metatranslator.cpp

QT 开发环境里面一个很重要的文件

/****************************************************************************
**
** Copyright (C) 1992-2006 Trolltech ASA. All rights reserved.
**
** This file is part of the Qt Linguist of the Qt Toolkit.
**
** This file may be used under the terms of the GNU General Public
** License version 2.0 as published by the Free Software Foundation
** and appearing in the file LICENSE.GPL included in the packaging of
** this file.  Please review the following information to ensure GNU
** General Public Licensing requirements will be met:
** http://www.trolltech.com/products/qt/opensource.html
**
** If you are unsure which license is appropriate for your use, please
** review the following information:
** http://www.trolltech.com/products/qt/licensing.html or contact the
** sales department at [e-mail].
**
** This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
** WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
**
****************************************************************************/

#include "metatranslator.h"
#include "translator.h"

#include <QApplication>
#include <QByteArray>
#include <QFile>
#include <QMessageBox>
#include <QTextCodec>
#include <QTextStream>
#include <QtXml>
#include <QtCore/QDir>
#include <QtCore/QFileInfo>

static bool encodingIsUtf8( const QXmlAttributes& atts )
{
    for ( int i = 0; i < atts.length(); i++ ) {
        // utf8="true" is a pre-3.0 syntax
        if ( atts.qName(i) == QString("utf8") ) {
            return ( atts.value(i) == QString("true") );
        } else if ( atts.qName(i) == QString("encoding") ) {
            return ( atts.value(i) == QString("UTF-8") );
        }
    }
    return false;
}

class TsHandler : public QXmlDefaultHandler
{
public:
    TsHandler( MetaTranslator *translator )
        : tor( translator ), type( MetaTranslatorMessage::Finished ),
          inMessage( false ), ferrorCount( 0 ), contextIsUtf8( false ),
          messageIsUtf8( false ), m_isPlural(false) { }

    virtual bool startElement( const QString& namespaceURI,
                               const QString& localName, const QString& qName,
                               const QXmlAttributes& atts );
    virtual bool endElement( const QString& namespaceURI,
                             const QString& localName, const QString& qName );
    virtual bool characters( const QString& ch );
    virtual bool fatalError( const QXmlParseException& exception );

    QString language() const { return m_language; }
private:
    MetaTranslator *tor;
    MetaTranslatorMessage::Type type;
    bool inMessage;
    QString m_language;
    QString context;
    QString source;
    QString comment;
    QStringList translations;
    QString m_fileName;
    int     m_lineNumber;

    QString accum;
    int ferrorCount;
    bool contextIsUtf8;
    bool messageIsUtf8;
    bool m_isPlural;
};

bool TsHandler::startElement( const QString& /* namespaceURI */,
                              const QString& /* localName */,
                              const QString& qName,
                              const QXmlAttributes& atts )
{
    if ( qName == QString("byte") ) {
        for ( int i = 0; i < atts.length(); i++ ) {
            if ( atts.qName(i) == QString("value") ) {
                QString value = atts.value( i );
                int base = 10;
                if ( value.startsWith("x") ) {
                    base = 16;
                    value = value.mid( 1 );
                }
                int n = value.toUInt( 0, base );
                if ( n != 0 )
                    accum += QChar( n );
            }
        }
    } else {
        if ( qName == QString("TS") ) {
            m_language = atts.value(QLatin1String("language"));
        } else if ( qName == QString("context") ) {
            context.clear();
            source.clear();
            comment.clear();
            translations.clear();
            contextIsUtf8 = encodingIsUtf8( atts );
        } else if ( qName == QString("message") ) {
            inMessage = true;
            type = MetaTranslatorMessage::Finished;
            source.clear();
            comment.clear();
            translations.clear();
            messageIsUtf8 = encodingIsUtf8( atts );
            m_isPlural = atts.value(QLatin1String("numerus")).compare(QLatin1String("yes")) == 0;
        } else if (qName == QString("location") && inMessage) {
            bool bOK;
            int lineNo = atts.value(QString("line")).toInt(&bOK);
            if (!bOK) lineNo = -1;
            m_fileName = atts.value(QString("filename"));
            m_lineNumber = lineNo;
        } else if ( qName == QString("translation") ) {
            for ( int i = 0; i < atts.length(); i++ ) {
                if ( atts.qName(i) == QString("type") ) {
                    if ( atts.value(i) == QString("unfinished") )
                        type = MetaTranslatorMessage::Unfinished;
                    else if ( atts.value(i) == QString("obsolete") )
                        type = MetaTranslatorMessage::Obsolete;
                    else
                        type = MetaTranslatorMessage::Finished;
                }
            }
        }
        accum.clear();
    }
    return true;
}

