j][ch];

  return (newch);
}


/*
 * 'htmlNewTree()' - Create a new tree node for the parent.
 */

tree_t *			/* O - New entry */
htmlNewTree(tree_t   *parent,	/* I - Parent entry */
            markup_t markup,	/* I - Markup code */
            uchar    *data)	/* I - Data/text */
{
  tree_t	*t;		/* New tree entry */


 /*
  * Allocate a new tree entry - use calloc() to get zeroed data...
  */

  t = (tree_t *)calloc(sizeof(tree_t), 1);
  if (t == NULL)
    return (NULL);

 /*
  * Set the markup code and copy the data if necessary...
  */

  t->markup = markup;
  if (data != NULL)
    t->data = (uchar *)strdup((char *)data);

 /*
  * Set/copy font characteristics...
  */

  if (parent == NULL)
  {
    t->halignment = ALIGN_LEFT;
    t->valignment = ALIGN_BOTTOM;
    t->typeface   = _htmlBodyFont;
    t->size       = SIZE_P;

    compute_color(t, _htmlTextColor);
  }
  else
  {
    t->link          = parent->link;
    t->halignment    = parent->halignment;
    t->valignment    = parent->valignment;
    t->typeface      = parent->typeface;
    t->size          = parent->size;
    t->style         = parent->style;
    t->preformatted  = parent->preformatted;
    t->indent        = parent->indent;
    t->red           = parent->red;
    t->green         = parent->green;
    t->blue          = parent->blue;
    t->underline     = parent->underline;
    t->strikethrough = parent->strikethrough;
  }

  switch (t->markup)
  {
    case MARKUP_NONE :
    case MARKUP_IMG :
       /*
	* Figure out the width & height of this fragment...
	*/

        compute_size(t);
	break;

    case MARKUP_H1 :
    case MARKUP_H2 :
    case MARKUP_H3 :
    case MARKUP_H4 :
    case MARKUP_H5 :
    case MARKUP_H6 :
        get_alignment(t);

        t->typeface      = _htmlHeadingFont;
        t->size          = (unsigned)(SIZE_H1 - t->markup + MARKUP_H1);
        t->subscript     = 0;
        t->superscript   = 0;
        t->strikethrough = 0;
        t->preformatted  = 0;
        t->style         = STYLE_BOLD;
        break;

    case MARKUP_P :
        get_alignment(t);

        t->typeface      = _htmlBodyFont;
        t->size          = SIZE_P;
        t->style         = STYLE_NORMAL;
        t->subscript     = 0;
        t->superscript   = 0;
        t->strikethrough = 0;
        t->preformatted  = 0;
        break;

    case MARKUP_PRE :
        t->typeface      = _htmlBodyFont >= TYPE_MONOSPACE ? TYPE_MONOSPACE
	                                                   : TYPE_COURIER;
        t->size          = SIZE_PRE;
        t->style         = STYLE_NORMAL;
        t->subscript     = 0;
        t->superscript   = 0;
        t->strikethrough = 0;
        t->preformatted  = 1;
        break;

    case MARKUP_DIV :
        get_alignment(t);
        break;

    case MARKUP_BLOCKQUOTE :
        t->style = STYLE_ITALIC;

    case MARKUP_UL :
    case MARKUP_DIR :
    case MARKUP_MENU :
    case MARKUP_OL :
    case MARKUP_DL :
        t->indent ++;
        break;

    case MARKUP_AREA :
    case MARKUP_BR :
    case MARKUP_COMMENT :
    case MARKUP_HR :
    case MARKUP_INPUT :
    case MARKUP_ISINDEX :
    case MARKUP_META :
    case MARKUP_WBR :
        break;

    case MARKUP_TH :
        t->style = STYLE_BOLD;
    case MARKUP_TD :
        get_alignment(t);
        break;

    case MARKUP_SUP :
        t->superscript = 1;
        t->size        = SIZE_P + SIZE_SUP;
        break;

    case MARKUP_SUB :
        t->subscript = 1;
        t->size      = SIZE_P + SIZE_SUB;
        break;

    case MARKUP_B :
        t->style = (style_t)(t->style | STYLE_BOLD);
        break;

    case MARKUP_DD :
        t->indent ++;
        break;

    case MARKUP_DT :
    case MARKUP_I :
        t->style = (style_t)(t->style | STYLE_ITALIC);
        break;

    case MARKUP_U :
    case MARKUP_INS :
        t->underline = 1;
        break;

    case MARKUP_STRIKE :
    case MARKUP_DEL :
        t->strikethrough = 1;
        break;

    default :
        break;
  }

  t->parent = parent;

  return (t);
}


/*
 * 'htmlGetText()' - Get all text from the given tree.
 */

uchar *				/* O - String containing text nodes */
htmlGetText(tree_t *t)		/* I - Tree to pick */
{
  uchar		*s,		// String
		*s2,		// New string
		*tdata = NULL,	// Temporary string data
		*talloc = NULL;	// Allocated string data
  size_t	slen,		// Length of string
		tlen;		// Length of node string


  // Loop through all of the nodes in the tree and collect text...
  slen = 0;
  s    = NULL;

  while (t != NULL)
  {
    if (t->child)
      tdata = talloc = htmlGetText(t->child);
    else
      tdata = t->data;

    tlen = tdata ? strlen((char *)tdata) : 0;

    if (tdata != NULL && tlen > 0)
    {
      // Add the text to this string...
      if (s)
        s2 = (uchar *)realloc(s, 1 + slen + tlen);
      else
        s2 = (uchar *)malloc(1 + tlen);

      if (!s2)
        break;

      s = s2;

      memcpy((char *)s + slen, (char *)tdata, tlen);

      slen += tlen;
    }

    if (talloc)
    {
      free(talloc);
      talloc = NULL;
    }

    t = t->next;
  }

  if (slen)
    s[slen] = '\0';

  if (talloc)
    free(talloc);

  return (s);
}


/*
 * 'htmlGetMeta()' - Get document "meta" data...
 */

uchar *				/* O - Content string */
htmlGetMeta(tree_t *tree,	/* I - Document tree */
            uchar  *name)	/* I - Metadata name */
{
  uchar	*tname,			/* Name value from tree entry */
	*tcontent;		/* Content value from tree entry */


  while (tree != NULL)
  {
   /*
    * Check this tree entry...
    */

    if (tree->markup == MARKUP_META &&
        (tname = htmlGetVariable(tree, (uchar *)"NAME")) != NULL &&
        (tcontent = htmlGetVariable(tree, (uchar *)"CONTENT")) != NULL)
    {
      if (strcasecmp((char *)name, (char *)tname) == 0)
        return (tcontent);
    }
    else if (tree->markup == MARKUP_HTML && !strcasecmp((char *)name, "LANG") && (tcontent = htmlGetVariable(tree, (uchar *)"LANG")) != NULL)
    {
      return (tcontent);
    }

