Re: Ragtag SE Questions

From: mathprof@bigfoot.com > >I had the following 'ragtag' questions related to Informix-SE: > >*Informix's SE doesn't include a lowercase() procedure; how do I write >one? Is there any place I can download several useful Informix-SE >procedures all at once (eg, some sort of "utility pack" for >Informix-SE)? There should be one at http://www.iiug.org >*I have a table where a given column is "almost" unique-- the only >value that can be repeated is the null value. In other words, the >table can have arbitrarily many rows with that column set to null, >but, once those rows are removed, the column is unique for the >remaining rows. How do I implement this? Beats me. :-) Don't think you can. >*Is there anyway to implement "invisible rows" in Informix-SE? These >are rows that wouldn't show up in SELECT statements unless >specifically requested. For example: > >SELECT * FROM mytab;> >would show the normal rows, while: > >SELECT * FROM mytab INCLUDE INVISIBLE;> >would show the normal rows PLUS the invisible rows. I think something >like this would be useful if you wanted to "delete" a row, but not >really delete it-- ie, be able to restore it if necessary. In other >words, something like: > >INVISIBLIZE FROM mytab where x<15; > >wouldn't actually delete the rows where x<15, it would just make them >invisible so that they wouldn't show up in normal SELECTs. Create a column called Invisible in your table; Create an SP or SQL to update that column based on your criteria; Create several views on your table: all, invisible only and visible only; Users can now select from the views, not the base table. >*Someone mentioned I could store "backup" copies of a given table at >various times using "schemas". How do I do this? What I'd really like >to do is have the table know how it looked before and after *every* >transaction. The only way I can think of doing this right now is to >write a TRIGGER that automatically does: > >UNLOAD TO '/tmp/foo.tabname' SELECT * FROM tabname> >[and even this isn't 100% accurate], and then does some sort of 'diff' >between this copy of '/tmp/foo.tabname' and the previously saved >one. That way, I'd have a list of 'patches' of how I got from the >original table to where I am today. But this seems horrendously ugly-- >there must be a better way? Let me know when you find it! :-) ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com