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

**

** 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 Qt Linguist 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 "translator.h"

#include <QtCore/QDebug>

#include <QtCore/QMap>

#include <QtCore/QStack>

#include <QtCore/QString>

#include <QtCore/QTextCodec>

#include <QtCore/QTextStream>

#include <QtXml/QXmlAttributes>

#include <QtXml/QXmlDefaultHandler>

#include <QtXml/QXmlParseException>

QT_BEGIN_NAMESPACE

/**

 * Implementation of XLIFF file format for Linguist

 */

//static const char *restypeDomain = "x-gettext-domain";

static const char *restypeContext = "x-trolltech-linguist-context";

static const char *restypePlurals = "x-gettext-plurals";

static const char *restypeDummy = "x-dummy";

static const char *dataTypeUIFile = "x-trolltech-designer-ui";

static const char *contextMsgctxt = "x-gettext-msgctxt"; // XXX Troll invention, so far.

static const char *contextOldMsgctxt = "x-gettext-previous-msgctxt"; // XXX Troll invention, so far.

static const char *attribPlural = "trolltech:plural";

static const char *XLIFF11namespaceURI = "urn:oasis:names:tc:xliff:document:1.1";

static const char *XLIFF12namespaceURI = "urn:oasis:names:tc:xliff:document:1.2";

static const char *TrollTsNamespaceURI = "urn:trolltech:names:ts:document:1.0";

#define COMBINE4CHARS(c1, c2, c3, c4) \

    (int(c1) << 24 | int(c2) << 16 | int(c3) << 8 | int(c4) )

static QString dataType(const TranslatorMessage &m)

{

    QByteArray fileName = m.fileName().toAscii();

    unsigned int extHash = 0;

    int pos = fileName.count() - 1;

    for (int pass = 0; pass < 4 && pos >=0; ++pass, --pos) {

        if (fileName.at(pos) == '.')

            break;

        extHash |= ((int)fileName.at(pos) << (8*pass));

    }

    switch (extHash) {

        case COMBINE4CHARS(0,'c','p','p'):

        case COMBINE4CHARS(0,'c','x','x'):

        case COMBINE4CHARS(0,'c','+','+'):

        case COMBINE4CHARS(0,'h','p','p'):

        case COMBINE4CHARS(0,'h','x','x'):

        case COMBINE4CHARS(0,'h','+','+'):

            return QLatin1String("cpp");

        case COMBINE4CHARS(0, 0 , 0 ,'c'):

        case COMBINE4CHARS(0, 0 , 0 ,'h'):

        case COMBINE4CHARS(0, 0 ,'c','c'):

        case COMBINE4CHARS(0, 0 ,'c','h'):

        case COMBINE4CHARS(0, 0 ,'h','h'):

            return QLatin1String("c");

        case COMBINE4CHARS(0, 0 ,'u','i'):

            return QLatin1String(dataTypeUIFile);   //### form?

        default:

            return QLatin1String("plaintext");      // we give up

    }

}

static void writeIndent(QTextStream &ts, int indent)

{

    ts << QString().fill(QLatin1Char(' '), indent * 2);

}

struct CharMnemonic

{

    char ch;

    char escape;

    const char *mnemonic;

};

static const CharMnemonic charCodeMnemonics[] = {

    {0x07, 'a', "bel"},

    {0x08, 'b', "bs"},

    {0x09, 't', "tab"},

    {0x0a, 'n', "lf"},

    {0x0b, 'v', "vt"},

    {0x0c, 'f', "ff"},

    {0x0d, 'r', "cr"}

};

static char charFromEscape(char escape)

{

    for (uint i = 0; i < sizeof(charCodeMnemonics)/sizeof(CharMnemonic); ++i) {

        CharMnemonic cm =  charCodeMnemonics[i];

        if (cm.escape == escape)

            return cm.ch;

    }

    Q_ASSERT(0);

    return escape;

}

static QString numericEntity(int ch, bool makePhs)

{

    // ### This needs to be reviewed, to reflect the updated XLIFF-PO spec.

    if (!makePhs || ch < 7 || ch > 0x0d)

        return QString::fromAscii("&#x%1;").arg(QString::number(ch, 16));

    CharMnemonic cm = charCodeMnemonics[int(ch) - 7];

    QString name = QLatin1String(cm.mnemonic);

    char escapechar = cm.escape;

    static int id = 0;

    return QString::fromAscii("<ph id=\"ph%1\" ctype=\"x-ch-%2\">\\%3</ph>")

