[ https://issues.apache.org/jira/browse/CASSANDRA-3647?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13294999#comment-13294999 ]
Sylvain Lebresne commented on CASSANDRA-3647: --------------------------------------------- bq. Fair point, but I feel like a Set is more like a Map with no value, than a List with no indexes Ok, let me rephrase a bit my argumentation. # The first point is that "S[4]" for sets is not a standard notation at all. That doesn't mean we cannot use it (there is no standard notation for that kind of operation that I know of anyway), but that does mean that another notation would be equaly good. And not reusing a very standard notation for something new could actually be a good point: I've learned that we don't all have the same intuition when it comes to syntax, and I would halfway expect some people to think that S[4] actually means "the 4th elements of the (sorted) set S". # Then there is the fact that we don't have a syntax for List.discard and I do think there is no good reason not to support it (not having a syntax is definitely not a good reason as far as I'm concerned; I prefer supporting it with a slightly ugly syntax than none at all). And the thing is, List.discard is very much the same operation than Set.discard, or at least it's much closer in semantic than List.discard_idx is of Set.discard. So inventing a syntax for both List.discard and Set.discard would be coherent. So the notation I'm suggesting is to use curly brackets instead of square ones, so S{4}. The meaning of such would be to 'select the value 4 in the set S if present'. I will note that supporting such notation with that meaning would for example allow us, if we so wish, to support "SELECT S{4} ..." as a simple way to test the existence of 4 in S (I'm not saying we need to support that, but at least I think it is a bit early to say we will never support such thing). Lastly, not reusing the square bracket notation will make it clear that "SET S[4] = 3" would be non-sensical (if S is a set). > Support set and map value types in CQL > -------------------------------------- > > Key: CASSANDRA-3647 > URL: https://issues.apache.org/jira/browse/CASSANDRA-3647 > Project: Cassandra > Issue Type: New Feature > Components: API, Core > Reporter: Jonathan Ellis > Assignee: Sylvain Lebresne > Labels: cql > Fix For: 1.2 > > > Composite columns introduce the ability to have arbitrarily nested data in a > Cassandra row. We should expose this through CQL. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira