Wow.....yikes. Hmmm. The unfortunate part is that if an XML document is not "well-formed", then as far as the parser is concerned, it might as well not even be an XML document at all. It either is or isn't, I know of no parsers that can say something like "Close enough."

In Unix, in a situation like this, I think I would head for the "strings" utility which scrapes out everythign that looks like text (i.e. 7 bit character strings of meaningful length) from a binary (i.e. 8bit) file. If the XML portion you seek is stored in a fairly unencrypted way inside the docx file, you could scrape the text out that way and see what it takes to do a manual repair?

Sorry, that's the best I've got for now. I've attached a link to a version of strings for Windows if you need it. If I think of anything else I'll come back and comment on myself.

UPDATE : Just tried strings on a typical docx file and I don't think that's what you need. Sorry I could not be of more help. Are you saying that you do in fact have a document.xml file? A text file that is just not well formed? Have you tried hitting it with a SAX parser or StAX (streaming) parser, which will only look at elements as they come up and might give you a higher quality parse?
Source(s):
http://blog.stevienova.com/2004/07/15/strings-for-windows-xp/

Asker's Rating:

Permalink | Report