htmlparser.cpp

monqueror一个很具有参考价值的源玛

/*
    This file is part of the KDE libraries

    Copyright (C) 1997 Martin Jones ([e-mail])
              (C) 1997 Torben Weis ([e-mail])
              (C) 1999 Lars Knoll ([e-mail])

    This library is free software; you can redistribute it and/or
    modify it under the terms of the GNU Library General Public
    License as published by the Free Software Foundation; either
    version 2 of the License, or (at your option) any later version.

    This library is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
    Library General Public License for more details.

    You should have received a copy of the GNU Library General Public License
    along with this library; see the file COPYING.LIB.  If not, write to
    the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
    Boston, MA 02111-1307, USA.
*/
//----------------------------------------------------------------------------
//
// KDE HTML Widget -- HTML Parser
// $Id: htmlparser.cpp,v 1.2 2002/01/28 04:31:02 leon Exp $

//#define PARSER_DEBUG
//#define COMMENTS_IN_DOM

#include "render_interface.h"
#include "htmlparser.h"

#include "dom_nodeimpl.h"
#include "dom_exception.h"


#include "html_baseimpl.h"
#include "html_blockimpl.h"
#include "html_documentimpl.h"
#include "html_elementimpl.h"
#include "html_formimpl.h"
#include "html_headimpl.h"
#include "html_imageimpl.h"
#include "html_inlineimpl.h"
#include "html_listimpl.h"
#include "html_miscimpl.h"
#include "html_tableimpl.h"
#include "html_objectimpl.h"


#include "dom_textimpl.h"

#include "htmlhashes.h"
#include "htmltoken.h"
#include "mghtmlview.h"
#include "mghtml_part.h"
#include "cssproperties.h"

#include "render_object.h"

#include "kdebug.h"

using namespace DOM;
using namespace khtml;

//----------------------------------------------------------------------------

// ### FIXME: move this list to dtd.cpp
//
// priority of tags. Closing tags of higher priority close tags of lower
// priority.
// Update this list, whenever you change htmltags.*
//
// 0 elements with forbidden close tag and text. They don't get pushed
//   to the stack.
// 1 inline elements
// 2 form elements
// 3 regular block level elements
// 4 lists (OL UL DIR MENU)
// 5 TD TH
// 6 TR
// 7 tbody thead tfoot caption
// 8 table
// 9 body frameset
// 10 html
const unsigned short tagPriority[] = {
    0, // 0
    1, // ID_A == 1
    1, // ID_ABBR
    1, // ID_ACRONYM
    3, // ID_ADDRESS
    1, // ID_APPLET
    0, // ID_AREA
    1, // ID_B
    0, // ID_BASE
    0, // ID_BASEFONT
    1, // ID_BDO
    1, // ID_BIG
    3, // ID_BLOCKQUOTE
    9, // ID_BODY
    0, // ID_BR
    1, // ID_BUTTON
    1, // ID_CAPTION
    3, // ID_CENTER
    1, // ID_CITE
    1, // ID_CODE
    0, // ID_COL
    1, // ID_COLGROUP
    3, // ID_DD
    1, // ID_DEL
    1, // ID_DFN
    4, // ID_DIR
    3, // ID_DIV
    4, // ID_DL
    3, // ID_DT
    1, // ID_EM
    0, // ID_EMBED
    3, // ID_FIELDSET
    1, // ID_FONT
    3, // ID_FORM
    0, // ID_FRAME
    9, // ID_FRAMESET
    3, // ID_H1
    3, // ID_H2
    3, // ID_H3
    3, // ID_H4
    3, // ID_H5
    3, // ID_H6
    9, // ID_HEAD
    0, // ID_HR
    10, // ID_HTML
    1, // ID_I
    1, // ID_IFRAME
    0, // ID_IMG
    0, // ID_INPUT
    1, // ID_INS
    0, // ID_ISINDEX
    1, // ID_KBD
    1, // ID_LABEL
    1, // ID_LEGEND
    3, // ID_LI
    0, // ID_LINK
    1, // ID_LISTING
    1, // ID_MAP
    4, // ID_MENU
    0, // ID_META
    9, // ID_NOEMBED
    9, // ID_NOFRAMES
    3, // ID_NOSCRIPT
    1, // ID_OBJECT
    4, // ID_OL
    1, // ID_OPTGROUP
    2, // ID_OPTION
    3, // ID_P
    0, // ID_PARAM
    1, // ID_PLAIN
    1, // ID_PRE
    1, // ID_Q
    1, // ID_S
    1, // ID_SAMP
    1, // ID_SCRIPT
    2, // ID_SELECT
    1, // ID_SMALL
    1, // ID_SPAN
    1, // ID_STRIKE
    1, // ID_STRONG
    1, // ID_STYLE
    1, // ID_SUB
    1, // ID_SUP
    8, // ID_TABLE
    7, // ID_TBODY
    5, // ID_TD
    1, // ID_TEXTAREA
    7, // ID_TFOOT
    5, // ID_TH
    7, // ID_THEAD
    1, // ID_TITLE
    6, // ID_TR
    1, // ID_TT
    1, // ID_U
    4, // ID_UL
    1, // ID_VAR
    0, // ID_TEXT
};


