QDataStream. How to skip data using serialization?

_Dima

Could somebody advise me how to skip data using serialized stream?

There is method "skip", but it needs just amount of raw bytes. I would like to use just Qt serialization mechanism.

Some details:

What am I trying to do?
I want to implement small file container which encapsulates few files inside.
!http://s22.postimg.org/j8g8roukx/FAT_schema.png(file container)!
What is problem?
That's reading of the certain file by index. For example, skip QByteArray's (from 0 to index-1) and read just QByteArray at "index" position.
Some proof code
@#include <QFile>
#include <QDataStream>

static const int QDataStreamVersion = QDataStream::Qt_4_8;
static const quint32 Magic = 0xA0B0C0D0;
static const quint32 Version = 1;

struct FatItem
{
QString fname;
quint32 index;
};

QDataStream & operator>> (QDataStream & _stream, FatItem & _fatItem)
{
_stream >> _fatItem.fname;
_stream >> _fatItem.index;
return _stream;
}

void openContainerAndExtractAllFiles(QString _containerPath, QString _destinationDir)
{
//open storage file
QFile container(_containerPath);
if(!container.open(QIODevice::ReadOnly))
{
qDebug() << "container"<< _containerPath << " error=" << container.errorString();
return;
}

// assign stream to file
QDataStream stream(&container);

//check for key number
quint32 magic;
stream &gt;&gt; magic;
if (magic != Magic)
{
    qDebug() &lt;&lt; "Storage file: bad format";
    return;
}

//check for correct version of the storage file
qint32 version;
stream >> version;
if (version < Version )
{
    qDebug() << "file version is too old";
    return;
}
else if (version > Version )
{
    qDebug() << "file version is too new";
    return;
}


// load FAT
QList<FatItem> fat;
stream >> fat;

// extract all files and put them to destination dir
for (quint32 i=0; i<fat.count() && !stream.atEnd(); i++)
{
    // extract
    QByteArray fData;
    stream >> fData;

    // put
    QFile f(_destinationDir + fat.at(i).fname);
    if (!f.open(QIODevice::WriteOnly))continue;
    f.write(fData);
    f.close();
}

container.close();

}@

andre

As the file layout is documented, you can find how the byte array is stored. I think it is basically just a field with the length, and then the raw data. You can read the length field yourself, and then you know how many bytes to skip to the next item.