> From: Carl Olivier [mailto:[EMAIL PROTECTED]]
> [...] > When we do a keyword search in content stored in our DB > (SAPDB) it would be great to be able to utilise a feature such as is available in > MS SQL Server (using the LIKE predicate and the % wildcard) as follows: > > select id from content where content like "%mykeyword%" > > obviously this is in the case where the colum content is a LONG.... I think this was clear. What Elke has asked is how much data typically is stored in your "content" field? Is it in the KB or GB range? They need this information since - i think - it's much easier for the SAP DB people to introduce a "NOT SO LONG VARCHAR" type with a maximum of - say - 64 KB. Am i right? But besides - what are the limitations of SQL Server for LONG VARCHAR? I think i have read about 8GB - and in this case i doubt SQL Server does accept a LIKE predicate on 8GB data, and - if it does - how is the performance? Does it scan the data on insert/update and creates some kind of hash table internally? I think this is why they need BINARY and IMAGE subtypes. Regards, Peter -- Peter Sch�fer EADS Radio Communication Systems GmbH & Co. KG NOT SPEAKING FOR MY EMPLOYER _______________________________________________ sapdb.general mailing list [EMAIL PROTECTED] http://listserv.sap.com/mailman/listinfo/sapdb.general