   /*
    * Check child entries...
    */

    if (tree->child != NULL)
      if ((tcontent = htmlGetMeta(tree->child, name)) != NULL)
        return (tcontent);

   /*
    * Next tree entry...
    */

    tree = tree->next;
  }

  return (NULL);
}


/*
 * 'htmlGetStyle()' - Get a style value from a node's STYLE attribute.
 */

uchar *				// O - Value or NULL
htmlGetStyle(tree_t *t,		// I - Node
             uchar  *name)	// I - Name (including ":")
{
  uchar		*ptr,		// Pointer in STYLE attribute
		*bufptr;	// Pointer in buffer
  size_t	ptrlen,		// Length of STYLE attribute
		namelen;	// Length of name
  static uchar	buffer[1024];	// Buffer for value


  // See if we have a STYLE attribute...
  if ((ptr = htmlGetVariable(t, (uchar *)"STYLE")) == NULL)
    return (NULL);

  // Loop through the STYLE attribute looking for the name...
  for (namelen = strlen((char *)name), ptrlen = strlen((char *)ptr);
       ptrlen > namelen;
       ptr ++, ptrlen --)
    if (strncasecmp((char *)name, (char *)ptr, namelen) == 0)
    {
      for (ptr += namelen; isspace(*ptr); ptr ++);

      for (bufptr = buffer;
           *ptr && *ptr != ';' && bufptr < (buffer + sizeof(buffer) - 1);
	   *bufptr++ = *ptr++);

      *bufptr = '\0';

      return (buffer);
    }

  return (NULL);
}


/*
 * 'htmlGetVariable()' - Get a variable value from a markup entry.
 */

uchar *				/* O - Value or NULL if variable does not exist */
htmlGetVariable(tree_t *t,	/* I - Tree entry */
                uchar  *name)	/* I - Variable name */
{
  var_t	*v,			/* Matching variable */
	key;			/* Search key */


  if (t == NULL || name == NULL || t->nvars == 0)
    return (NULL);

  key.name = name;

  v = (var_t *)bsearch(&key, t->vars, (size_t)t->nvars, sizeof(var_t), (compare_func_t)compare_variables);
  if (v == NULL)
    return (NULL);
  else if (v->value == NULL)
    return ((uchar *)"");
  else
    return (v->value);
}


/*
 * 'htmlLoadFontWidths()' - Load all of the font width files.
 */

void
htmlLoadFontWidths(int typeface, int style)
{
  char		filename[1024];		/* Filenames */
  FILE		*fp;			/* Files */
  int		ch;			/* Character */
  float		width;			/* Width value */
  char		glyph[64];		/* Glyph name */
  char		line[1024];		/* Line from AFM file */


 /*
  * Now read all of the font widths...
  */

  for (ch = 0; ch < 256; ch ++)
    _htmlWidths[typeface][style][ch] = 600;

  if (_htmlUTF8)
  {
    for (ch = 0; ch < 65536; ch ++)
      _htmlWidthsAll[typeface][style][ch] = 600;
  }

  snprintf(filename, sizeof(filename), "%s/fonts/%s.afm", _htmlData, _htmlFonts[typeface][style]);
  if ((fp = fopen(filename, "r")) == NULL)
  {
#ifndef DEBUG
    progress_error(HD_ERROR_FILE_NOT_FOUND, "Unable to open font width file %s!", filename);
#endif /* !DEBUG */
    return;
  }

  while (fgets(line, sizeof(line), fp) != NULL)
  {
    if (strncmp(line, "C ", 2) != 0)
      continue;

    if (typeface < TYPE_SYMBOL)
    {
     /*
      * Handle encoding of regular fonts using assigned charset...
      */

      if (sscanf(line, "%*s%*s%*s%*s%f%*s%*s%63s", &width, glyph) != 2)
	continue;

      for (ch = 0; ch < 256; ch ++)
      {
	if (_htmlGlyphs[ch] && !strcmp(_htmlGlyphs[ch], glyph))
	{
	  _htmlWidths[typeface][style][ch] = (short)width;
	  break;
	}
      }

      if (_htmlUTF8)
      {
	for (ch = 0; ch < 65536; ch ++)
	{
	  if (_htmlGlyphsAll[ch] && !strcmp(_htmlGlyphsAll[ch], glyph))
	  {
	    _htmlWidthsAll[typeface][style][ch] = (short)width;
	    break;
	  }
	}
      }
    }
    else
    {
     /*
      * Symbol and Dingbats fonts uses their own encoding...
      */

      if (sscanf(line, "%*s%d%*s%*s%f", &ch, &width) != 2)
	continue;

      if (ch < 256 && ch >= 0)
      {
	_htmlWidths[typeface][style][ch]    = (short)width;
	_htmlWidthsAll[typeface][style][ch] = (short)width;
      }
    }
  }

  fclose(fp);

  // Make sure that non-breaking space has the same width as a breaking space...
  _htmlWidths[typeface][style][160]    = _htmlWidths[typeface][style][32];
  _htmlWidthsAll[typeface][style][160] = _htmlWidthsAll[typeface][style][32];

  _htmlWidthsLoaded[typeface][style] = 1;
}


/*
 * 'htmlSetVariable()' - Set a variable for a markup entry.
 */

int				/* O - Set status: 0 = success, -1 = fail */
htmlSetVariable(tree_t *t,	/* I - Tree entry */
                uchar  *name,	/* I - Variable name */
                uchar  *value)	/* I - Variable value */
{
  var_t	*v,			/* Matching variable */
	key;			/* Search key */


  DEBUG_printf(("%shtmlSetVariable(%p, \"%s\", \"%s\")\n", indent, (void *)t, name,
                value ? (const char *)value : "(null)"));

  if (t->nvars == 0)
    v = NULL;
  else
  {
    key.name = name;

    v = (var_t *)bsearch(&key, t->vars, (size_t)t->nvars, sizeof(var_t), (compare_func_t)compare_variables);
  }

  if (v == NULL)
  {
    if (t->nvars == 0)
      v = (var_t *)malloc(sizeof(var_t));
    else
      v = (var_t *)realloc(t->vars, sizeof(var_t) * (size_t)(t->nvars + 1));

    if (v == NULL)
    {
      DEBUG_printf(("%s==== MALLOC/REALLOC FAILED! ====\n", indent));

      return (-1);
    }

    t->vars  = v;
    v        += t->nvars;
    t->nvars ++;
    v->name  = (uchar *)strdup((char *)name);
    if (value != NULL)
      v->value = (uchar *)strdup((char *)value);
    else
      v->value = NULL;

    if (strcasecmp((char *)name, "HREF") == 0)
    {
      DEBUG_printf(("%s---- Set link to %s ----\n", indent, value));
      t->link = t;
    }

    if (t->nvars > 1)
      qsort(t->vars, (size_t)t->nvars, sizeof(var_t), (compare_func_t)compare_variables);
  }
  else if (v->value != value)
  {
    if (v->value != NULL)
      free(v->value);
    if (value != NULL)
      v->value = (uchar *)strdup((char *)value);
    else
      v->value = NULL;
  }

  return (0);
}


/*
 * 'htmlSetBaseSize()' - Set the font sizes and spacings...
 */

void
htmlSetBaseSize(double p,	/* I - Point size of paragraph font */
                double s)	/* I - Spacing */
{
  int	i;			/* Looping var */


  p /= 1.2 * 1.2 * 1.2;
  for (i = 0; i < 8; i ++, p *= 1.2)
  {
    _htmlSizes[i]    = p;
    _htmlSpacings[i] = p * s;
  }
}