bool TsHandler::endElement( const QString& /* namespaceURI */,
                            const QString& /* localName */,
                            const QString& qName )
{
    if ( qName == QString("codec") || qName == QString("defaultcodec") ) {
        // "codec" is a pre-3.0 syntax
        tor->setCodec( accum.toLatin1() );
    } else if ( qName == QString("name") ) {
        context = accum;
    } else if ( qName == QString("source") ) {
        source = accum;
    } else if ( qName == QString("comment") ) {
        if ( inMessage ) {
            comment = accum;
        } else {
            if ( contextIsUtf8 )
                tor->insert( MetaTranslatorMessage(context.toUtf8(),
                             ContextComment, accum.toUtf8(), QString(), 0,
                             QStringList(), true,
                             MetaTranslatorMessage::Unfinished) );
            else
                tor->insert( MetaTranslatorMessage(context.toAscii(),
                             ContextComment, accum.toAscii(), QString(), 0,
                             QStringList(), false,
                             MetaTranslatorMessage::Unfinished) );
        }
    } else if ( qName == QString("numerusform") ) {
        translations.append(accum);
        m_isPlural = true;
    } else if ( qName == QString("translation") ) {
        if (translations.isEmpty())
            translations.append(accum);
    } else if ( qName == QString("message") ) {
        if ( messageIsUtf8 )
            tor->insert( MetaTranslatorMessage(context.toUtf8(), source.toUtf8(),
                                            comment.toUtf8(), m_fileName, m_lineNumber, 
                                            translations, true, type, m_isPlural) );
        else
            tor->insert( MetaTranslatorMessage(context.toAscii(), source.toAscii(),
                                            comment.toAscii(), m_fileName, m_lineNumber, 
                                            translations, false, type, m_isPlural) );
        inMessage = false;
    }
    return true;
}

bool TsHandler::characters( const QString& ch )
{
    QString t = ch;
    t.replace( "\r", "" );
    accum += t;
    return true;
}

bool TsHandler::fatalError( const QXmlParseException& exception )
{
    if ( ferrorCount++ == 0 ) {
        QString msg;
        msg.sprintf( "Parse error at line %d, column %d (%s).",
                     exception.lineNumber(), exception.columnNumber(),
                     exception.message().toLatin1().data() );
        if ( qApp == 0 )
            fprintf( stderr, "XML error: %s\n", msg.toLatin1().data() );
        else
            QMessageBox::information(0,
                                      QObject::tr("Qt Linguist"), msg );
    }
    return false;
}

static QString numericEntity( int ch )
{
    return QString( ch <= 0x20 ? "<byte value=\"x%1\"/>" : "&#x%1;" )
           .arg( ch, 0, 16 );
}

static QString protect( const QByteArray& str )
{
    QString result;
    int len = (int) str.length();
    for ( int k = 0; k < len; k++ ) {
        switch( str[k] ) {
        case '\"':
            result += QString( "&quot;" );
            break;
        case '&':
            result += QString( "&amp;" );
            break;
        case '>':
            result += QString( "&gt;" );
            break;
        case '<':
            result += QString( "&lt;" );
            break;
        case '\'':
            result += QString( "&apos;" );
            break;
        default:
            if ( (uchar) str[k] < 0x20 && str[k] != '\n' )
                result += numericEntity( (uchar) str[k] );
            else
                result += str[k];
        }
    }
    return result;
}

static QString evilBytes( const QByteArray& str, bool utf8 )
{
    if ( utf8 ) {
        return protect( str );
    } else {
        QString result;
        QByteArray t = protect( str ).toLatin1();
        int len = (int) t.length();
        for ( int k = 0; k < len; k++ ) {
            if ( (uchar) t[k] >= 0x7f )
                result += numericEntity( (uchar) t[k] );
            else
                result += QChar( t[k] );
        }
        return result;
    }
}

MetaTranslatorMessage::MetaTranslatorMessage()
    : utfeight( false ), ty( Unfinished ), m_plural(false)
{
}

MetaTranslatorMessage::MetaTranslatorMessage( const char *context,
                                              const char *sourceText,
                                              const char *comment,
                                              const QString &fileName,
                                              int lineNumber,
                                              const QStringList& translations,
                                              bool utf8, Type type, bool plural )
    : TranslatorMessage( context, sourceText, comment, fileName, lineNumber, translations ),
      utfeight( false ), ty( type ), m_plural(plural)
{
    /*
      Don't use UTF-8 if it makes no difference. UTF-8 should be
      reserved for the real problematic case: non-ASCII (possibly
      non-Latin1) characters in .ui files.
    */
    if ( utf8 ) {
        if ( sourceText != 0 ) {
            int i = 0;
            while ( sourceText[i] != '\0' ) {
                if ( (uchar) sourceText[i] >= 0x80 ) {
                    utfeight = true;
                    break;
                }
                i++;
            }
        }
        if ( !utfeight && comment != 0 ) {
            int i = 0;
            while ( comment[i] != '\0' ) {
                if ( (uchar) comment[i] >= 0x80 ) {
                    utfeight = true;
                    break;
                }
                i++;
            }
        }

        if ( !utfeight && context != 0 ) {
            int i = 0;
            while ( context[i] != '\0' ) {
                if ( (uchar) context[i] >= 0x80 ) {
                    utfeight = true;
                    break;
                }
                i++;
            }
        }
    }
}


MetaTranslatorMessage::MetaTranslatorMessage( const MetaTranslatorMessage& m )
    : TranslatorMessage( m ), utfeight( m.utfeight ), ty( m.ty ), m_plural(m.m_plural)
{
}

MetaTranslatorMessage& MetaTranslatorMessage::operator=(
        const MetaTranslatorMessage& m )
{
    TranslatorMessage::operator=( m );
    utfeight = m.utfeight;
    ty = m.ty;
    m_plural = m.m_plural;
    return *this;
}

bool MetaTranslatorMessage::operator==( const MetaTranslatorMessage& m ) const
{
    return qstrcmp( context(), m.context() ) == 0 &&
           qstrcmp( sourceText(), m.sourceText() ) == 0 &&
           qstrcmp( comment(), m.comment() ) == 0;
}

bool MetaTranslatorMessage::operator<( const MetaTranslatorMessage& m ) const
{
    int delta = qstrcmp( context(), m.context() );
    if ( delta == 0 )
        delta = qstrcmp( sourceText(), m.sourceText() );
    if ( delta == 0 )
        delta = qstrcmp( comment(), m.comment() );
    return delta < 0;
}

MetaTranslator::MetaTranslator()
{
    clear();
}

MetaTranslator::MetaTranslator( const MetaTranslator& tor )
    : mm( tor.mm ), codecName( tor.codecName ), codec( tor.codec )
{
}

MetaTranslator& MetaTranslator::operator=( const MetaTranslator& tor )
{
    mm = tor.mm;
    codecName = tor.codecName;
    codec = tor.codec;
    return *this;
}

void MetaTranslator::clear()
{
    mm.clear();
    codecName = "ISO-8859-1";
    codec = 0;
}

bool MetaTranslator::load( const QString& filename )
{
    QFile f( filename );
    if ( !f.open(QIODevice::ReadOnly) )
        return false;

    QXmlInputSource in( &f );
    QXmlSimpleReader reader;
    reader.setFeature( "http://xml.org/sax/features/namespaces", false );
    reader.setFeature( "http://xml.org/sax/features/namespace-prefixes", true );
    TsHandler *hand = new TsHandler( this );
    reader.setContentHandler( static_cast<QXmlDefaultHandler*>(hand) );
    reader.setErrorHandler( static_cast<QXmlDefaultHandler*>(hand) );

    bool ok = reader.parse( in );
    reader.setContentHandler( 0 );
    reader.setErrorHandler( 0 );

    m_language = hand->language();
    makeFileNamesAbsolute(QFileInfo(filename).absoluteDir());

    delete hand;
    f.close();
    return ok;