Re: Illustra System Catalog

In article <341902E0.4377@wumpus.uow.edu.au>, Carlos Fernando Enguix <cfe01@wumpus.uow.edu.au> writes >Currently I am involved in creating a customised catalog created >automatically with queries that extract information from the Illustra >system catalog. >For this reason I need to know which table,column from the system >catalog specifies the components of the primary key of a particular >table. > >I had a look to the name of all columns from the system catalog >but I could not figure out which is the one that I need. > >I suppose that probably somebody had the same problem previously, >I would really appreciate if you can give me such information. Yes, we've solved this problem in two ways! We have our own shell/awk scripts which dumps the Illustra schema for named tables, (it doesn't do rules, or multiple column constraints). I believe my colleague, David Barnwell posted an early version to the Illustra mailing list. I've developed this a bit further into a Perl script that will dump an Illustra database as an IUS schema. This handles quite a lot of the conversion for you, so oid types and converted to integers, and a serial named oid is added to each table. Timestamps and other things are converted to a large degree. Identifiers longer than 18 characters are truncated, unless there's an alternative in the perl script. If there's some interest, I'll try and get permission for these scripts posted with the hope that others can make further improvements. We've done some conversion from Illustra to IUS and would be interested in hearing any war stories. (We run both in parallel on a Sun Ultra) -- Chad Hanna Project Co-ordinator Berkshire Family History Society