/****************************************************************************

**

** Copyright (C) 2009 Nokia Corporation and/or its subsidiary(-ies).

** All rights reserved.

** Contact: Nokia Corporation (qt-info@nokia.com)

**

** This file is part of the QtGui module of the Qt Toolkit.

**

** $QT_BEGIN_LICENSE:LGPL$

** No Commercial Usage

** This file contains pre-release code and may not be distributed.

** You may use this file in accordance with the terms and conditions

** contained in the Technology Preview License Agreement accompanying

** this package.

**

** GNU Lesser General Public License Usage

** Alternatively, this file may be used under the terms of the GNU Lesser

** General Public License version 2.1 as published by the Free Software

** Foundation and appearing in the file LICENSE.LGPL included in the

** packaging of this file.  Please review the following information to

** ensure the GNU Lesser General Public License version 2.1 requirements

** will be met: http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.

**

** In addition, as a special exception, Nokia gives you certain additional

** rights.  These rights are described in the Nokia Qt LGPL Exception

** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.

**

** If you have questions regarding the use of this file, please contact

** Nokia at qt-info@nokia.com.

**

**

**

**

**

**

**

**

** $QT_END_LICENSE$

**

****************************************************************************/

#include "qtexthtmlparser_p.h"

#include <qbytearray.h>

#include <qtextcodec.h>

#include <qapplication.h>

#include <qstack.h>

#include <qdebug.h>

#include <qthread.h>

#include "qtextdocument.h"

#include "qtextformat_p.h"

#include "qtextdocument_p.h"

#include "qtextcursor.h"

#include "qfont_p.h"

#include "private/qunicodetables_p.h"

#include "private/qfunctions_p.h"

#ifndef QT_NO_TEXTHTMLPARSER

QT_BEGIN_NAMESPACE

// see also tst_qtextdocumentfragment.cpp

#define MAX_ENTITY 258

static const struct QTextHtmlEntity { const char *name; quint16 code; } entities[MAX_ENTITY]= {

