Murray Altheim wrote:
So am I understand right that <movie> will become a row (in relational database therms) and each subtag (title, directors, ...) a column? Or am I bound to relational model too much?
To answer your last question: yes. Xindice will allow you to store the entire document as a "record" in the database. You can then build indices or further processes into the application. The advantage is that the content is stored *as XML*, i.e., not just as a text string. All of the documents are stored in collections, and all of the collections in one root collection, and that root collection is *one big document*.
Thanks, Murray. I thought that Xindice is a (some kind of) replacement for database, where one can walk along XML files like he did with relational database (SELECTs, INSERTs, DELETEs etc., but now with XPath language). But is some persistent storage like database still needed?
May be XQuery implementation will be more useful in this case.
