Dynamic xsd and Biztalk 2006

Question

Hi all,

I'm sort of a biztalk newbie and we run 2006 here with plans to migrate to 2006 R2 in the future. I have a schema that I need to work with called Siri, which is public at:

link text or link text for the root xsd document.

I have successfully imported this schema into my biztalk project, however, I am finding it difficult to get my head wrapped around how to use this since this xsd schema is all xsd:choice and xsd:group nodes. Also there is only one targetnamespace at the root element.

My project is to consume estimated timetables of buses and then insert this data into a database table. After working with the schema in Biztalk I'm finding it difficult to promote elements in order to filter and create/work with mappings (haven't really got that far yet).

Can anyone point me in the right direction in some documentation/websites or basically how I should approach this? Has anyone worked with a schema like this in biztalk 2006 and if so how did you approach message mapping, retrieval of data from the xml and inserting into the DB? I can post up a sample xml payload if anyone would like to have a look..

thanks for the help in advance..

Answer 1

Hi,

Heres a good book to read which will answer most of your questions:

BizTalk 2006 Recipes

Cheers Hemil

Answer 2

Hi, How many parties are delivering input file to you? is it possible to correlate on incoming party and then have a more predefined mapping for each?

Offcourse you can also do If/Else on the Choice/Group element and map accordingly.

• or a mix of both the the strategies mentioned above.

Is it possible for you to give me a sample-input-file?

good luck!

Answer 3

Schemas like this are definitely workable with BizTalk, but it may take some more effort and some good understanding of XSD, no doubt.

I think that, on top of reading one or two good books on BizTalk, which always helps (and there are several good ones out there), you could get a lot of help on the Biztalk specific forums and newsgroups as well as here if you post specific questions, as it's quite hard to describe an overall solution to property promotion or mapping.

THere's one other thing to consider, although with great care, and that is to not use their schema directly.

Schemas that [try to] define a standard have to be very generic; often that leads to complexities as well as varaieties through extension points which makes life hard for development and maintenance.

If you were able to derive from that schema a schema that describes only the valid strucutre that you are expecting to get (which might be a small subset of what the standard can suppoort) you might end up with a much smaller and simpler one.

The advantage is easier development and maintenance, the disadvantage is that if you reach a point where you need to support more of the original schema you have to change your schema and re-deploy. Also - you risk making your life harder if and when you want to move to newer version of the standard, although that might be the case anyway

Hope that hekos