Peter Eisentraut wrote: > Joe Conway writes: > >>2) Implement the following new builtin functions >> array_push(anyarray, anyelement) returns anyarray >> array_pop(anyarray) returns anyelement >> array_subscript(anyarray, int) yields anyelement >> singleton_array(anyelement) returns anyarray >> - any reason not to call this one simply "array"? >> split(text, text) returns text[] >> - split string into array on delimiter >> implode(text[], text) returns text >> - join array elements into a string using given string delimiter
I think this goes too far. It is just an invitation to people to create bad database designs by using arrays as lists. Create an array support package on gborg if you like, but I feel this should not be in the mainline.
Sorry, I don't agree with that assessment. There are lots of ways people can create bad database designs using the CREATE TABLE statement too ;-)
Arrays *do* have a place, and they are supported in SQL99+. We get complaints almost daily regarding the poor array support -- this is just a step toward improving that. If you want to debate the merit of individual functions on that list, let's do that, but to dismiss them all with a wave-of-the-hand is too far.
Joe
---------------------------(end of broadcast)--------------------------- TIP 5: Have you checked our extensive FAQ?
http://www.postgresql.org/docs/faqs/FAQ.html