/*
 * 'htmlSetCharSet()' - Set the character set for output.
 */

void
htmlSetCharSet(const char *cs)		/* I - Character set file to load */
{
  int		i;			/* Looping var */
  char		filename[1024];		/* Filenames */
  FILE		*fp;			/* Files */
  int		ch, unicode;		/* Character values */
  char		glyph[64];		/* Glyph name */
  char		line[1024];		/* Line from charset file */
  int		chars[256];		/* Character encoding array */


  strlcpy(_htmlCharSet, cs, sizeof(_htmlCharSet));

  if (!_htmlInitialized)
  {
   /*
    * Load the PostScript glyph names for all of Unicode...
    */

    memset(_htmlGlyphsAll, 0, sizeof(_htmlGlyphsAll));

    snprintf(line, sizeof(line), "%s/data/psglyphs", _htmlData);
    if ((fp = fopen(line, "r")) != NULL)
    {
      while (fscanf(fp, "%x%63s", &unicode, glyph) == 2)
        _htmlGlyphsAll[unicode] = strdup(glyph);

      fclose(fp);

      _htmlInitialized = 1;
    }
#ifndef DEBUG
    else
      progress_error(HD_ERROR_FILE_NOT_FOUND,
                     "Unable to open psglyphs data file!");
#endif /* !DEBUG */
  }

  memset(_htmlGlyphs, 0, sizeof(_htmlGlyphs));

  if (!strcmp(cs, "utf-8"))
  {
    // Generate a dynamic mapping of Unicode to an 8-bit charset with the
    // bottom 128 characters matching US ASCII...
    _htmlUTF8 = 0x80;

    for (i = 0; i < 128; i ++)
    {
     /*
      * Add the glyph to the charset array...
      */

      _htmlGlyphs[i]  = _htmlGlyphsAll[i];
      _htmlUnicode[i] = i;
    }

    memset(_htmlWidthsLoaded, 0, sizeof(_htmlWidthsLoaded));
    return;
  }

  if (strncmp(cs, "8859-", 5) == 0)
    snprintf(filename, sizeof(filename), "%s/data/iso-%s", _htmlData, cs);
  else
    snprintf(filename, sizeof(filename), "%s/data/%s", _htmlData, cs);

  if ((fp = fopen(filename, "r")) == NULL)
  {
   /*
    * Can't open charset file; use ISO-8859-1...
    */

#ifndef DEBUG
    progress_error(HD_ERROR_FILE_NOT_FOUND,
                   "Unable to open character set file %s!", cs);
#endif /* !DEBUG */

    for (i = 0; i < 256; i ++)
      chars[i] = i;
  }
  else
  {
   /*
    * Read the <char> <unicode> lines from the file...
    */

    memset(chars, 0, sizeof(chars));

    while (fscanf(fp, "%x%x", &ch, &unicode) == 2)
      chars[ch] = unicode;

    fclose(fp);
  }

 /*
  * Build the glyph array...
  */

  for (i = 0; i < 256; i ++)
  {
   /*
    * Add the glyph to the charset array...
    */

    if (chars[i] == 0)
    {
      _htmlGlyphs[i] = NULL;
      continue;
    }
    else
      _htmlGlyphs[i] = _htmlGlyphsAll[chars[i]];

    if (_htmlGlyphs[i])
      _htmlUnicode[i] = chars[i];
  }

  memset(_htmlWidthsLoaded, 0, sizeof(_htmlWidthsLoaded));
}


/*
 * 'htmlSetTextColor()' - Set the default text color.
 */

void
htmlSetTextColor(uchar *color)	/* I - Text color */
{
  strlcpy((char *)_htmlTextColor, (char *)color, sizeof(_htmlTextColor));
}


/*
 * 'compare_variables()' - Compare two markup variables.
 */

static int			/* O - -1 if v0 < v1, 0 if v0 == v1, 1 if v0 > v1 */
compare_variables(var_t *v0,	/* I - First variable */
                  var_t *v1)	/* I - Second variable */
{
  return (strcasecmp((char *)v0->name, (char *)v1->name));
}


/*
 * 'compare_markups()' - Compare two markup strings...
 */

static int			/* O - -1 if m0 < m1, 0 if m0 == m1, 1 if m0 > m1 */
compare_markups(uchar **m0,	/* I - First markup */
                uchar **m1)	/* I - Second markup */
{
  if (tolower((*m0)[0]) == 'h' && isdigit((*m0)[1]) &&
      tolower((*m1)[0]) == 'h' && isdigit((*m1)[1]))
    return (atoi((char *)*m0 + 1) - atoi((char *)*m1 + 1));
  else
    return (strcasecmp((char *)*m0, (char *)*m1));
}


/*
 * 'delete_node()' - Free all memory associated with a node...
 */

static void
delete_node(tree_t *t)		/* I - Node to delete */
{
  int		i;		/* Looping var */
  var_t		*var;		/* Current variable */


  if (t == NULL)
    return;

  if (t->data != NULL)
    free(t->data);

  for (i = t->nvars, var = t->vars; i > 0; i --, var ++)
  {
    free(var->name);
    if (var->value != NULL)
      free(var->value);
  }

  if (t->vars != NULL)
    free(t->vars);

  free(t);
}


//
// 'insert_space()' - Insert a whitespace character before the
//                    specified node.
//

static void
insert_space(tree_t *parent,	// I - Parent node
             tree_t *t)		// I - Node to insert before
{
  tree_t	*space;		// Space node


  // Allocate memory for the whitespace...
  space = (tree_t *)calloc(sizeof(tree_t), 1);
  if (space == NULL)
  {
#ifndef DEBUG
    progress_error(HD_ERROR_OUT_OF_MEMORY,
                   "Unable to allocate memory for HTML tree node!");
#endif /* !DEBUG */
    return;
  }

