[
https://issues.apache.org/jira/browse/TINKERPOP3-851?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14906798#comment-14906798
]
Marko A. Rodriguez commented on TINKERPOP3-851:
-----------------------------------------------
We shouldn't do this with a {{groupCount}}-by modulator. In general, I don't
think a second {{by}} modulator is good at all given {{Long}}. What we need is
a way to determine from the third reduce-{{by}}-modulator of the more general
{{GroupStep}} if we can do the reduction "on the fly."
Also, perhaps we can get rid of {{GroupCountStep}} as it is a special case of
{{GroupStep}} and thus, reduce the instruction set.
> GroupCountStep needs a by() for the count.
> ------------------------------------------
>
> Key: TINKERPOP3-851
> URL: https://issues.apache.org/jira/browse/TINKERPOP3-851
> Project: TinkerPop 3
> Issue Type: New Feature
> Components: process
> Reporter: Marko A. Rodriguez
> Assignee: Marko A. Rodriguez
> Fix For: 3.1.0-incubating
>
>
> Currently {{groupCount()}} takes one {{by}}-modulator. That {{by()}} is the
> key-projection. In TP2, we had a second "by modulator" that was the number to
> "incr." In TP3, the number to incr is currently {{traverser.bulk()}}. We
> should keep this default, but we should support a user defined second
> {{by()}} modulator.
> {code}
> g.V.out.groupCount.by('name').by(sack())
> g.V.out.groupCount.by('name').by('age')
> {code}
> This is necessary for 3.1.0 as it is necessary for the "signal processing"
> features available in the new sack-work.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
