Re: [HACKERS] syntax for drop if exists
Tom Lane wrote: Andrew Dunstan [EMAIL PROTECTED] writes: The MySQL syntax is actually drop table if exists foo Implementing this unfortunately generates a shift/reduce conflict, What did you try exactly? I don't see any fundamental reason for a conflict here. You may just need to rearrange the grammar to postpone the reduction a bit. You're right, as usual. I had factored out the IF EXISTS bit into a seperate rule. When I undid that and instead used 2 rules for DropStmt, the problem disappeared. (This is because it gives bison more info about the context of each IF - this has often caught me with bison - I should have known better). cheers andrew ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
[HACKERS] syntax for drop if exists
I was just looking briefly at doing drop if exists as we discussed recently. The MySQL syntax is actually drop table if exists foo Implementing this unfortunately generates a shift/reduce conflict, unless I put IF in the func_name_keyword list, which strikes me as a bad idea. Alternatively, we could use the syntax drop if exists table foo ... which seems more natural to me, and generates no conflict. Or we could live with the conflict, which I think would be harmless unless you wanted to delete a table called if, in which case you might need to say drop table if exists if ;-) I'm inclined to live with it, annoying as it is. I looked around to see what other DBs do - but AFAICS most don't support this. Thoughts? cheers andrew ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
Re: [HACKERS] syntax for drop if exists
Andrew Dunstan [EMAIL PROTECTED] writes: The MySQL syntax is actually drop table if exists foo Implementing this unfortunately generates a shift/reduce conflict, What did you try exactly? I don't see any fundamental reason for a conflict here. You may just need to rearrange the grammar to postpone the reduction a bit. Or we could live with the conflict, Utterly unacceptable; see previous discussions. regards, tom lane ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
