Parsing hyperlinks from HTML using QTextDocument?

[Violet Giraffe]

I'm having trouble with parsing simple things from a simple HTML page without using Webkit and DOM (I don't want a Webkit dependency).

1. This page says QTextDocument can parse stuff from HTML.
2. I do the following and I can see that my HTML had been parsed nicely, but only the user-visible text, no markup here:

for (QTextBlock block = doc.begin(), end = doc.end(); block != end; block = block.next())
{
	qDebug() << block.text();
}

3. I read some QTextBlock docs and try this, but anchorHref and anchorNames is always empty even for the blocks that I know are <a href>.

for (QTextBlock block = doc.begin(), end = doc.end(); block != end; block = block.next())
{
	qDebug() << block.text();
	qDebug() << block.charFormat().anchorNames();
	qDebug() << block.charFormat().anchorHref();
	qDebug() << "-------------------------------";
}

Is there any way to get the hyperlink URLs?

[raven-worx]

@Violet-Giraffe
This should work (untested though)

void searchLink(QTextFrame * parent)
{
    for( QTextFrame::iterator it = parent->begin(); !it.atEnd(); ++it )
    {
        QTextFrame *textFrame = it.currentFrame();
        QTextBlock textBlock = it.currentBlock();

        if( textFrame )
        {
            this->searchLink(textFrame);
        }
        else if( textBlock.isValid() )
        {
            this->searchLink(textBlock);
        }
    }
}

void searchLink(QTextBlock & parent)
{
    for(QTextBlock::iterator it = parent.begin(); !it.atEnd(); ++it)
    {
        QTextFragment textFragment = it.fragment();
        if( textFragment.isValid() )
        {
            QTextCharFormat textCharFormat = textFragment.charFormat();
            if( textCharFormat.isAnchor() )
            {
                 textCharFormat.anchorHref();  // <-- URL
            }
        }
    }
}

The searchLink() method searches recursively.

searchLink( textDocument->rootFrame() );

[Violet Giraffe]

Aha! So my mistake was that I only looked at blocks and not fragments. Thank you.