"Ian Harding" <iharding160@hotmail.com> wrote in message
news:42kr5dF1jo2i9U1@individual.net...
>I have a class derived from XmlTextReader. It implements the following
>methods:
> public override void ReadEndElement()
> public override void ReadStartElement()
> public override string ReadString()
> public override bool Read()
>
> When the class is used to read a document, the Read override is called
> many times and I can see the document contents being stepped through using
> base.LocalName etc. What I don't understand is that the other three
> overrides are never called. Samples I've found on Google seem to indicate
> that all ReadXXX methods are called when a relevant document item is
> encountered but that isn't happening in this case.
>
> My aim is to replace an attribute value on all occurances of a specific
> element. I can add code into the Read() override to identify when the
> reader is processing that element, but need to override the behaviour of
> ReadString() so I can substitute the new attribute value.
>
> Thanks,
> Ian

Naraendira Kuma R.R. wrote:
> Seems like you could accomplish the same goal, easily using XSLT without
> using XmlTextReader.
>

"Ian Harding" <iharding160@hotmail.com> wrote in message
news:42mog8F1jp5c4U1@individual.net...
> Naraendira Kuma R.R. wrote:
>> Seems like you could accomplish the same goal, easily using XSLT without
>> using XmlTextReader.
>>
>
> I was trying to keep overhead to a minimum. What I've read indicates that
> reading a document with an XmlReader should be more efficient that XSLT in
> this case. The XML document arrives in a string, the XmlTextReader
> isolates one part of the document (an ADO Recordset) into another string,
> then the Recordset is de-serialized from the XML.
>
> On the other hand, if there's no way of substituting an attribute value
> using an XmlReader, I might have to go the XSLT route.
>
> Thanks,
> Ian