    { "AElig", 0x00c6 },

    { "AMP", 38 },

    { "Aacute", 0x00c1 },

    { "Acirc", 0x00c2 },

    { "Agrave", 0x00c0 },

    { "Alpha", 0x0391 },

    { "Aring", 0x00c5 },

    { "Atilde", 0x00c3 },

    { "Auml", 0x00c4 },

    { "Beta", 0x0392 },

    { "Ccedil", 0x00c7 },

    { "Chi", 0x03a7 },

    { "Dagger", 0x2021 },

    { "Delta", 0x0394 },

    { "ETH", 0x00d0 },

    { "Eacute", 0x00c9 },

    { "Ecirc", 0x00ca },

    { "Egrave", 0x00c8 },

    { "Epsilon", 0x0395 },

    { "Eta", 0x0397 },

    { "Euml", 0x00cb },

    { "GT", 62 },

    { "Gamma", 0x0393 },

    { "Iacute", 0x00cd },

    { "Icirc", 0x00ce },

    { "Igrave", 0x00cc },

    { "Iota", 0x0399 },

    { "Iuml", 0x00cf },

    { "Kappa", 0x039a },

    { "LT", 60 },

    { "Lambda", 0x039b },

    { "Mu", 0x039c },

    { "Ntilde", 0x00d1 },

    { "Nu", 0x039d },

    { "OElig", 0x0152 },

    { "Oacute", 0x00d3 },

    { "Ocirc", 0x00d4 },

    { "Ograve", 0x00d2 },

    { "Omega", 0x03a9 },

    { "Omicron", 0x039f },

    { "Oslash", 0x00d8 },

    { "Otilde", 0x00d5 },

    { "Ouml", 0x00d6 },

    { "Phi", 0x03a6 },

    { "Pi", 0x03a0 },

    { "Prime", 0x2033 },

    { "Psi", 0x03a8 },

    { "QUOT", 34 },

    { "Rho", 0x03a1 },

    { "Scaron", 0x0160 },

    { "Sigma", 0x03a3 },

    { "THORN", 0x00de },

    { "Tau", 0x03a4 },

    { "Theta", 0x0398 },

    { "Uacute", 0x00da },

    { "Ucirc", 0x00db },

    { "Ugrave", 0x00d9 },

    { "Upsilon", 0x03a5 },

    { "Uuml", 0x00dc },

    { "Xi", 0x039e },

    { "Yacute", 0x00dd },

    { "Yuml", 0x0178 },

    { "Zeta", 0x0396 },

    { "aacute", 0x00e1 },

    { "acirc", 0x00e2 },

    { "acute", 0x00b4 },

    { "aelig", 0x00e6 },

    { "agrave", 0x00e0 },

    { "alefsym", 0x2135 },

    { "alpha", 0x03b1 },

    { "amp", 38 },

    { "and", 0x22a5 },

    { "ang", 0x2220 },

    { "apos", 0x0027 },

    { "aring", 0x00e5 },

    { "asymp", 0x2248 },

    { "atilde", 0x00e3 },

    { "auml", 0x00e4 },

    { "bdquo", 0x201e },

    { "beta", 0x03b2 },

    { "brvbar", 0x00a6 },

    { "bull", 0x2022 },

    { "cap", 0x2229 },

    { "ccedil", 0x00e7 },

    { "cedil", 0x00b8 },

    { "cent", 0x00a2 },

    { "chi", 0x03c7 },

    { "circ", 0x02c6 },

    { "clubs", 0x2663 },

    { "cong", 0x2245 },

    { "copy", 0x00a9 },

    { "crarr", 0x21b5 },

    { "cup", 0x222a },

    { "curren", 0x00a4 },

    { "dArr", 0x21d3 },

    { "dagger", 0x2020 },

    { "darr", 0x2193 },

    { "deg", 0x00b0 },

    { "delta", 0x03b4 },

    { "diams", 0x2666 },

    { "divide", 0x00f7 },

    { "eacute", 0x00e9 },

    { "ecirc", 0x00ea },

    { "egrave", 0x00e8 },

    { "empty", 0x2205 },

    { "emsp", 0x2003 },

    { "ensp", 0x2002 },

    { "epsilon", 0x03b5 },

    { "equiv", 0x2261 },

    { "eta", 0x03b7 },

    { "eth", 0x00f0 },

    { "euml", 0x00eb },

    { "euro", 0x20ac },

    { "exist", 0x2203 },

    { "fnof", 0x0192 },

    { "forall", 0x2200 },

    { "frac12", 0x00bd },

    { "frac14", 0x00bc },

    { "frac34", 0x00be },

    { "frasl", 0x2044 },

    { "gamma", 0x03b3 },

    { "ge", 0x2265 },

    { "gt", 62 },

    { "hArr", 0x21d4 },

    { "harr", 0x2194 },

    { "hearts", 0x2665 },

    { "hellip", 0x2026 },

    { "iacute", 0x00ed },

    { "icirc", 0x00ee },

    { "iexcl", 0x00a1 },

    { "igrave", 0x00ec },

    { "image", 0x2111 },

    { "infin", 0x221e },

    { "int", 0x222b },

    { "iota", 0x03b9 },

    { "iquest", 0x00bf },

    { "isin", 0x2208 },

    { "iuml", 0x00ef },

    { "kappa", 0x03ba },

    { "lArr", 0x21d0 },

    { "lambda", 0x03bb },

    { "lang", 0x2329 },

    { "laquo", 0x00ab },

    { "larr", 0x2190 },

    { "lceil", 0x2308 },

    { "ldquo", 0x201c },

    { "le", 0x2264 },

    { "lfloor", 0x230a },

    { "lowast", 0x2217 },

    { "loz", 0x25ca },

    { "lrm", 0x200e },

    { "lsaquo", 0x2039 },

    { "lsquo", 0x2018 },

    { "lt", 60 },

    { "macr", 0x00af },

    { "mdash", 0x2014 },

    { "micro", 0x00b5 },

