>>>>>>So what happens if you remove the xsd part with an editor? If this solves the problem you might do this using string operations or the like.
>>>>>>
>>>>>>If not, you might be unable to solve the problem with VFP internal commands
>>>>>
>>>>>I cannot remove the xsd part from the xml (it is used by too many customers). Then I will look for a different approach.
>>>>
>>>>Why not? Simply create a copy to some temp and manipulate the snapshot. Just to test if you can use it at all.
>>>
>>>Testing it on my computer works. But it is a too big of a maintenance headache to implement. Different customers have all the EmailList.xsd in their XML files. But the XML files are of different "structure". Up to now, this XML was used only by .NET application which works well since it ignores the XSD part.
>>>Now I need to access the same XML from the VFP and running into the problem with the XSD.
>>
>>If it works on your comp, why not on any other comp? You need to find the reason, not the symptom.
>>
>>How, w/o a XSD, do you think VFP can determine a complex datastructure of a XML? Guesswork?
>
>Ideally if the XMLTOCURSOR() could be used (I started another thread) then I can define the structure in the cursor and then populate this cursor from XML. I was hoping that the same could be done with XMLAdapter but alas.

I hav seen the other thread. This function is even more limited. It asumes one dataset in a VFP like structure. XML Adapter can read more then one table from one XML. But the tables must be composed in a way VFP can understand. This must not be true for any XML.

Possibly see it as the following. HTML is a special kind of XML, but still valid well formated XML. What do you think those VFP XML stuff will do if you provide it a syntactical correct HTML file? Creating cursors or fail?