              .arg(++id) .arg(name) .arg(escapechar);

}

static QString protect(const QString &str, bool makePhs = true)

{

    QString result;

    int len = str.size();

    for (int i = 0; i != len; ++i) {

        uint c = str.at(i).unicode();

        switch (c) {

        case '\"':

            result += QLatin1String(""");

            break;

        case '&':

            result += QLatin1String("&");

            break;

        case '>':

            result += QLatin1String(">");

            break;

        case '<':

            result += QLatin1String("<");

            break;

        case '\'':

            result += QLatin1String("'");

            break;

        default:

            if (c < 0x20 && c != '\r' && c != '\n' && c != '\t')

                result += numericEntity(c, makePhs);

            else // this also covers surrogates

                result += QChar(c);

        }

    }

    return result;

}

static void writeExtras(QTextStream &ts, int indent,

                        const TranslatorMessage::ExtraData &extras, const QRegExp &drops)

{

    for (Translator::ExtraData::ConstIterator it = extras.begin(); it != extras.end(); ++it) {

        if (!drops.exactMatch(it.key())) {

            writeIndent(ts, indent);

            ts << "<trolltech:" << it.key() << '>'

               << protect(it.value())

               << "</trolltech:" << it.key() << ">\n";

        }

    }

}

static void writeLineNumber(QTextStream &ts, const TranslatorMessage &msg, int indent)

{

    if (msg.lineNumber() == -1)

        return;

    writeIndent(ts, indent);

    ts << "<context-group purpose=\"location\"><context context-type=\"linenumber\">"

       << msg.lineNumber() << "</context></context-group>\n";

    foreach (const TranslatorMessage::Reference &ref, msg.extraReferences()) {

        writeIndent(ts, indent);

        ts << "<context-group purpose=\"location\">";

        if (ref.fileName() != msg.fileName())

            ts << "<context context-type=\"sourcefile\">" << ref.fileName() << "</context>";

        ts << "<context context-type=\"linenumber\">" << ref.lineNumber()

           << "</context></context-group>\n";

    }

}

static void writeComment(QTextStream &ts, const TranslatorMessage &msg, const QRegExp &drops, int indent)

{

    if (!msg.comment().isEmpty()) {

        writeIndent(ts, indent);

        ts << "<context-group><context context-type=\"" << contextMsgctxt << "\">"

           << protect(msg.comment(), false)

           << "</context></context-group>\n";

    }

    if (!msg.oldComment().isEmpty()) {

        writeIndent(ts, indent);

        ts << "<context-group><context context-type=\"" << contextOldMsgctxt << "\">"

           << protect(msg.oldComment(), false)

           << "</context></context-group>\n";

    }

    writeExtras(ts, indent, msg.extras(), drops);

    if (!msg.extraComment().isEmpty()) {

        writeIndent(ts, indent);

        ts << "<note annotates=\"source\" from=\"developer\">"

           << protect(msg.extraComment()) << "</note>\n";

    }

    if (!msg.translatorComment().isEmpty()) {

        writeIndent(ts, indent);

        ts << "<note from=\"translator\">"

           << protect(msg.translatorComment()) << "</note>\n";

    }

}

static void writeTransUnits(QTextStream &ts, const TranslatorMessage &msg, const QRegExp &drops, int indent)

{

    static int msgid;

    QString msgidstr = !msg.id().isEmpty() ? msg.id() : QString::fromAscii("_msg%1").arg(++msgid);

    QStringList translns = msg.translations();

    QHash<QString, QString>::const_iterator it;

    QString pluralStr;

    QStringList sources(msg.sourceText());

    if ((it = msg.extras().find(QString::fromLatin1("po-msgid_plural"))) != msg.extras().end())

        sources.append(*it);

    QStringList oldsources;

    if (!msg.oldSourceText().isEmpty())

        oldsources.append(msg.oldSourceText());

    if ((it = msg.extras().find(QString::fromLatin1("po-old_msgid_plural"))) != msg.extras().end()) {

        if (oldsources.isEmpty()) {

            if (sources.count() == 2)

                oldsources.append(QString());

            else

                pluralStr = QLatin1Char(' ') + QLatin1String(attribPlural) + QLatin1String("=\"yes\"");

        }

        oldsources.append(*it);