    { "middot", 0x00b7 },

    { "minus", 0x2212 },

    { "mu", 0x03bc },

    { "nabla", 0x2207 },

    { "nbsp", 0x00a0 },

    { "ndash", 0x2013 },

    { "ne", 0x2260 },

    { "ni", 0x220b },

    { "not", 0x00ac },

    { "notin", 0x2209 },

    { "nsub", 0x2284 },

    { "ntilde", 0x00f1 },

    { "nu", 0x03bd },

    { "oacute", 0x00f3 },

    { "ocirc", 0x00f4 },

    { "oelig", 0x0153 },

    { "ograve", 0x00f2 },

    { "oline", 0x203e },

    { "omega", 0x03c9 },

    { "omicron", 0x03bf },

    { "oplus", 0x2295 },

    { "or", 0x22a6 },

    { "ordf", 0x00aa },

    { "ordm", 0x00ba },

    { "oslash", 0x00f8 },

    { "otilde", 0x00f5 },

    { "otimes", 0x2297 },

    { "ouml", 0x00f6 },

    { "para", 0x00b6 },

    { "part", 0x2202 },

    { "percnt", 0x0025 },

    { "permil", 0x2030 },

    { "perp", 0x22a5 },

    { "phi", 0x03c6 },

    { "pi", 0x03c0 },

    { "piv", 0x03d6 },

    { "plusmn", 0x00b1 },

    { "pound", 0x00a3 },

    { "prime", 0x2032 },

    { "prod", 0x220f },

    { "prop", 0x221d },

    { "psi", 0x03c8 },

    { "quot", 34 },

    { "rArr", 0x21d2 },

    { "radic", 0x221a },

    { "rang", 0x232a },

    { "raquo", 0x00bb },

    { "rarr", 0x2192 },

    { "rceil", 0x2309 },

    { "rdquo", 0x201d },

    { "real", 0x211c },

    { "reg", 0x00ae },

    { "rfloor", 0x230b },

    { "rho", 0x03c1 },

    { "rlm", 0x200f },

    { "rsaquo", 0x203a },

    { "rsquo", 0x2019 },

    { "sbquo", 0x201a },

    { "scaron", 0x0161 },

    { "sdot", 0x22c5 },

    { "sect", 0x00a7 },

    { "shy", 0x00ad },

    { "sigma", 0x03c3 },

    { "sigmaf", 0x03c2 },

    { "sim", 0x223c },

    { "spades", 0x2660 },

    { "sub", 0x2282 },

    { "sube", 0x2286 },

    { "sum", 0x2211 },

    { "sup", 0x2283 },

    { "sup1", 0x00b9 },

    { "sup2", 0x00b2 },

    { "sup3", 0x00b3 },

    { "supe", 0x2287 },

    { "szlig", 0x00df },

    { "tau", 0x03c4 },

    { "there4", 0x2234 },

    { "theta", 0x03b8 },

    { "thetasym", 0x03d1 },

    { "thinsp", 0x2009 },

    { "thorn", 0x00fe },

    { "tilde", 0x02dc },

    { "times", 0x00d7 },

    { "trade", 0x2122 },

    { "uArr", 0x21d1 },

    { "uacute", 0x00fa },

    { "uarr", 0x2191 },

    { "ucirc", 0x00fb },

    { "ugrave", 0x00f9 },

    { "uml", 0x00a8 },

    { "upsih", 0x03d2 },

    { "upsilon", 0x03c5 },

    { "uuml", 0x00fc },

    { "weierp", 0x2118 },

    { "xi", 0x03be },

    { "yacute", 0x00fd },

    { "yen", 0x00a5 },

    { "yuml", 0x00ff },

    { "zeta", 0x03b6 },

    { "zwj", 0x200d },

    { "zwnj", 0x200c }

};

Q_STATIC_GLOBAL_OPERATOR bool operator<(const QString &entityStr, const QTextHtmlEntity &entity)

{

    return entityStr < QLatin1String(entity.name);

}

Q_STATIC_GLOBAL_OPERATOR bool operator<(const QTextHtmlEntity &entity, const QString &entityStr)

{

    return QLatin1String(entity.name) < entityStr;

}

static QChar resolveEntity(const QString &entity)

{

    const QTextHtmlEntity *start = &entities[0];

    const QTextHtmlEntity *end = &entities[MAX_ENTITY];

    const QTextHtmlEntity *e = qBinaryFind(start, end, entity);

    if (e == end)

        return QChar();

    return e->code;

}

static const ushort windowsLatin1ExtendedCharacters[0xA0 - 0x80] = {

    0x20ac, // 0x80

    0x0081, // 0x81 direct mapping

    0x201a, // 0x82

    0x0192, // 0x83

    0x201e, // 0x84

    0x2026, // 0x85

    0x2020, // 0x86

    0x2021, // 0x87

    0x02C6, // 0x88

    0x2030, // 0x89

    0x0160, // 0x8A

    0x2039, // 0x8B

    0x0152, // 0x8C

    0x008D, // 0x8D direct mapping

    0x017D, // 0x8E

    0x008F, // 0x8F directmapping

    0x0090, // 0x90 directmapping

    0x2018, // 0x91

    0x2019, // 0x92

    0x201C, // 0x93

    0X201D, // 0x94

    0x2022, // 0x95

    0x2013, // 0x96

    0x2014, // 0x97

    0x02DC, // 0x98

    0x2122, // 0x99

    0x0161, // 0x9A

    0x203A, // 0x9B

    0x0153, // 0x9C

    0x009D, // 0x9D direct mapping

    0x017E, // 0x9E

    0x0178  // 0x9F

};

