XML Follies

[billroper]

Things I learned about XML today:

• It would be easier to work with if I was using someone else's parser, instead of the undocumented home-grown one that I'm stuck with for the moment.
  
• UTF-16 is all well and good, but even if Excel reads an XML file in that format, it will damned well write it back out in UTF-8.
  
• Exactly why Excel complains about denormalized data remains a mystery, but by deleting the attribute of a tag that was nested two levels out, you can get it to quit complaining and do what you want.
  
• Word will also write out the file in UTF-8 format and will thoughtfully remove the CR/LF information. There's nothing like a 59,000 character line.
  
• Unicode isn't more trouble than it's worth, but it would be nice if Microsoft's support for it was a bit better.

Comments

[(Anonymous)]

But if it were only a 58,000 character line it would be okay :)

Is there a particular reason why you aren't just opening it in a browser rather than in Excel or Word? IE actually parses it quite nicely and you can collapse elements as though it were in a tree.

[madfilkentist]

Is there some reason Xerces isn't being used there? It works nicely for everything I know of.

[sdorn.livejournal.com]

Someone recommended Altova XMLSpy to me when I needed to create an odd XML document.

[billroper]

When this particular piece of code was written, there were various licensing questions about XML parsers, so this was put together.

We'll be using something else in the future.