Mining Structured arrays in BLOBs (170 Views)
Reply
Occasional Contributor
Andy Plant
Posts: 7
Registered: ‎01-05-2010
Message 1 of 2 (170 Views)

Mining Structured arrays in BLOBs

This isn't a question it is a bit of knowledge that I just gleaned out of HP after 3 months of bashing on the door I finally got throught to R&D. If you want to use a toolset like Crystal reports and the like to mine data stored in structured arrays inside SM7 there is a quick fix that works a treat. In quotes is the response from HP.
"I have checked this with one of our lead engineers and he has told me of the following process where you can map fields in the labor structure to a new table and you can then run your reports off that instead of the probsummary table.

Go to database dictionary and bring up the probsummary table. Double click on the labor structure which will open a popup window where you can enter a Attribute file name and unique Attribute. Add a file name in both fields, eg: laborattribute and laborunique. Click OK and OK again to exit the dbdict.
The laborattribute table will have the fields in a normal scalar fashion, ie: text, date/time, number etc. Not a blob/clob.
You can then report off of the new table from the rdbms. The unique id is also included in the new table, so you will have the ticket number."
Please use plain text.
Respected Contributor
Kelalek2
Posts: 180
Registered: ‎08-05-2011
Message 2 of 2 (117 Views)

Re: Mining Structured arrays in BLOBs

Just wanted to raise this - great tip by Andy Plant. If you need to make reports by making queries to SQL database, this is essential to get data out from for example "clocks" table. We're tracking how long different assingment groups are holding the tickets...

Please use plain text.
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation