qxml.cpp

奇趣公司比较新的qt/emd版本

    d->inputDevice = 0;
    d->inputStream = 0;

    setData(QString());
#ifndef QT_NO_TEXTCODEC
    d->encMapper = 0;
#endif
    d->nextReturnedEndOfData = true; // first call to next() will call fetchData()

    d->encodingDeclBytes.clear();
    d->encodingDeclChars.clear();
    d->lookingForEncodingDecl = true;
}

/*!
    Constructs an input source which contains no data.

    \sa setData()
*/
QXmlInputSource::QXmlInputSource()
{
    init();
}

/*!
    Constructs an input source and gets the data from device \a dev.
    If \a dev is not open, it is opened in read-only mode. If \a dev
    is 0 or it is not possible to read from the device, the input
    source will contain no data.

    \sa setData() fetchData() QIODevice
*/
QXmlInputSource::QXmlInputSource(QIODevice *dev)
{
    init();
    d->inputDevice = dev;
    d->inputDevice->setTextModeEnabled(false);
}

#ifdef QT3_SUPPORT
/*!
    Use the QXmlInputSource(QIODevice *) constructor instead, with
    the device used by \a stream.

    \sa QTextStream::device()
*/
QXmlInputSource::QXmlInputSource(QTextStream& stream)
{
    init();
    d->inputStream = &stream;
}

/*!
    Use QXmlInputSource(&\a file) instead.
*/
QXmlInputSource::QXmlInputSource(QFile& file)
{
    init();
    d->inputDevice = &file;
}
#endif

/*!
    Destructor.
*/
QXmlInputSource::~QXmlInputSource()
{
#ifndef QT_NO_TEXTCODEC
    delete d->encMapper;
#endif
    delete d;
}

/*!
Returns the next character of the input source. If this function
reaches the end of available data, it returns
QXmlInputSource::EndOfData. If you call next() after that, it
tries to fetch more data by calling fetchData(). If the
fetchData() call results in new data, this function returns the
first character of that data; otherwise it returns
QXmlInputSource::EndOfDocument.

Readers, such as QXmlSimpleReader, will assume that the end of
the XML document has been reached if the this function returns
QXmlInputSource::EndOfDocument, and will check that the
supplied input is well-formed. Therefore, when reimplementing
this function, it is important to ensure that this behavior is
duplicated.

\sa reset() fetchData() QXmlSimpleReader::parse() QXmlSimpleReader::parseContinue()
*/
QChar QXmlInputSource::next()
{
    if (d->pos >= d->length) {
        if (d->nextReturnedEndOfData) {
            d->nextReturnedEndOfData = false;
            fetchData();
            if (d->pos >= d->length) {
                return EndOfDocument;
            }
            return next();
        }
        d->nextReturnedEndOfData = true;
        return EndOfData;
    }

    // QXmlInputSource has no way to signal encoding errors. The best we can do
    // is return EndOfDocument. We do *not* return EndOfData, because the reader
    // will then just call this function again to get the next char.
    QChar c = d->unicode[d->pos++];
    if (c.unicode() == EndOfData)
        c = EndOfDocument;
    return c;
}

/*!
    This function sets the position used by next() to the beginning of
    the data returned by data(). This is useful if you want to use the
    input source for more than one parse.

    \sa next()
*/
void QXmlInputSource::reset()
{
    d->nextReturnedEndOfData = false;
    d->pos = 0;
}

/*!
    Returns the data the input source contains or an empty string if the
    input source does not contain any data.

    \sa setData() QXmlInputSource() fetchData()
*/
QString QXmlInputSource::data() const
{
    if (d->nextReturnedEndOfData) {
        QXmlInputSource *that = const_cast<QXmlInputSource*>(this);
        that->d->nextReturnedEndOfData = false;
        that->fetchData();
    }
    return d->str;
}

/*!
    Sets the data of the input source to \a dat.

    If the input source already contains data, this function deletes
    that data first.

    \sa data()
*/
void QXmlInputSource::setData(const QString& dat)
{
    d->str = dat;
    d->unicode = dat.unicode();
    d->pos = 0;
    d->length = d->str.length();
    d->nextReturnedEndOfData = false;
}

/*!
    \overload

    The data \a dat is passed through the correct text-codec, before
    it is set.
*/
void QXmlInputSource::setData(const QByteArray& dat)
{
    setData(fromRawData(dat));
}

/*!
    This function reads more data from the device that was set during
    construction. If the input source already contained data, this
    function deletes that data first.

    This object contains no data after a call to this function if the
    object was constructed without a device to read data from or if
    this function was not able to get more data from the device.

    There are two occasions where a fetch is done implicitly by
    another function call: during construction (so that the object
    starts out with some initial data where available), and during a
    call to next() (if the data had run out).

    You don't normally need to use this function if you use next().

    \sa data() next() QXmlInputSource()
*/

void QXmlInputSource::fetchData()
{
    enum
    {
        BufferSize = 1024
    };

    QByteArray rawData;

    if (d->inputDevice || d->inputStream) {
        QIODevice *device = d->inputDevice ? d->inputDevice : d->inputStream->device();

        if (!device) {
            if (d->inputStream && d->inputStream->string()) {
                QString *s = d->inputStream->string();
                rawData = QByteArray((const char *) s->constData(), s->size() * sizeof(QChar));
            }
        } else if (device->isOpen() || device->open(QIODevice::ReadOnly)) {
            rawData.resize(BufferSize);
            qint64 size = device->read(rawData.data(), BufferSize);

            if (size != -1) {
                // We don't want to give fromRawData() less than four bytes if we can avoid it.
                while (size < 4) {
                    if (!device->waitForReadyRead(-1))
                        break;
                    int ret = device->read(rawData.data() + size, BufferSize - size);
                    if (ret <= 0)
                        break;
                    size += ret;
                }
            }

            rawData.resize(qMax(qint64(0), size));
        }

        /* We do this inside the "if (d->inputDevice ..." scope
         * because if we're not using a stream or device, that is,
         * the user set a QString manually, we don't want to set
         * d->str. */
        setData(fromRawData(rawData));
    }
}

#ifndef QT_NO_TEXTCODEC
static QString extractEncodingDecl(const QString &text, bool *needMoreText)
{
    *needMoreText = false;

    int l = text.length();
    QString snip = QString::fromLatin1("<?xml").left(l);
    if (l > 0 && !text.startsWith(snip))
        return QString();

    int endPos = text.indexOf(QLatin1Char('>'));
    if (endPos == -1) {
        *needMoreText = l < 255; // we won't look forever
        return QString();
    }

    int pos = text.indexOf(QLatin1String("encoding"));
    if (pos == -1 || pos >= endPos)
        return QString();

    while (pos < endPos) {
        ushort uc = text.at(pos).unicode();
        if (uc == '\'' || uc == '"')
            break;
        ++pos;
    }

    if (pos == endPos)
        return QString();

    QString encoding;
    ++pos;
    while (pos < endPos) {
        ushort uc = text.at(pos).unicode();
        if (uc == '\'' || uc == '"')
            break;
        encoding.append(uc);
        ++pos;
    }

    return encoding;
}
#endif // QT_NO_TEXTCODEC

/*!
    This function reads the XML file from \a data and tries to
    recognize the encoding. It converts the raw data \a data into a
    QString and returns it. It tries its best to get the correct
    encoding for the XML file.