  // Set/copy font characteristics...
  if (parent)
  {
    space->typeface = parent->typeface;
    space->size     = parent->size;
    space->style    = parent->style;
  }
  else
  {
    space->typeface = _htmlBodyFont;
    space->size     = SIZE_P;
  }

  // Initialize element data...
  space->markup = MARKUP_NONE;
  space->data   = (uchar *)strdup(" ");

  // Set tree pointers...
  space->parent = parent;
  space->prev   = t->prev;
  space->next   = t;

  if (space->prev)
    space->prev->next = space;
  else if (parent)
    parent->child = space;

  t->prev = space;

  compute_size(space);
}


/*
 * 'parse_markup()' - Parse a markup string.
 */

static int			/* O - -1 on error, MARKUP_nnnn otherwise */
parse_markup(tree_t *t,		/* I - Current tree entry */
             FILE   *fp,	/* I - Input file */
	     int    *linenum)	/* O - Current line number */
{
  int	ch, ch2;		/* Characters from file */
  uchar	markup[255],		/* Markup string... */
	*mptr,			/* Current character... */
	comment[10240],		/* Comment string */
	*cptr,			/* Current char... */
	**temp;			/* Markup variable entry */


  mptr = markup;

  while ((ch = getc(fp)) != EOF && mptr < (markup + sizeof(markup) - 1))
    if (ch == '>' || isspace(ch))
      break;
    else if (ch == '/' && mptr > markup)
    {
      // Look for "/>"...
      ch = getc(fp);

      if (ch != '>')
        return (MARKUP_ERROR);

      break;
    }
    else
    {
      if ((ch & 0x80) && _htmlUTF8)
      {
        // Collect UTF-8 value...
        ch = utf8_getc(ch, fp);
      }

      if (ch)
        *mptr++ = (uchar)ch;

      // Handle comments without whitespace...
      if ((mptr - markup) == 3 && strncmp((const char *)markup, "!--", 3) == 0)
      {
        ch = getc(fp);
        break;
      }
    }

  *mptr = '\0';

  if (ch == EOF)
    return (MARKUP_ERROR);

  mptr = markup;
  temp = (uchar **)bsearch(&mptr, _htmlMarkups,
                           sizeof(_htmlMarkups) / sizeof(_htmlMarkups[0]),
                           sizeof(_htmlMarkups[0]),
                           (compare_func_t)compare_markups);

  if (temp == NULL)
  {
   /*
    * Unrecognized markup stuff...
    */

    t->markup = MARKUP_UNKNOWN;
    strlcpy((char *)comment, (char *)markup, sizeof(comment));
    cptr = comment + strlen((char *)comment);

    DEBUG_printf(("%s%s (unrecognized!)\n", indent, markup));
  }
  else
  {
    t->markup = (markup_t)((const char **)temp - _htmlMarkups);
    cptr      = comment;

    DEBUG_printf(("%s%s, line %d\n", indent, markup, *linenum));
  }

  if (t->markup == MARKUP_COMMENT || t->markup == MARKUP_UNKNOWN)
  {
    int lastch = ch;			// Last character seen

    while (ch != EOF && cptr < (comment + sizeof(comment) - 2))
    {
      if (ch == '>' && temp == NULL)
        break;

      if (ch == '\n')
        (*linenum) ++;

      if (ch == '-' && lastch == '-')
      {
        *cptr++ = (uchar)ch;

        if ((ch2 = getc(fp)) == '>')
	{
	  // Erase trailing -->
	  cptr -= 2;

          if (cptr < comment)
            cptr = comment; // Issue #316: buffer underflow
	  break;
        }
	else
	  ch = ch2;
      }
      else
      {
        if (ch == '&')
	{
          // Handle character entities...
	  uchar	entity[16],		// Character entity name
		*eptr;			// Pointer into name


	  eptr = entity;
	  while (eptr < (entity + sizeof(entity) - 1) &&
		 (ch = getc(fp)) != EOF)
	    if (!isalnum(ch) && ch != '#')
	      break;
	    else
	      *eptr++ = (uchar)ch;

	  if (ch != ';')
	  {
	    ungetc(ch, fp);
	    ch = 0;
	  }

	  *eptr = '\0';
	  if (!ch)
	  {
	    progress_error(HD_ERROR_HTML_ERROR, "Unquoted & on line %d of %s.",
	                   *linenum, _htmlCurrentFile);

            if (cptr < (comment + sizeof(comment) - 1))
	      *cptr++ = '&';
            strlcpy((char *)cptr, (char *)entity, sizeof(comment) - (size_t)(cptr - comment));
	    cptr += strlen((char *)cptr);
	  }
	  else if ((ch = iso8859(entity)) == 0)
	  {
	    progress_error(HD_ERROR_HTML_ERROR,
	                   "Unknown character entity \"&%s;\" on line %d of %s.",
	                   entity, *linenum, _htmlCurrentFile);

            if (cptr < (comment + sizeof(comment) - 1))
	      *cptr++ = '&';
            strlcpy((char *)cptr, (char *)entity, sizeof(comment) - (size_t)(cptr - comment));
	    cptr += strlen((char *)cptr);
            if (cptr < (comment + sizeof(comment) - 1))
	      *cptr++ = ';';
	  }
	  else
	    *cptr++ = (uchar)ch;
	}
	else
        {
          if ((ch & 0x80) && _htmlUTF8)
          {
            // Collect UTF-8 value...
            ch = utf8_getc(ch, fp);
          }

          if (ch)
            *cptr++ = (uchar)ch;
        }

        lastch = ch;
        ch     = getc(fp);
      }
    }

    *cptr = '\0';
    t->data = (uchar *)strdup((char *)comment);
  }
  else
  {
    while (ch != EOF && ch != '>')
    {
      if (ch == '\n')
        (*linenum) ++;

      if (!isspace(ch))
      {
        ungetc(ch, fp);
        parse_variable(t, fp, linenum);
      }

      ch = getc(fp);

      if (ch == '/')
      {
	// Look for "/>"...
	ch = getc(fp);

	if (ch != '>')
          return (MARKUP_ERROR);

	break;
      }
    }
  }

  return (t->markup);
}


/*
 * 'parse_variable()' - Parse a markup variable string.
 */

static int				// O - -1 on error, 0 on success
parse_variable(tree_t *t,		// I - Current tree entry
               FILE   *fp,		// I - Input file
	       int    *linenum)		// I - Current line number
{
  uchar	name[1024],			// Name of variable
	value[10240],			// Value of variable
	*ptr,				// Temporary pointer
	entity[16],			// Character entity name
	*eptr;				// Pointer into name
  int	ch;				// Character from file


  ptr = name;
  while ((ch = getc(fp)) != EOF)
    if (isspace(ch) || ch == '=' || ch == '>' || ch == '\r')
      break;
    else if (ch == '/' && ptr == name)
      break;
    else if (ptr < (name + sizeof(name) - 1))
    {
      if ((ch & 0x80) && _htmlUTF8)
      {
        // Collect UTF-8 value...
        ch = utf8_getc(ch, fp);
      }