    }

    QStringList::const_iterator

        srcit = sources.begin(), srcend = sources.end(),

        oldsrcit = oldsources.begin(), oldsrcend = oldsources.end(),

        transit = translns.begin(), transend = translns.end();

    int plural = 0;

    QString source;

    while (srcit != srcend || oldsrcit != oldsrcend || transit != transend) {

        QByteArray attribs;

        QByteArray state;

        if (msg.type() == TranslatorMessage::Obsolete) {

            if (!msg.isPlural())

                attribs = " translate=\"no\"";

        } else if (msg.type() == TranslatorMessage::Finished) {

            attribs = " approved=\"yes\"";

        } else if (transit != transend && !transit->isEmpty()) {

            state = " state=\"needs-review-translation\"";

        }

        writeIndent(ts, indent);

        ts << "<trans-unit id=\"" << msgidstr;

        if (msg.isPlural())

            ts << "[" << plural++ << "]";

        ts << "\"" << attribs << ">\n";

        ++indent;

        writeIndent(ts, indent);

        if (srcit != srcend) {

            source = *srcit;

            ++srcit;

        } // else just repeat last element

        ts << "<source xml:space=\"preserve\">" << protect(source) << "</source>\n";

        bool puttrans = false;

        QString translation;

        if (transit != transend) {

            translation = *transit;

            translation.replace(QChar(Translator::BinaryVariantSeparator),

                                QChar(Translator::TextVariantSeparator));

            ++transit;

            puttrans = true;

        }

        do {

            if (oldsrcit != oldsrcend && !oldsrcit->isEmpty()) {

                writeIndent(ts, indent);

                ts << "<alt-trans>\n";

                ++indent;

                writeIndent(ts, indent);

                ts << "<source xml:space=\"preserve\"" << pluralStr << '>' << protect(*oldsrcit) << "</source>\n";

                if (!puttrans) {

                    writeIndent(ts, indent);

                    ts << "<target restype=\"" << restypeDummy << "\"/>\n";

                }

            }

            if (puttrans) {

                writeIndent(ts, indent);

                ts << "<target xml:space=\"preserve\"" << state << ">" << protect(translation) << "</target>\n";

            }

            if (oldsrcit != oldsrcend) {

                if (!oldsrcit->isEmpty()) {

                    --indent;

                    writeIndent(ts, indent);

                    ts << "</alt-trans>\n";

                }

                ++oldsrcit;

            }

            puttrans = false;

        } while (srcit == srcend && oldsrcit != oldsrcend);

        if (!msg.isPlural()) {

            writeLineNumber(ts, msg, indent);

            writeComment(ts, msg, drops, indent);

        }

        --indent;

        writeIndent(ts, indent);

        ts << "</trans-unit>\n";

    }

}

static void writeMessage(QTextStream &ts, const TranslatorMessage &msg, const QRegExp &drops, int indent)

{

    if (msg.isPlural()) {

        writeIndent(ts, indent);

        ts << "<group restype=\"" << restypePlurals << "\"";

        if (!msg.id().isEmpty())

            ts << " id=\"" << msg.id() << "\"";

        if (msg.type() == TranslatorMessage::Obsolete)

            ts << " translate=\"no\"";

        ts << ">\n";

        ++indent;

        writeLineNumber(ts, msg, indent);

        writeComment(ts, msg, drops, indent);

        writeTransUnits(ts, msg, drops, indent);

        --indent;

        writeIndent(ts, indent);

        ts << "</group>\n";

    } else {

        writeTransUnits(ts, msg, drops, indent);

    }

}

class XLIFFHandler : public QXmlDefaultHandler

{

public:

    XLIFFHandler(Translator &translator, ConversionData &cd);

    bool startElement(const QString& namespaceURI, const QString &localName,

        const QString &qName, const QXmlAttributes &atts );

    bool endElement(const QString& namespaceURI, const QString &localName,

        const QString &qName );

    bool characters(const QString &ch);

    bool fatalError(const QXmlParseException &exception);

    bool endDocument();

private:

    enum XliffContext {

        XC_xliff,

        XC_group,

        XC_trans_unit,

        XC_context_group,

        XC_context_group_any,

        XC_context,

        XC_context_filename,

        XC_context_linenumber,

        XC_context_context,

        XC_context_comment,

        XC_context_old_comment,

        XC_ph,

        XC_extra_comment,

        XC_translator_comment,

        XC_restype_context,

        XC_restype_translation,

        XC_restype_plurals,

        XC_alt_trans

    };

    void pushContext(XliffContext ctx);

    bool popContext(XliffContext ctx);

