From: "Guennadi V. Liakhovetski" Newsgroups: comp.databases.theory,comp.databases Subject: SQL over tree-type database? Date: Sat, 5 May 2001 21:51:11 +0100 Organization: Sheffield University Lines: 26 Sender: ap1gvl@acms23 Message-ID: NNTP-Posting-Host: acms23.shef.ac.uk Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: hermes.shef.ac.uk 989095875 13074 143.167.4.62 (5 May 2001 20:51:15 GMT) X-Complaints-To: usenet@shef.ac.uk NNTP-Posting-Date: Sat, 5 May 2001 20:51:15 +0000 (UTC) Hello Sorry, I am not a specialist in databases, so, perhaps, my question is simple or meaningless... We've got a relational database (PostgreSQL), which is currently VERY redundant. Ok, I know how to reduce the redundancy by splitting the table into several smaller (2-column in our case) ones. But, anyway the redundancy would only be decreased, and not removed completely, and building SQL-queries would become more difficult (nested selects, etc.), unless you can create something like a view on the top of several tables?... Anyway, the data naturally have tree-structure, so, ideally it should be kept that way (not sure if RDBMSs can perform such optimisation / redundancy removal automatically?), but I still want to be able to use SQL-queries. So, ideally, the data should be stored in a tree, but logically be representable as a table with rows spanning all tree-levels... Is this possible? and - if yes - how? Thanks Guennadi ___ Dr. Guennadi V. Liakhovetski Department of Applied Mathematics University of Sheffield, U.K. email: G.Liakhovetski@sheffield.ac.uk