      if (ch)
        *ptr++ = (uchar)ch;
    }

  *ptr = '\0';

  if (ch == '\n')
    (*linenum) ++;

  while (isspace(ch) || ch == '\r')
  {
    ch = getc(fp);

    if (ch == '\n')
      (*linenum) ++;
  }

  switch (ch)
  {
    default :
        ungetc(ch, fp);
        return (htmlSetVariable(t, name, NULL));
    case EOF :
        return (-1);
    case '=' :
        ptr = value;
        ch  = getc(fp);

        while (isspace(ch) || ch == '\r')
          ch = getc(fp);

        if (ch == EOF)
          return (-1);

        if (ch == '\'')
        {
          while ((ch = getc(fp)) != EOF)
	  {
            if (ch == '\'')
              break;
	    else if (ch == '&')
	    {
	      // Possibly a character entity...
	      eptr = entity;
	      while (eptr < (entity + sizeof(entity) - 1) &&
	             (ch = getc(fp)) != EOF)
	        if (!isalnum(ch) && ch != '#')
		  break;
		else
		  *eptr++ = (uchar)ch;

              if (ch != ';')
	      {
	        ungetc(ch, fp);
		ch = 0;
	      }

              *eptr = '\0';
              if (!ch)
	      {
		progress_error(HD_ERROR_HTML_ERROR,
		               "Unquoted & on line %d of %s.",
	                       *linenum, _htmlCurrentFile);

        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = '&';
                strlcpy((char *)ptr, (char *)entity, sizeof(value) - (size_t)(ptr - value));
		ptr += strlen((char *)ptr);
	      }
	      else if ((ch = iso8859(entity)) == 0)
	      {
		progress_error(HD_ERROR_HTML_ERROR,
		               "Unknown character entity \"&%s;\" on line %d of %s.",
	                       entity, *linenum, _htmlCurrentFile);

        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = '&';
                strlcpy((char *)ptr, (char *)entity, sizeof(value) - (size_t)(ptr - value));
		ptr += strlen((char *)ptr);
        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = ';';
	      }
	      else if (ptr < (value + sizeof(value) - 1))
	        *ptr++ = (uchar)ch;
	    }
            else if (ptr < (value + sizeof(value) - 1) &&
	             ch != '\n' && ch != '\r')
            {
              if ((ch & 0x80) && _htmlUTF8)
              {
                // Collect UTF-8 value...
                ch = utf8_getc(ch, fp);
              }

              if (ch)
                *ptr++ = (uchar)ch;
            }
	    else if (ch == '\n')
	    {
	      if (ptr < (value + sizeof(value) - 1))
	        *ptr++ = ' ';

	      (*linenum) ++;
	    }
	  }

          *ptr = '\0';
        }
        else if (ch == '\"')
        {
          while ((ch = getc(fp)) != EOF)
	  {
            if (ch == '\"')
              break;
	    else if (ch == '&')
	    {
	      // Possibly a character entity...
	      eptr = entity;
	      while (eptr < (entity + sizeof(entity) - 1) &&
	             (ch = getc(fp)) != EOF)
	        if (!isalnum(ch) && ch != '#')
		  break;
		else
		  *eptr++ = (uchar)ch;

              if (ch != ';')
	      {
	        ungetc(ch, fp);
		ch = 0;
	      }

              *eptr = '\0';
              if (!ch)
	      {
		progress_error(HD_ERROR_HTML_ERROR, "Unquoted & on line %d of %s.",
	                       *linenum, _htmlCurrentFile);

        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = '&';
                strlcpy((char *)ptr, (char *)entity, sizeof(value) - (size_t)(ptr - value));
		ptr += strlen((char *)ptr);
	      }
	      else if ((ch = iso8859(entity)) == 0)
	      {
		progress_error(HD_ERROR_HTML_ERROR,
		               "Unknown character entity \"&%s;\" on line %d of %s.",
	                       entity, *linenum, _htmlCurrentFile);

        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = '&';
                strlcpy((char *)ptr, (char *)entity, sizeof(value) - (size_t)(ptr - value));
		ptr += strlen((char *)ptr);
        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = ';';
	      }
	      else if (ptr < (value + sizeof(value) - 1))
	        *ptr++ = (uchar)ch;
	    }
            else if (ptr < (value + sizeof(value) - 1) &&
	             ch != '\n' && ch != '\r')
            {
              if ((ch & 0x80) && _htmlUTF8)
              {
                // Collect UTF-8 value...
                ch = utf8_getc(ch, fp);
              }

              if (ch)
                *ptr++ = (uchar)ch;
            }
	    else if (ch == '\n')
	    {
	      if (ptr < (value + sizeof(value) - 1))
	        *ptr++ = ' ';

	      (*linenum) ++;
	    }
	  }

          *ptr = '\0';
        }
        else
        {
          *ptr++ = (uchar)ch;
          while ((ch = getc(fp)) != EOF)
	  {
            if (isspace(ch) || ch == '>' || ch == '\r')
              break;
	    else if (ch == '&')
	    {
	      // Possibly a character entity...
	      eptr = entity;
	      while (eptr < (entity + sizeof(entity) - 1) &&
	             (ch = getc(fp)) != EOF)
	        if (!isalnum(ch) && ch != '#')
		  break;
		else
		  *eptr++ = (uchar)ch;

              if (ch != ';')
	      {
	        ungetc(ch, fp);
		ch = 0;
	      }

              *eptr = '\0';
              if (!ch)
	      {
		progress_error(HD_ERROR_HTML_ERROR, "Unquoted & on line %d of %s.",
	                       *linenum, _htmlCurrentFile);

        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = '&';
                strlcpy((char *)ptr, (char *)entity, sizeof(value) - (size_t)(ptr - value));
		ptr += strlen((char *)ptr);
	      }
	      else if ((ch = iso8859(entity)) == 0)
	      {
		progress_error(HD_ERROR_HTML_ERROR,
		               "Unknown character entity \"&%s;\" on line %d of %s.",
	                       entity, *linenum, _htmlCurrentFile);

        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = '&';
                strlcpy((char *)ptr, (char *)entity, sizeof(value) - (size_t)(ptr - value));
		ptr += strlen((char *)ptr);
        	if (ptr < (value + sizeof(value) - 1))
		  *ptr++ = ';';
	      }
	      else if (ptr < (value + sizeof(value) - 1))
	        *ptr++ = (uchar)ch;
	    }
            else if (ptr < (value + sizeof(value) - 1))
            {
              if ((ch & 0x80) && _htmlUTF8)
              {
                // Collect UTF-8 value...
                ch = utf8_getc(ch, fp);
              }

              if (ch)
                *ptr++ = (uchar)ch;
            }
	  }

	  if (ch == '\n')
	    (*linenum) ++;

