Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
Author |
Topic |
giantg
Starting Member
1 Post |
Posted - 2002-07-19 : 14:16:10
|
First off Rob, I would like to say how much I like this design for a hierarchy within a relational database. I've seen other's that use many more tables (i.e. not adjacent).. I've been trying to implement this system on MS SQL Server and I have come up with an ADD procedure.. I believe MOVE and PROMOTE will be pretty trivial, but I am stuck on the DELETE procedure. The only way I can figure to even start it is to use recursion. It seems that would be the only way to make sure no nodes are truncated (i.e. no parent, not root, just floating). Or would it be better to take this approach:If a node with children is deleted, then the children will become children of that nodes parent...i.e. (root) (root) | | / \ ----- / \ / | \ / \ Delete node 2 / | \ (node 1) (node 2) ============> / | | | / | ----- / \ / | --- / \ / | - / \ / | / \ (node 1) (node 3) (node 4) (node 3) (node 4) | | | | | ----- ----- ----- ----- ----- --- --- --- --- --- - - - - ---------Key---------+ | |----- = Terminator | --- | - |--------------------+ Any insight is appreciated, Thanks. -Greg |
|
|
|
|