/**
 * @internal
 */
class HTMLStackElem
{
public:
    HTMLStackElem( int _id,
                   int _level,
                   DOM::NodeImpl *_node,
                   HTMLStackElem * _next
        )
        :
        id(_id),
        level(_level),
        node(_node),
        next(_next)
        { }

    int       id;
    int       level;
    NodeImpl *node;
    HTMLStackElem *next;
};


/**
 * @internal
 *
 * The parser parses tokenized input into the document, building up the
 * document tree. If the document is wellformed, parsing it is
 * straightforward.
 * Unfortunately, people can't write wellformed HTML documents, so the parser
 * has to be tolerant about errors.
 *
 * We have to take care of the following error conditions:
 * 1. The element being added is explicitly forbidden inside some outer tag.
 *    In this case we should close all tags up to the one, which forbids
 *    the element, and add it afterwards.
 * 2. We are not allowed to add the element directly. It could be, that
 *    the person writing the document forgot some tag inbetween (or that the
 *    tag inbetween is optional...) This could be the case with the following
 *    tags: HTML HEAD BODY TBODY TR TD LI (did I forget any?)
 * 3. We wan't to add a block element inside to an inline element. Close all
 *    inline elements up to the next higher block element.
 * 4. If this doesn't help close elements, until we are allowed to add the
 *    element or ignore the tag.
 *
 */
KHTMLParser::KHTMLParser( MGHTMLView *_parent,
                          HTMLDocumentImpl *doc)
{
    ////kdDebug( 6035 ) << "parser constructor" << endl;

    HTMLWidget    = _parent;
    document      = doc;

    blockStack = 0;

    // ID_CLOSE_TAG == Num of tags
    forbiddenTag = new ushort[ID_CLOSE_TAG+1];

    reset();
}

KHTMLParser::~KHTMLParser()
{
    freeBlock();

    delete [] forbiddenTag;
}

void KHTMLParser::reset()
{
    current = document;

    freeBlock();

    // before parsing no tags are forbidden...
    memset(forbiddenTag, 0, (ID_CLOSE_TAG+1)*sizeof(ushort));

    inBody = false;
    haveFrameSet = false;
    _inline = false;

    form = 0;
    map = 0;
    head = 0;
    end = false;

    discard_until = 0;
}

void KHTMLParser::parseToken(Token *t)
{
    if (t->id > 2*ID_CLOSE_TAG)
    {
      //kdDebug( 6035 ) << "Unknown tag!! tagID = " << t->id << endl;
      return;
    }
    if(discard_until)
    {
        if(t->id == discard_until)
            discard_until = 0;
        return;
    }