    If \a beginning is true, this function assumes that the data
    starts at the beginning of a new XML document and looks for an
    encoding declaration. If \a beginning is false, it converts the
    raw data using the encoding determined from prior calls.
*/
QString QXmlInputSource::fromRawData(const QByteArray &data, bool beginning)
{
#ifdef QT_NO_TEXTCODEC
    Q_UNUSED(beginning);
    return QString::fromAscii(data.constData(), data.size());
#else
    if (data.size() == 0)
        return QString();
    if (beginning) {
        delete d->encMapper;
        d->encMapper = 0;
    }

    int mib = 106; // UTF-8

    // This is the initial UTF codec we will read the encoding declaration with
    if (d->encMapper == 0) {
        d->encodingDeclBytes.clear();
        d->encodingDeclChars.clear();
        d->lookingForEncodingDecl = true;

        // look for byte order mark and read the first 5 characters
        if (data.size() >= 2) {
            uchar ch1 = data.at(0);
            uchar ch2 = data.at(1);

            if (ch1 == 0xfe && ch2 == 0xff || ch1 == 0xff && ch2 == 0xfe)
                mib = 1015; // UTF-16 with byte order mark
            else if (ch1 == 0x3c && ch2 == 0x00)
                mib = 1014; // UTF-16LE
            else if (ch1 == 0x00 && ch2 == 0x3c)
                mib = 1013; // UTF-16BE
        }

        QTextCodec *codec = QTextCodec::codecForMib(mib);
        Q_ASSERT(codec);

        d->encMapper = codec->makeDecoder();
    }

    QString input = d->encMapper->toUnicode(data, data.size());

    if (d->lookingForEncodingDecl) {
        d->encodingDeclChars += input;

        bool needMoreText;
        QString encoding = extractEncodingDecl(d->encodingDeclChars, &needMoreText);

        if (!encoding.isEmpty()) {
            if (QTextCodec *codec = QTextCodec::codecForName(encoding.toLatin1())) {
                /* If the encoding is the same, we don't have to do toUnicode() all over again. */
                if(codec->mibEnum() != mib) {
                    delete d->encMapper;
                    d->encMapper = codec->makeDecoder();

                    /* The variable input can potentially be large, so we deallocate
                     * it before calling toUnicode() in order to avoid having two
                     * large QStrings in memory simultaneously. */
                    input.clear();

                    // prime the decoder with the data so far
                    d->encMapper->toUnicode(d->encodingDeclBytes, d->encodingDeclBytes.size());
                    // now feed it the new data
                    input = d->encMapper->toUnicode(data, data.size());
                }
            }
        }

        d->encodingDeclBytes += data;
        d->lookingForEncodingDecl = needMoreText;
    }

    return input;
#endif
}


/*********************************************
 *
 * QXmlDefaultHandler
 *
 *********************************************/

/*!
    \class QXmlContentHandler
    \reentrant
    \brief The QXmlContentHandler class provides an interface to
    report the logical content of XML data.

    \module XML
    \ingroup xml-tools

    If the application needs to be informed of basic parsing events,
    it can implement this interface and activate it using
    QXmlReader::setContentHandler(). The reader can then report basic
    document-related events like the start and end of elements and
    character data through this interface.

    The order of events in this interface is very important, and
    mirrors the order of information in the document itself. For
    example, all of an element's content (character data, processing
    instructions, and sub-elements) appears, in order, between the
    startElement() event and the corresponding endElement() event.

    The class QXmlDefaultHandler provides a default implementation for
    this interface; subclassing from the QXmlDefaultHandler class is
    very convenient if you only want to be informed of some parsing
    events.

    The startDocument() function is called at the start of the
    document, and endDocument() is called at the end. Before parsing
    begins setDocumentLocator() is called. For each element
    startElement() is called, with endElement() being called at the
    end of each element. The characters() function is called with
    chunks of character data; ignorableWhitespace() is called with
    chunks of whitespace and processingInstruction() is called with
    processing instructions. If an entity is skipped skippedEntity()
    is called. At the beginning of prefix-URI scopes
    startPrefixMapping() is called.

    \sa QXmlDTDHandler, QXmlDeclHandler, QXmlEntityResolver, QXmlErrorHandler,
        QXmlLexicalHandler, {Introduction to SAX2}
*/

/*!
    \fn QXmlContentHandler::~QXmlContentHandler()