2

I have a problem of dealing with close to 100000 xml records. The problem is to construct a schema mapping from xml schema of these records to relations Any ideas in this field are welcome. Please propose an algorithm / a mthodology that can be followed to achieve this schema mapping.

Or any related work would surely be helpful.

Thanks Peter

Peter Savioli
  • 21
  • 2
  • 1
    It would be helpful to provide more information here: a) are the XML records homogeneous and if so how much? b) are they shallow or are they quite hierarchical? Also, consider removing the tag 'distributed' since the question does not seem to be relevant to the tag (or add more info alternatively). – Lucas Zamboulis Mar 09 '11 at 01:41
  • A lot depends on what you mean by "lossless". For example, the problem becomes more difficult if the order of the records in the XML is significant. You've implied that the general algorithm in a published paper is more complex than necessary because your data is simple, but you haven't shared with us any information about this simplicity. – Michael Kay Mar 09 '11 at 09:09
  • Roughly how many different XML DTDs or XML Schemas are we talking about here? A few, hundreds, or thousands? – Mike Sherrill 'Cat Recall' Mar 09 '11 at 11:46
  • >> The xml records are homogeneous totally. & they are not very hierarchical. Cna say they are kind of shallow itself. – Peter Savioli Mar 09 '11 at 12:15
  • >> There is a single xml DTD which has to be mapped onto 2/3 distinct relational schemas for the same xml records. >> lossless - here means that none of the information in the xml records is lost and everything should be presented in the relational database. >> THe order of the records may change in the xml document & hence I have the task to come up with an invertible & query preserving schema mapping regardless of the ordering of XML elements. – Peter Savioli Mar 09 '11 at 12:19

1 Answers1

0

We're think on details, but it might be that you don't need an algorithm or methodology as much as you need really good tools. Altova has a set of XML tools; some of them can help you map XML documents to a SQL database. (I'm not sure whether they will help you create tables based on XML document elements.) You can download Altova Missionkit here and use it free for 30 days.

I'm sure they're not the only player in this market.

Disclaimer: I have no relationship with Altova. I've used XMLSpy briefly during a contract job for a Fortune 500 a while back. It worked well, and without surprises.

Mike Sherrill 'Cat Recall'
  • 91,602
  • 17
  • 122
  • 185
  • I would have been happy to use these tools if it was not for learning purposes. But I would really like to do it myself instead of relying on some tools. I have used xmlSpy & it does work fine without surprises – Peter Savioli Mar 09 '11 at 12:22