#ifdef PARSER_DEBUG
    //kdDebug( 6035 ) << "\n\n==> parser: processing token " << t->id << " current = " << current->id() << endl;
        //kdDebug(6035) << "inline=" << _inline << " inBody=" << inBody << endl;
#endif

    if(t->id > ID_CLOSE_TAG)
    {
        processCloseTag(t);
        return;
    }

    // ignore spaces, if we're not inside a paragraph or other inline code
    if( t->id == ID_TEXT && (!_inline  || !inBody) )
    {
#ifdef PARSER_DEBUG
        //kdDebug(6035) << "length="<< t->text.length() << "text='" << t->text.string() << "'" << endl;
#endif
        if(t->text.length() == 1 && t->text[0].latin1() == ' ')
            return;
    }


    NodeImpl *n = getElement(t);
    // just to be sure, and to catch currently unimplemented stuff
    if(!n)
        return;

    // set attributes
    if(n->isElementNode())
    {
        ElementImpl *e = static_cast<ElementImpl *>(n);
        e->setAttribute(t->attrs);

        // take care of optional close tags
        if(e->endTag() == DOM::OPTIONAL)
            popBlock(t->id);
    }

    // if this tag is forbidden inside the current context, pop
    // blocks until we are allowed to add it...
    while(forbiddenTag[t->id]) popOneBlock();

    try
    {
        insertNode(n);
    }
    catch(DOMException)
    {
									
        // we couldn't insert the node...
#ifdef PARSER_DEBUG
        //kdDebug( 6035 ) << "insertNode failed current=" << current->id() << ", new=" << n->id() << "!" << endl;
#endif
        if (map == n)
        {
#ifdef PARSER_DEBUG
            //kdDebug( 6035 ) << "  --> resetting map!" << endl;
#endif
            map = 0;
        }
        if (form == n)
        {
#ifdef PARSER_DEBUG
            //kdDebug( 6035 ) << "   --> resetting form!" << endl;
#endif
            form = 0;
        }
        delete n;
    }

}

void KHTMLParser::insertNode(NodeImpl *n)
{
    int id = n->id();

    // let's be stupid and just try to insert it.
    // this should work if the document is wellformed
    try
    {
#ifdef PARSER_DEBUG
        NodeImpl *tmp = current;
#endif
        NodeImpl *newNode = current->addChild(n);

        //kdDebug(0) << "hrere" << endl;
#ifdef PARSER_DEBUG
        //kdDebug( 6035 ) << "added " << n->nodeName().string() << " to " << tmp->nodeName().string() << ", new current=" << newNode->nodeName().string() << endl;
#endif
        // don't push elements without end tag on the stack
        if(tagPriority[id] != 0)
        {
            pushBlock(id, tagPriority[id]);
            current = newNode;
            n->attach(HTMLWidget);
            // ### HACK!!!
            if(n->id() == ID_BODY)
                document->createSelector();
            if(current->isInline()) _inline = true;
        }
        else
            n->attach(HTMLWidget);

        if(tagPriority[id] == 0 && n->renderer())
        {
            n->renderer()->calcMinMaxWidth();
            if (n->id() == ID_EMBED) n->renderer()->close();
        }

    }
    catch(DOMException exception)
    {
#ifdef PARSER_DEBUG
        //kdDebug( 6035 ) << "ADDING NODE FAILED!!!! current = " << current->nodeName().string() << ", new = " << n->nodeName().string() << endl;
#endif
        // error handling...
        HTMLElementImpl *e;
        bool handled = false;

        // switch according to the element to insert
        switch(id)
        {
        case ID_COMMENT:
            break;
        case ID_HEAD:
            // ### alllow not having <HTML> in at all, as per HTML spec
            if (!current->isDocumentNode() && !current->id() == ID_HTML )