// the displayMode value is according to the what are blocks in the piecetable, not

// what the w3c defines.

static const QTextHtmlElement elements[Html_NumElements]= {

    { "a",          Html_a,          QTextHtmlElement::DisplayInline },

    { "address",    Html_address,    QTextHtmlElement::DisplayInline },

    { "b",          Html_b,          QTextHtmlElement::DisplayInline },

    { "big",        Html_big,        QTextHtmlElement::DisplayInline },

    { "blockquote", Html_blockquote, QTextHtmlElement::DisplayBlock },

    { "body",       Html_body,       QTextHtmlElement::DisplayBlock },

    { "br",         Html_br,         QTextHtmlElement::DisplayInline },

    { "caption",    Html_caption,    QTextHtmlElement::DisplayBlock },

    { "center",     Html_center,     QTextHtmlElement::DisplayBlock },

    { "cite",       Html_cite,       QTextHtmlElement::DisplayInline },

    { "code",       Html_code,       QTextHtmlElement::DisplayInline },

    { "dd",         Html_dd,         QTextHtmlElement::DisplayBlock },

    { "dfn",        Html_dfn,        QTextHtmlElement::DisplayInline },

    { "div",        Html_div,        QTextHtmlElement::DisplayBlock },

    { "dl",         Html_dl,         QTextHtmlElement::DisplayBlock },

    { "dt",         Html_dt,         QTextHtmlElement::DisplayBlock },

    { "em",         Html_em,         QTextHtmlElement::DisplayInline },

    { "font",       Html_font,       QTextHtmlElement::DisplayInline },

    { "h1",         Html_h1,         QTextHtmlElement::DisplayBlock },

    { "h2",         Html_h2,         QTextHtmlElement::DisplayBlock },

    { "h3",         Html_h3,         QTextHtmlElement::DisplayBlock },

    { "h4",         Html_h4,         QTextHtmlElement::DisplayBlock },

    { "h5",         Html_h5,         QTextHtmlElement::DisplayBlock },

    { "h6",         Html_h6,         QTextHtmlElement::DisplayBlock },

    { "head",       Html_head,       QTextHtmlElement::DisplayNone },

    { "hr",         Html_hr,         QTextHtmlElement::DisplayBlock },

    { "html",       Html_html,       QTextHtmlElement::DisplayInline },

    { "i",          Html_i,          QTextHtmlElement::DisplayInline },

    { "img",        Html_img,        QTextHtmlElement::DisplayInline },

    { "kbd",        Html_kbd,        QTextHtmlElement::DisplayInline },

    { "li",         Html_li,         QTextHtmlElement::DisplayBlock },

    { "link",       Html_link,       QTextHtmlElement::DisplayNone },

    { "meta",       Html_meta,       QTextHtmlElement::DisplayNone },

    { "nobr",       Html_nobr,       QTextHtmlElement::DisplayInline },

    { "ol",         Html_ol,         QTextHtmlElement::DisplayBlock },

    { "p",          Html_p,          QTextHtmlElement::DisplayBlock },

    { "pre",        Html_pre,        QTextHtmlElement::DisplayBlock },

    { "qt",         Html_body /*deliberate mapping*/, QTextHtmlElement::DisplayBlock },