          *ptr = '\0';
          if (ch == '>')
            ungetc(ch, fp);
        }

        return (htmlSetVariable(t, name, value));
  }
}


/*
 * 'compute_size()' - Compute the width and height of a tree entry.
 */

static int			/* O - 0 = success, -1 = failure */
compute_size(tree_t *t)		/* I - Tree entry */
{
  uchar		*ptr;		/* Current character */
  float		width;		/* Current width */
  int		int_width;	/* Integer width */
  uchar		*width_ptr,	/* Pointer to width string */
		*height_ptr,	/* Pointer to height string */
		*size_ptr,	/* Pointer to size string */
		*type_ptr;	/* Pointer to spacer type string */
  image_t	*img;		/* Image */
  char		number[255];	/* Width or height value */


  if (!_htmlInitialized)
    htmlSetCharSet("iso-8859-1");

  if (t->markup == MARKUP_IMG)
  {
    width_ptr  = htmlGetVariable(t, (uchar *)"WIDTH");
    height_ptr = htmlGetVariable(t, (uchar *)"HEIGHT");

    img = image_load((char *)htmlGetVariable(t, (uchar *)"REALSRC"),
                     _htmlGrayscale);

    if (width_ptr != NULL && height_ptr != NULL)
    {
      t->width  = (float)(atoi((char *)width_ptr) / _htmlPPI * 72.0f);
      t->height = (float)(atoi((char *)height_ptr) / _htmlPPI * 72.0f);

      return (0);
    }

    if (img == NULL)
      return (-1);

    if (width_ptr != NULL)
    {
      t->width  = (float)(atoi((char *)width_ptr) / _htmlPPI * 72.0f);
      t->height = (float)(t->width * img->height / img->width);

      snprintf(number, sizeof(number), "%d", atoi((char *)width_ptr) * img->height / img->width);
      if (strchr((char *)width_ptr, '%') != NULL)
        strlcat(number, "%", sizeof(number));
      htmlSetVariable(t, (uchar *)"HEIGHT", (uchar *)number);
    }
    else if (height_ptr != NULL)
    {
      t->height = (float)(atoi((char *)height_ptr) / _htmlPPI * 72.0f);
      t->width  = (float)(t->height * img->width / img->height);

      snprintf(number, sizeof(number), "%d", atoi((char *)height_ptr) * img->width / img->height);
      if (strchr((char *)height_ptr, '%') != NULL)
        strlcat(number, "%", sizeof(number));
      htmlSetVariable(t, (uchar *)"WIDTH", (uchar *)number);
    }
    else
    {
      t->width  = (float)(img->width / _htmlPPI * 72.0f);
      t->height = (float)(img->height / _htmlPPI * 72.0f);

      snprintf(number, sizeof(number), "%d", img->width);
      htmlSetVariable(t, (uchar *)"WIDTH", (uchar *)number);

      snprintf(number, sizeof(number), "%d", img->height);
      htmlSetVariable(t, (uchar *)"HEIGHT", (uchar *)number);
    }

    return (0);
  }
  else if (t->markup == MARKUP_SPACER)
  {
    width_ptr  = htmlGetVariable(t, (uchar *)"WIDTH");
    height_ptr = htmlGetVariable(t, (uchar *)"HEIGHT");
    size_ptr   = htmlGetVariable(t, (uchar *)"SIZE");
    type_ptr   = htmlGetVariable(t, (uchar *)"TYPE");

    if (width_ptr != NULL)
      t->width = (float)(atoi((char *)width_ptr) / _htmlPPI * 72.0f);
    else if (size_ptr != NULL)
      t->width = (float)(atoi((char *)size_ptr) / _htmlPPI * 72.0f);
    else
      t->width = 1.0f;

    if (height_ptr != NULL)
      t->height = (float)(atoi((char *)height_ptr) / _htmlPPI * 72.0f);
    else if (size_ptr != NULL)
      t->height = (float)(atoi((char *)size_ptr) / _htmlPPI * 72.0f);
    else
      t->height = 1.0f;

    if (type_ptr == NULL)
      return (0);

    if (strcasecmp((char *)type_ptr, "horizontal") == 0)
      t->height = 0.0;
    else if (strcasecmp((char *)type_ptr, "vertical") == 0)
      t->width = 0.0;

    return (0);
  }
  else if (t->markup == MARKUP_BR)
  {
    t->width  = 0.0;
    t->height = (float)_htmlSizes[t->size];

    return (0);
  }
  else if (t->preformatted && t->data)
  {
    int		max_width = 0;		/* Maximum width */

    for (int_width = 0, ptr = t->data; *ptr != '\0'; ptr ++)
    {
      if (*ptr == '\n')
      {
        if (int_width > max_width)
          max_width = int_width;

	int_width = 0;
      }
      else if (*ptr == '\t')
        int_width = (int_width + 7) & ~7;
      else
        int_width ++;
    }

    if (int_width > max_width)
      max_width = int_width;

    if (!_htmlWidthsLoaded[t->typeface][t->style])
      htmlLoadFontWidths(t->typeface, t->style);

    width = _htmlWidths[t->typeface][t->style][0x20] * max_width * 0.001f;
  }
  else if (t->data)
  {
    if (!_htmlWidthsLoaded[t->typeface][t->style])
      htmlLoadFontWidths(t->typeface, t->style);

    for (int_width = 0, ptr = t->data; *ptr != '\0'; ptr ++)
      int_width += _htmlWidths[t->typeface][t->style][(int)*ptr & 255];

    width = 0.001f * int_width;
  }
  else
    width = 0.0f;

  t->width  = (float)(width * _htmlSizes[t->size]);
  t->height = (float)_htmlSizes[t->size];

