skipping invalid nodes with XmlTextReader

[Stephen S Kelley]

I want to do something like:

bool bContinueReading=true;
XmlTextReader Reader=new XmlTextReader("file.xml");
while(bContinueReading)
{
try
{
Reader.Read();
...
}
catch
{
Reader.Skip();
}
}

To skip over any xml formatting problems (invalid characters) in the file.
However, the Skip() function does not seem to move the cursor forward and an
infinite loop results when an error occurs.

[Stephen S Kelley]

The code should read:

bool bContinueReading=true;
XmlTextReader Reader=new XmlTextReader("file.xml");
while(bContinueReading)
{
try
{
bContinueReading=Reader.Read();
...
}
catch
{
Reader.Skip();
}
}

[Oleg Tkachenko [MVP]]

Stephen S Kelley wrote:

To skip over any xml formatting problems (invalid characters) in the file.
However, the Skip() function does not seem to move the cursor forward and an
infinite loop results when an error occurs.

You can't by design. XML spec requires draconian error handling - any
malformed XML document must not be parsed successfully.

--
Oleg Tkachenko [XML MVP, MCAD]
http://blog.tkachenko.com

[Chris Lovett]

And the XmlReader implements this W3C XML Spec. The problem with error
recovery in general is it becomes ambiguous which is generally not good for
mission critical business applications. So you will have to employ a more
forgiving parser (see SgmlReader on www.gotdotnet.com for an example).

"Oleg Tkachenko [MVP]" <oleg@NO!SPAM!PLEASEtkachenko.com> wrote in message
news:et**************@TK2MSFTNGP10.phx.gbl...

Stephen S Kelley wrote:

To skip over any xml formatting problems (invalid characters) in the
file. However, the Skip() function does not seem to move the cursor
forward and an infinite loop results when an error occurs.

You can't by design. XML spec requires draconian error handling - any
malformed XML document must not be parsed successfully.

--
Oleg Tkachenko [XML MVP, MCAD]
http://blog.tkachenko.com