[
https://issues.apache.org/jira/browse/TINKERPOP-1924?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16408155#comment-16408155
]
stephen mallette commented on TINKERPOP-1924:
---------------------------------------------
hmmm - didn't realize that you specifically wanted that kind of output. Seems a
bit of an odd way to want shape your results though, no? is there a specific
use case you could share? in other words, is there some reason that you would
want to return that mix of ids (and possibly labels) and property data as a
stream?
for fun:
{code}
gremlin> g.V().project('name','id').by('name').by(id).select(values).unfold()
==>marko
==>1
==>vadas
==>2
==>lop
==>3
==>josh
==>4
==>ripple
==>5
==>peter
==>6
{code}
obviously, you're {{union()}} approach is nicer. :)
> Make meta-properties accessible via values(...)-step
> ----------------------------------------------------
>
> Key: TINKERPOP-1924
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1924
> Project: TinkerPop
> Issue Type: Improvement
> Components: structure
> Affects Versions: 3.3.1
> Reporter: Daniel Weber
> Priority: Minor
>
> I'm currently running into a situation where I'd like to access
> meta-properties through the values(...)-method. Unfortunately, the signature
> of values() only takes strings, so this is not possible:
> {code:java}
> g.V().values(T.id)
> {code}
> Of course, there's the trivial workaround:
> {code:java}
> g.V().id()
> {code}
> However, I'd like to get values for multiple properties, so the workaround
> becomes ugly:
> {code:java}
> g.V().union(__.id(), __.values('key'))
> {code}
> That there would be no overload of values(...) supporting T-accessors is a
> bit odd. has(...) has an extra overload for T-accessors, property(...) allows
> arbitrary objects to be passed in as key. The valueMap-method allows
> including T-accessors in the returned map.
> Would it be possible to relax the signature of values(...) to take an
> arbitrary list of objects instead of property keys of type string? If the
> discussion here turns out in favor of such a relaxation, I could try to
> implement the change.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