  DEBUG_printf(("%swidth = %.1f, height = %.1f\n", indent, t->width, t->height));

  return (0);
}


/*
 * 'compute_color()' - Compute the red, green, blue color from the given
 *                     string.
 */

static int
compute_color(tree_t *t,	/* I - Tree entry */
              uchar  *color)	/* I - Color string */
{
  float	rgb[3];			/* RGB color */


  get_color(color, rgb);

  t->red   = (uchar)(rgb[0] * 255.0f + 0.5f);
  t->green = (uchar)(rgb[1] * 255.0f + 0.5f);
  t->blue  = (uchar)(rgb[2] * 255.0f + 0.5f);

  return (0);
}


/*
 * 'get_alignment()' - Get horizontal & vertical alignment values.
 */

static int			/* O - 0 for success, -1 for failure */
get_alignment(tree_t *t)	/* I - Tree entry */
{
  uchar	*align;			/* Alignment string */


  if ((align = htmlGetVariable(t, (uchar *)"ALIGN")) == NULL)
    align = htmlGetStyle(t, (uchar *)"text-align");

  if (align != NULL)
  {
    if (!strcasecmp((char *)align, "left"))
      t->halignment = ALIGN_LEFT;
    else if (!strcasecmp((char *)align, "center"))
      t->halignment = ALIGN_CENTER;
    else if (!strcasecmp((char *)align, "right"))
      t->halignment = ALIGN_RIGHT;
    else if (!strcasecmp((char *)align, "justify"))
      t->halignment = ALIGN_JUSTIFY;
    else if (!strcasecmp((char *)align, "top"))
      t->valignment = ALIGN_TOP;
    else if (!strcasecmp((char *)align, "middle") ||
             !strcasecmp((char *)align, "absmiddle"))
      t->valignment = ALIGN_MIDDLE;
    else if (!strcasecmp((char *)align, "bottom"))
      t->valignment = ALIGN_BOTTOM;
  }

  if ((align = htmlGetVariable(t, (uchar *)"VALIGN")) == NULL)
    align = htmlGetStyle(t, (uchar *)"vertical-align");

  if (align != NULL)
  {
    if (!strcasecmp((char *)align, "top"))
      t->valignment = ALIGN_TOP;
    else if (!strcasecmp((char *)align, "middle"))
      t->valignment = ALIGN_MIDDLE;
    else if (!strcasecmp((char *)align, "center"))
      t->valignment = ALIGN_MIDDLE;
    else if (!strcasecmp((char *)align, "bottom"))
      t->valignment = ALIGN_BOTTOM;
  }

  return (0);
}


/*
 * 'fix_filename()' - Fix a filename to be relative to the base directory.
 */

static const char *			/* O - Fixed filename */
fix_filename(char *filename,		/* I - Original filename */
             char *base)		/* I - Base directory */
{
  char		*slash;			/* Location of slash */
  char		*tempptr;		/* Pointer into filename */
  static char	temp[1024];		/* Temporary filename */
  static char	newfilename[1024];	/* New filename */


//  printf("fix_filename(filename=\"%s\", base=\"%s\")\n", filename, base);

  if (filename == NULL)
    return (NULL);

#ifdef DEBUG // to silence Clang static analyzer, totally unnecessary
  memset(temp, 0, sizeof(temp));
#endif // DEBUG

  // Unescape filenames as needed...
  if (strchr(filename, '%') && !strstr(filename, "//"))
  {
    for (tempptr = temp; *filename && tempptr < (temp + sizeof(temp) - 1);)
    {
      if (*filename == '%')
      {
        // Decode hex-escaped filename character...
	filename ++;
	if (isxdigit(filename[0] & 255) && isxdigit(filename[1] & 255))
	{
	  if (isdigit(filename[0] & 255))
	    *tempptr = (char)((filename[0] - '0') << 4);
	  else
	    *tempptr = (char)((tolower(filename[0]) - 'a' + 10) << 4);

	  if (isdigit(filename[1] & 255))
	    *tempptr |= filename[1] - '0';
	  else
	    *tempptr |= tolower(filename[0]) - 'a' + 10;

          tempptr ++;
	  filename += 2;
	}
	else
	  *tempptr++ = '%';
      }
      else
        *tempptr++ = *filename++;
    }

    *tempptr = '\0';
    filename = temp;
  }

  if (strcmp(base, ".") == 0 || strstr(filename, "//") != NULL)
    return (file_find(Path, filename));

  if (strncmp(filename, "./", 2) == 0 ||
      strncmp(filename, ".\\", 2) == 0)
    filename += 2;

  if (strncmp(base, "http://", 7) == 0 || strncmp(base, "https://", 8) == 0)
  {
    // Base is a URL...
    char	scheme[32],		// URI scheme
		userpass[256],		// Username:password
		host[256],		// Hostname or IP address
		resource[256];		// Resource path
    int		port;			// Port number

    httpSeparateURI(HTTP_URI_CODING_ALL, base, scheme, sizeof(scheme), userpass, sizeof(userpass), host, sizeof(host), &port, resource, sizeof(resource));

    if (filename[0] == '/')
    {
      // Absolute path, so just use the server...
      httpAssembleURI(HTTP_URI_CODING_ALL, newfilename, sizeof(newfilename), scheme, userpass, host, port, filename);
    }
    else
    {
      // Relative path, strip the last component from the resource...
      if ((slash = strrchr(resource, '/')) != NULL)
	*slash = '\0';

      // Handle "../" in filename...
      while (!strncmp(filename, "../", 3))
      {
	// Strip one level of directory in the resource
	filename += 3;

	if ((slash = strrchr(resource, '/')) != NULL)
	  *slash = '\0';
      }

      // Combine the resource and remaining relative filename to make a URL...
      httpAssembleURIf(HTTP_URI_CODING_ALL, newfilename, sizeof(newfilename), scheme, userpass, host, port, "%s/%s", resource, filename);
    }
  }
  else
  {
    // Base is a filename...
    if (filename[0] == '/' || filename[0] == '\\' || base == NULL ||
	base[0] == '\0' || (isalpha(filename[0]) && filename[1] == ':'))
    {
      // No change needed for absolute path...
      return (file_find(Path, filename));
    }

    strlcpy(newfilename, base, sizeof(newfilename));
    base = newfilename;

#if defined(WIN32) || defined(__EMX__)
    while (!strncmp(filename, "../", 3) || !strncmp(filename, "..\\", 3))
#else
    while (!strncmp(filename, "../", 3))
#endif // WIN32 || __EMX__
    {
      filename += 3;
#if defined(WIN32) || defined(__EMX__)
      if ((slash = strrchr(base, '/')) != NULL)
        *slash = '\0';
      else if ((slash = strrchr(base, '\\')) != NULL)
        *slash = '\0';
#else
      if ((slash = strrchr(base, '/')) != NULL)
        *slash = '\0';
#endif // WIN32 || __EMX__
      else
      {
	filename -= 3;
	break;
      }
    }

    if (filename[0] != '/' && *base && base[strlen(base) - 1] != '/')
      strlcat(newfilename, "/", sizeof(newfilename));

    strlcat(newfilename, filename, sizeof(newfilename));
  }

//  printf("    newfilename=\"%s\"\n", newfilename);

  return (file_find(Path, newfilename));
}


//
// 'html_memory_used()' - Figure out the amount of memory that was used.
//

static int				// O - Bytes used
html_memory_used(tree_t *t)		// I - Tree node
{
  int	i;				// Looping var
  int	bytes;				// Bytes used


  if (t == NULL)
    return (0);

  bytes = 0;

  while (t != NULL)
  {
    bytes += sizeof(tree_t);
    bytes += (size_t)t->nvars * sizeof(var_t);