    { "s",          Html_s,          QTextHtmlElement::DisplayInline },

    { "samp",       Html_samp,       QTextHtmlElement::DisplayInline },

    { "script",     Html_script,     QTextHtmlElement::DisplayNone },

    { "small",      Html_small,      QTextHtmlElement::DisplayInline },

    { "span",       Html_span,       QTextHtmlElement::DisplayInline },

    { "strong",     Html_strong,     QTextHtmlElement::DisplayInline },

    { "style",      Html_style,      QTextHtmlElement::DisplayNone },

    { "sub",        Html_sub,        QTextHtmlElement::DisplayInline },

    { "sup",        Html_sup,        QTextHtmlElement::DisplayInline },

    { "table",      Html_table,      QTextHtmlElement::DisplayTable },

    { "tbody",      Html_tbody,      QTextHtmlElement::DisplayTable },

    { "td",         Html_td,         QTextHtmlElement::DisplayBlock },

    { "tfoot",      Html_tfoot,      QTextHtmlElement::DisplayTable },

    { "th",         Html_th,         QTextHtmlElement::DisplayBlock },

    { "thead",      Html_thead,      QTextHtmlElement::DisplayTable },

    { "title",      Html_title,      QTextHtmlElement::DisplayNone },

    { "tr",         Html_tr,         QTextHtmlElement::DisplayTable },

    { "tt",         Html_tt,         QTextHtmlElement::DisplayInline },

    { "u",          Html_u,          QTextHtmlElement::DisplayInline },

    { "ul",         Html_ul,         QTextHtmlElement::DisplayBlock },

    { "var",        Html_var,        QTextHtmlElement::DisplayInline },

};

Q_STATIC_GLOBAL_OPERATOR bool operator<(const QString &str, const QTextHtmlElement &e)

{

    return str < QLatin1String(e.name);

}

Q_STATIC_GLOBAL_OPERATOR bool operator<(const QTextHtmlElement &e, const QString &str)

{

    return QLatin1String(e.name) < str;

}

static const QTextHtmlElement *lookupElementHelper(const QString &element)

{

    const QTextHtmlElement *start = &elements[0];

    const QTextHtmlElement *end = &elements[Html_NumElements];

    const QTextHtmlElement *e = qBinaryFind(start, end, element);

    if (e == end)

        return 0;

    return e;

}

int QTextHtmlParser::lookupElement(const QString &element)

{

    const QTextHtmlElement *e = lookupElementHelper(element);

    if (!e)

        return -1;

    return e->id;

}

// quotes newlines as "\\n"

static QString quoteNewline(const QString &s)

{

    QString n = s;

    n.replace(QLatin1Char('\n'), QLatin1String("\\n"));

    return n;

}

QTextHtmlParserNode::QTextHtmlParserNode()

    : parent(0), id(Html_unknown),

      cssFloat(QTextFrameFormat::InFlow), hasOwnListStyle(false),

      hasCssListIndent(false), isEmptyParagraph(false), isTextFrame(false), isRootFrame(false),

      displayMode(QTextHtmlElement::DisplayInline), hasHref(false),

      listStyle(QTextListFormat::ListStyleUndefined), imageWidth(-1), imageHeight(-1), tableBorder(0),

      tableCellRowSpan(1), tableCellColSpan(1), tableCellSpacing(2), tableCellPadding(0),

      borderBrush(Qt::darkGray), borderStyle(QTextFrameFormat::BorderStyle_Outset),

      userState(-1), cssListIndent(0), wsm(WhiteSpaceModeUndefined)

{

    margin[QTextHtmlParser::MarginLeft] = 0;

    margin[QTextHtmlParser::MarginRight] = 0;

    margin[QTextHtmlParser::MarginTop] = 0;

    margin[QTextHtmlParser::MarginBottom] = 0;

}

void QTextHtmlParser::dumpHtml()

{

    for (int i = 0; i < count(); ++i) {

        qDebug().nospace() << qPrintable(QString(depth(i)*4, QLatin1Char(' ')))

                           << qPrintable(at(i).tag) << ':'

                           << quoteNewline(at(i).text);

            ;

    }

}

QTextHtmlParserNode *QTextHtmlParser::newNode(int parent)

{

    QTextHtmlParserNode *lastNode = &nodes.last();

    QTextHtmlParserNode *newNode = 0;

    bool reuseLastNode = true;

    if (nodes.count() == 1) {

        reuseLastNode = false;

    } else if (lastNode->tag.isEmpty()) {

        if (lastNode->text.isEmpty()) {

            reuseLastNode = true;

        } else { // last node is a text node (empty tag) with some text

            if (lastNode->text.length() == 1 && lastNode->text.at(0).isSpace()) {

                int lastSibling = count() - 2;

                while (lastSibling

                       && at(lastSibling).parent != lastNode->parent

                       && at(lastSibling).displayMode == QTextHtmlElement::DisplayInline) {