htmltokenizer.cpp

konqueror3 embedded版本, KDE环境下的当家浏览器的嵌入式版本源码包.

/*
    This file is part of the KDE libraries

    Copyright (C) 1997 Martin Jones ([e-mail])
              (C) 1997 Torben Weis ([e-mail])
              (C) 1998 Waldo Bastian ([e-mail])
              (C) 1999 Lars Knoll ([e-mail])
              (C) 1999 Antti Koivisto ([e-mail])
              (C) 2001-2003 Dirk Mueller ([e-mail])
              (C) 2002 Apple Computer, Inc.

    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., 51 Franklin Street, Fifth Floor,
    Boston, MA 02110-1301, USA.
*/
//----------------------------------------------------------------------------
//
// KDE HTML Widget - Tokenizers

//#define TOKEN_DEBUG 1
//#define TOKEN_DEBUG 2

#ifdef HAVE_CONFIG_H
#include "config.h"
#endif

#include "html/htmltokenizer.h"
#include "html/html_documentimpl.h"
#include "html/htmlparser.h"
#include "html/dtd.h"

#include "misc/loader.h"
#include "misc/htmlhashes.h"

#include "khtmlview.h"
#include "khtml_part.h"
#include "xml/dom_docimpl.h"
#include "css/csshelper.h"
#include "ecma/kjs_proxy.h"
#include <kcharsets.h>
#include <kglobal.h>
#include <ctype.h>
#include <assert.h>
#include <qvariant.h>
#include <kdebug.h>
#include <stdlib.h>

#include "kentities.c"

using namespace khtml;

static const QChar commentStart [] = { '<','!','-','-', QChar::null };

static const char scriptEnd [] = "</script";
static const char xmpEnd [] = "</xmp";
static const char styleEnd [] =  "</style";
static const char textareaEnd [] = "</textarea";
static const char titleEnd [] = "</title";

#define KHTML_ALLOC_QCHAR_VEC( N ) (QChar*) malloc( sizeof(QChar)*( N ) )
#define KHTML_REALLOC_QCHAR_VEC(P, N ) (QChar*) realloc(P, sizeof(QChar)*( N ))
#define KHTML_DELETE_QCHAR_VEC( P ) free((char*)( P ))

// Full support for MS Windows extensions to Latin-1.
// Technically these extensions should only be activated for pages
// marked "windows-1252" or "cp1252", but
// in the standard Microsoft way, these extensions infect hundreds of thousands
// of web pages.  Note that people with non-latin-1 Microsoft extensions
// are SOL.
//
// See: http://www.microsoft.com/globaldev/reference/WinCP.asp
//      http://www.bbsinc.com/iso8859.html
//      http://www.obviously.com/
//
// There may be better equivalents
#if 0
#define fixUpChar(x)
#else
#define fixUpChar(x) \
            switch ((x).unicode()) \
            { \
            /* ALL of these should be changed to Unicode SOON */ \
            case 0x80: (x) = 0x20ac; break; \
            case 0x82: (x) = ',';    break; \
            case 0x83: (x) = 0x0192; break; \
            case 0x84: (x) = '"';    break; \
            case 0x85: (x) = 0x2026; break; \
            case 0x86: (x) = 0x2020; break; \
            case 0x87: (x) = 0x2021; break; \
            case 0x88: (x) = 0x02C6; break; \
            case 0x89: (x) = 0x2030; break; \
            case 0x8A: (x) = 0x0160; break; \
            case 0x8b: (x) = '<';    break; \
            case 0x8C: (x) = 0x0152; break; \
            case 0x8E: (x) = 0x017D; break; \
            case 0x91: (x) = '\'';   break; \
            case 0x92: (x) = '\'';   break; \
            case 0x93: (x) = '"';    break; \
            case 0x94: (x) = '"';    break; \
            case 0x95: (x) = '*';    break; \
            case 0x96: (x) = '-';    break; \
            case 0x97: (x) = '-';    break; \
            case 0x98: (x) = '~';    break; \
            case 0x99: (x) = 0x2122; break; \
            case 0x9A: (x) = 0x0161; break; \
            case 0x9b: (x) = '>';    break; \
            case 0x9C: (x) = 0x0153; break; \
            case 0x9E: (x) = 0x017E; break; \
            case 0x9F: (x) = 0x0178; break; \
            default: break; \
            }
#endif