    for (i = 0; i < t->nvars; i ++)
    {
      bytes += (strlen((char *)t->vars[i].name) + 8) & (size_t)~7;

      if (t->vars[i].value != NULL)
        bytes += (strlen((char *)t->vars[i].value) + 8) & (size_t)~7;
    }

    if (t->data != NULL)
      bytes += (strlen((char *)t->data) + 8) & (size_t)~7;

    bytes += html_memory_used(t->child);

    t = t->next;
  }

  return (bytes);
}


//
// 'htmlDebugStats()' - Display debug statistics for HTML tree memory use.
//

void
htmlDebugStats(const char *title,	// I - Title
               tree_t     *t)		// I - Document root node
{
  const char	*debug;			/* HTMLDOC_DEBUG env var */


  if ((debug = getenv("HTMLDOC_DEBUG")) == NULL ||
      (strstr(debug, "all") == NULL && strstr(debug, "memory") == NULL))
    return;

  progress_error(HD_ERROR_NONE, "DEBUG: %s = %d kbytes", title,
                 (html_memory_used(t) + 1023) / 1024);
}


//
// 'htmlFindFile()' - Find a file in the document.
//

tree_t *				// O - Node for file
htmlFindFile(tree_t *doc,		// I - Document pointer
             uchar  *filename)		// I - Filename
{
  tree_t	*tree;			// Current node
  uchar		*treename;		// Filename from node


  if (!filename || !doc)
    return (NULL);

  for (tree = doc; tree; tree = tree->next)
    if ((treename = htmlGetVariable(tree, (uchar *)"_HD_FILENAME")) != NULL &&
        !strcmp((char *)treename, (char *)filename))
      return (tree);

  return (NULL);
}


//
// 'htmlFixLinks()' - Fix the external links in the document.
//

void
htmlFixLinks(tree_t *doc,		// I - Top node
             tree_t *tree,		// I - Current node
	     uchar  *base)		// I - Base directory/path
{
  uchar		*href;			// HREF attribute
  char		full_href[1024];	// Full HREF value
  const char	*debug;			// HTMLDOC_DEBUG environment variable
  static int	show_debug = -1;	// Show debug messages?


  if (show_debug < 0)
  {
    if ((debug = getenv("HTMLDOC_DEBUG")) == NULL ||
	(strstr(debug, "all") == NULL && strstr(debug, "links") == NULL))
      show_debug = 0;
    else
      show_debug = 1;

    if (show_debug)
      progress_error(HD_ERROR_NONE, "DEBUG: Updating links in document.");
  }

  while (tree)
  {
    if (tree->markup == MARKUP_A && base && base[0] &&
        (href = htmlGetVariable(tree, (uchar *)"HREF")) != NULL)
    {
      // Check if the link needs to be localized...
      if (href[0] != '#' && file_method((char *)href) == NULL &&
          file_method((char *)base) != NULL &&
	  htmlFindFile(doc, (uchar *)file_basename((char *)href)) == NULL)
      {
        // Yes, localize it...
	if (href[0] == '/')
	{
	  // Absolute URL, just copy scheme, server, etc.
	  char *ptr;			// Pointer into URL...

	  strlcpy(full_href, (char *)base, sizeof(full_href));

          if (href[1] == '/')
	  {
	    // Just use scheme...
	    if ((ptr = strstr(full_href, "//")) != NULL)
	      *ptr ='\0';
	  }
	  else if ((ptr = strstr(full_href, "//")) != NULL  &&
	           (ptr = strchr(ptr + 2, '/')) != NULL)
	    *ptr ='\0';

	  strlcat(full_href, (char *)href, sizeof(full_href));
	}
	else if (!strncmp((char *)href, "./", 2))
	{
	  // Relative URL of the form "./foo/bar", append href sans
	  // "./" to base to form full href...
	  snprintf(full_href, sizeof(full_href), "%s/%s", base, href + 2);
	}
	else
	{
	  // Relative URL, append href to base to form full href...
	  snprintf(full_href, sizeof(full_href), "%s/%s", base, href);
	}

        if (show_debug)
          progress_error(HD_ERROR_NONE, "DEBUG: Mapping \"%s\" to \"%s\"...", href, full_href);

	htmlSetVariable(tree, (uchar *)"_HD_FULL_HREF", (uchar *)full_href);
      }
      else
      {
        // No, just mirror the link in the _HD_FULL_HREF attribute...
	htmlSetVariable(tree, (uchar *)"_HD_FULL_HREF", href);
      }
    }
    else if (tree->markup == MARKUP_FILE)
      base = htmlGetVariable(tree, (uchar *)"_HD_BASE");

    if (tree->child)
      htmlFixLinks(doc, tree->child, base);

    tree = tree->next;
  }
}


//
// 'utf8_getc()' - Get a UTF-8 encoded character.
//

static int                              // O - Unicode equivalent
utf8_getc(int  ch,                      // I - Initial character
          FILE *fp)                     // I - File to read from
{
  int  ch2 = -1, ch3 = -1;              // Temporary characters


  if ((ch & 0xe0) == 0xc0)
  {
   /*
    * Two-byte sequence for 0x80 to 0x7ff...
    */

    ch  = (ch & 0x1f) << 6;
    ch2 = getc(fp);

    if ((ch2 & 0xc0) == 0x80)
      ch |= ch2 & 0x3f;
    else
      goto bad_sequence;
  }
  else if ((ch & 0xf0) == 0xe0)
  {
   /*
    * Three-byte sequence from 0x800 to 0xffff...
    */

    ch  = (ch & 0x0f) << 12;
    ch2 = getc(fp);

    if ((ch2 & 0xc0) == 0x80)
      ch |= (ch2 & 0x3f) << 6;
    else
      goto bad_sequence;

    ch3 = getc(fp);

    if ((ch3 & 0xc0) == 0x80)
      ch |= ch3 & 0x3f;
    else
      goto bad_sequence;
  }
  else if (ch & 0x80)
    goto bad_sequence;

  if (ch == 0xfeff)
  {
    // BOMs are invalid in UTF-8 text, but Microsoft insists on still using
    // them...  Try reading another character...
    //
    // TODO: Emit a warning about this...
    return (utf8_getc(getc(fp), fp));
  }

  return (htmlMapUnicode(ch));

  bad_sequence:

  if (ch3 >= 0)
    progress_error(HD_ERROR_READ_ERROR, "Bad UTF-8 character sequence %02X %02X %02X.", ch, ch2, ch3);
  else if (ch2 >= 0)
    progress_error(HD_ERROR_READ_ERROR, "Bad UTF-8 character sequence %02X %02X.", ch, ch2);
  else
    progress_error(HD_ERROR_READ_ERROR, "Bad UTF-8 character sequence %02X.", ch);

  return (0);
}
Anon7 - 2022
AnonSec Team