    XliffContext currentContext() const;

    bool hasContext(XliffContext ctx) const;

    bool finalizeMessage(bool isPlural);

private:

    Translator &m_translator;

    ConversionData &m_cd;

    TranslatorMessage::Type m_type;

    QString m_language;

    QString m_sourceLanguage;

    QString m_context;

    QString m_id;

    QStringList m_sources;

    QStringList m_oldSources;

    QString m_comment;

    QString m_oldComment;

    QString m_extraComment;

    QString m_translatorComment;

    bool m_isPlural;

    bool m_hadAlt;

    QStringList m_translations;

    QString m_fileName;

    int     m_lineNumber;

    QString m_extraFileName;

    TranslatorMessage::References m_refs;

    TranslatorMessage::ExtraData m_extra;

    QString accum;

    QString m_ctype;

    const QString m_URITT;  // convenience and efficiency

    const QString m_URI;  // ...

    const QString m_URI12;  // ...

    QStack<int> m_contextStack;

};

XLIFFHandler::XLIFFHandler(Translator &translator, ConversionData &cd)

  : m_translator(translator), m_cd(cd),

    m_type(TranslatorMessage::Finished),

    m_lineNumber(-1),

    m_URITT(QLatin1String(TrollTsNamespaceURI)),

    m_URI(QLatin1String(XLIFF11namespaceURI)),

    m_URI12(QLatin1String(XLIFF12namespaceURI))

{}

void XLIFFHandler::pushContext(XliffContext ctx)

{

    m_contextStack.push_back(ctx);

}

// Only pops it off if the top of the stack contains ctx

bool XLIFFHandler::popContext(XliffContext ctx)

{

    if (!m_contextStack.isEmpty() && m_contextStack.top() == ctx) {

        m_contextStack.pop();

        return true;

    }

    return false;

}

XLIFFHandler::XliffContext XLIFFHandler::currentContext() const

{

    if (!m_contextStack.isEmpty())

        return (XliffContext)m_contextStack.top();

    return XC_xliff;

}

// traverses to the top to check all of the parent contexes.

bool XLIFFHandler::hasContext(XliffContext ctx) const

{

    for (int i = m_contextStack.count() - 1; i >= 0; --i) {

        if (m_contextStack.at(i) == ctx)

            return true;

    }

    return false;

}

bool XLIFFHandler::startElement(const QString& namespaceURI,

    const QString &localName, const QString &qName, const QXmlAttributes &atts )

{

    Q_UNUSED(qName);

    if (namespaceURI == m_URITT)

        goto bail;

    if (namespaceURI != m_URI && namespaceURI != m_URI12)

        return false;

    if (localName == QLatin1String("xliff")) {

        // make sure that the stack is not empty during parsing

        pushContext(XC_xliff);

    } else if (localName == QLatin1String("file")) {

        m_fileName = atts.value(QLatin1String("original"));

        m_language = atts.value(QLatin1String("target-language"));

        m_language.replace(QLatin1Char('-'), QLatin1Char('_'));

        m_sourceLanguage = atts.value(QLatin1String("source-language"));

        m_sourceLanguage.replace(QLatin1Char('-'), QLatin1Char('_'));

    } else if (localName == QLatin1String("group")) {

        if (atts.value(QLatin1String("restype")) == QLatin1String(restypeContext)) {

            m_context = atts.value(QLatin1String("resname"));

            pushContext(XC_restype_context);

        } else {

            if (atts.value(QLatin1String("restype")) == QLatin1String(restypePlurals)) {

                pushContext(XC_restype_plurals);

                m_id = atts.value(QLatin1String("id"));

                if (atts.value(QLatin1String("translate")) == QLatin1String("no"))

                    m_type = TranslatorMessage::Obsolete;

            } else {

                pushContext(XC_group);

            }

        }

    } else if (localName == QLatin1String("trans-unit")) {

        if (!hasContext(XC_restype_plurals) || m_sources.isEmpty() /* who knows ... */)

            if (atts.value(QLatin1String("translate")) == QLatin1String("no"))

                m_type = TranslatorMessage::Obsolete;

        if (!hasContext(XC_restype_plurals)) {

            m_id = atts.value(QLatin1String("id"));

            if (m_id.startsWith(QLatin1String("_msg")))

                m_id.clear();

        }