Hi,
Creating a very "abstract" xml parser is not easy and it's quite error-prone. You have to know some (or at least one) parent or common tag names in order to have a starting point for any deeper "abstract" parsing. Basically, you have to move the parser step by step checking everything in the way. For example, you could use a simple function as an entry point for the following function when you meet a specific tag:
void XmlReader::readFragment()
{
Q_ASSERT(m_reader->isStartElement()); //Make sure the xmlreader is at an opening tag
const QString p = m_reader->qualifiedName().toString(); // From now on, you will work with this opening tag
//Do what you want, for example handle attributes and their values
for (auto i = 0; i < m_reader->attributes().count(); ++i) {
...
}
m_reader->readNext(); //Move on
...
//Do everything manually like this
if (m_reader->isCharacters) {
...
} else if (m_reader->isEndElement()) {
...
}
//or like this
//Eventually, you have to check for the closing current tag, exit the function,
//and return the control to the "entry point" function
while (!m_reader->isEndElement() && !m_reader->qualifiedName().toString() == p) {
... //Do your work
m_reader->readNext();
}
}
However, with this approach, things can get extremely complicated if there are nested tags, for example:
<p>
<note>
<p>...</p>
</note>
</p>
Maybe it would be easier to subclass QXmlStreamReader but it would help to provide some sample xml data and more info on what exactly you want to do.