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

HTMLTokenizer::HTMLTokenizer(DOM::DocumentPtr *_doc, KHTMLView *_view)
{
    view = _view;
    buffer = 0;
    scriptCode = 0;
    scriptCodeSize = scriptCodeMaxSize = scriptCodeResync = 0;
    charsets = KGlobal::charsets();
    parser = new KHTMLParser(_view, _doc);
    m_executingScript = 0;
    m_autoCloseTimer = 0;
    onHold = false;

    reset();
}

HTMLTokenizer::HTMLTokenizer(DOM::DocumentPtr *_doc, DOM::DocumentFragmentImpl *i)
{
    view = 0;
    buffer = 0;
    scriptCode = 0;
    scriptCodeSize = scriptCodeMaxSize = scriptCodeResync = 0;
    charsets = KGlobal::charsets();
    parser = new KHTMLParser( i, _doc );
    m_executingScript = 0;
    m_autoCloseTimer = 0;
    onHold = false;

    reset();
}

void HTMLTokenizer::reset()
{
    assert(m_executingScript == 0);
    Q_ASSERT(onHold == false);
    m_abort = false;

    while (!cachedScript.isEmpty())
        cachedScript.dequeue()->deref(this);

    if ( buffer )
        KHTML_DELETE_QCHAR_VEC(buffer);
    buffer = dest = 0;
    size = 0;

    if ( scriptCode )
        KHTML_DELETE_QCHAR_VEC(scriptCode);
    scriptCode = 0;
    scriptCodeSize = scriptCodeMaxSize = scriptCodeResync = 0;

    if (m_autoCloseTimer) {
        killTimer(m_autoCloseTimer);
        m_autoCloseTimer = 0;
    }

    currToken.reset();
}

void HTMLTokenizer::begin()
{
    m_executingScript = 0;
    onHold = false;
    reset();
    size = 254;
    buffer = KHTML_ALLOC_QCHAR_VEC( 255 );
    dest = buffer;
    tag = NoTag;
    pending = NonePending;
    discard = NoneDiscard;
    pre = false;
    prePos = 0;
    plaintext = false;
    xmp = false;
    processingInstruction = false;
    script = false;
    escaped = false;
    style = false;
    skipLF = false;
    select = false;
    comment = false;
    server = false;
    textarea = false;
    title = false;
    startTag = false;
    tquote = NoQuote;
    searchCount = 0;
    Entity = NoEntity;
    noMoreData = false;
    brokenComments = false;
    brokenServer = false;
    brokenScript = false;
    lineno = 0;
    scriptStartLineno = 0;
    tagStartLineno = 0;
}

void HTMLTokenizer::processListing(TokenizerString list)
{
    bool old_pre = pre;

    // This function adds the listing 'list' as
    // preformatted text-tokens to the token-collection
    // thereby converting TABs.
    if(!style) pre = true;
    prePos = 0;

    while ( !list.isEmpty() )
    {
        checkBuffer(3*TAB_SIZE);

        if (skipLF && ( *list != '\n' ))
        {
            skipLF = false;
        }

        if (skipLF)
        {
            skipLF = false;
            ++list;
        }
        else if (( *list == '\n' ) || ( *list == '\r' ))
        {
            if (discard == LFDiscard)
            {
                // Ignore this LF
                discard = NoneDiscard; // We have discarded 1 LF
            }
            else
            {
                // Process this LF
                if (pending)
                    addPending();

                // we used to do it not at all and we want to have
                // it fixed for textarea. So here we are
                if ( textarea ) {
                    prePos++;
                    *dest++ = *list;
                } else
                    pending = LFPending;
            }
            /* Check for MS-DOS CRLF sequence */
            if (*list == '\r')
            {
                skipLF = true;
            }
            ++list;
        }
        else if (( *list == ' ' ) || ( *list == '\t'))
        {
            if (pending)
                addPending();
            if (*list == ' ')
                pending = SpacePending;
            else
                pending = TabPending;

            ++list;
        }
        else
        {
            discard = NoneDiscard;
            if (pending)
                addPending();

            prePos++;
            *dest++ = *list;
            ++list;
        }

    }

    if ((pending == SpacePending) || (pending == TabPending))
        addPending();
    else
        pending = NonePending;

    prePos = 0;
    pre = old_pre;
}

void HTMLTokenizer::parseSpecial(TokenizerString &src)
{
    assert( textarea || title || !Entity );
    assert( !tag );
    assert( xmp+textarea+title+style+script == 1 );
    if (script)
        scriptStartLineno = lineno+src.lineCount();

    if ( comment ) parseComment( src );

    while ( !src.isEmpty() ) {
        checkScriptBuffer();
        unsigned char ch = src->latin1();
        if ( !scriptCodeResync && !brokenComments && !textarea && !xmp && !title && ch == '-' && scriptCodeSize >= 3 && !src.escaped() && QConstString( scriptCode+scriptCodeSize-3, 3 ).string() == "<!-" ) {
            comment = true;
            scriptCode[ scriptCodeSize++ ] = ch;
            ++src;
            parseComment( src );
            continue;
        }
        if ( scriptCodeResync && !tquote && ( ch == '>' ) ) {
            ++src;
            scriptCodeSize = scriptCodeResync-1;
            scriptCodeResync = 0;
            scriptCode[ scriptCodeSize ] = scriptCode[ scriptCodeSize + 1 ] = 0;
            if ( script )
                scriptHandler();
            else {
                processListing(TokenizerString(scriptCode, scriptCodeSize));
                processToken();
                if ( style )         { currToken.tid = ID_STYLE + ID_CLOSE_TAG; }
                else if ( textarea ) { currToken.tid = ID_TEXTAREA + ID_CLOSE_TAG; }
                else if ( title ) { currToken.tid = ID_TITLE + ID_CLOSE_TAG; }
                else if ( xmp )  { currToken.tid = ID_XMP + ID_CLOSE_TAG; }
                processToken();
                script = style = textarea = title = xmp = false;
                tquote = NoQuote;
                scriptCodeSize = scriptCodeResync = 0;
            }
            return;
        }
        // possible end of tagname, lets check.
        if ( !scriptCodeResync && !escaped && !src.escaped() && ( ch == '>' || ch == '/' || ch <= ' ' ) && ch &&
             scriptCodeSize >= searchStopperLen &&
             !QConstString( scriptCode+scriptCodeSize-searchStopperLen, searchStopperLen ).string().find( searchStopper, 0, false )) {
            scriptCodeResync = scriptCodeSize-searchStopperLen+1;
            tquote = NoQuote;
            continue;
        }
        if ( scriptCodeResync && !escaped ) {
            if(ch == '\"')
                tquote = (tquote == NoQuote) ? DoubleQuote : ((tquote == SingleQuote) ? SingleQuote : NoQuote);
            else if(ch == '\'')
                tquote = (tquote == NoQuote) ? SingleQuote : (tquote == DoubleQuote) ? DoubleQuote : NoQuote;
            else if (tquote != NoQuote && (ch == '\r' || ch == '\n'))
                tquote = NoQuote;
        }
        escaped = ( !escaped && ch == '\\' );
        if (!scriptCodeResync && (textarea||title) && !src.escaped() && ch == '&') {
            QChar *scriptCodeDest = scriptCode+scriptCodeSize;
            ++src;
            parseEntity(src,scriptCodeDest,true);
            scriptCodeSize = scriptCodeDest-scriptCode;
        }
        else {
            scriptCode[ scriptCodeSize++ ] = *src;
            ++src;
        }
    }
}

void HTMLTokenizer::scriptHandler()
{
    QString currentScriptSrc = scriptSrc;
    scriptSrc = QString::null;

    processListing(TokenizerString(scriptCode, scriptCodeSize));
    QString exScript( buffer, dest-buffer );

    processToken();
    currToken.tid = ID_SCRIPT + ID_CLOSE_TAG;
    processToken();

    TokenizerString prependingSrc;

    if ( !parser->skipMode() ) {
        CachedScript* cs = 0;

        // forget what we just got, load from src url instead
        if ( !currentScriptSrc.isEmpty() &&
             (cs = parser->doc()->docLoader()->requestScript(currentScriptSrc, scriptSrcCharset) ))
            cachedScript.enqueue(cs);

        if (cs) {
            pendingSrc.prepend(src);
            setSrc(TokenizerString());
            scriptCodeSize = scriptCodeResync = 0;
            cs->ref(this);
        }
        else if (currentScriptSrc.isEmpty() && view && javascript ) {
            if ( !m_executingScript )
                pendingSrc.prepend(src);
            else
                prependingSrc = src;
            setSrc(TokenizerString());
            scriptCodeSize = scriptCodeResync = 0;
            scriptExecution( exScript, QString::null, tagStartLineno /*scriptStartLineno*/ );
        }
    }

    script = false;
    scriptCodeSize = scriptCodeResync = 0;

    if ( !m_executingScript && cachedScript.isEmpty() ) {
        // kdDebug( 6036 ) << "adding pending Output to parsed string" << endl;
        src.append(pendingSrc);
        pendingSrc.clear();
    } else if ( !prependingSrc.isEmpty() )
        write( prependingSrc, false );
}

void HTMLTokenizer::scriptExecution( const QString& str, const QString& scriptURL,
                                     int baseLine)
{
    bool oldscript = script;
    m_executingScript++;
    script = false;
    QString url;
    if (scriptURL.isNull() && view)
      url = static_cast<DocumentImpl*>(view->part()->document().handle())->URL().url();
    else
      url = scriptURL;

    if (view)
	view->part()->executeScript(url,baseLine+1,Node(),str);