[
https://issues.apache.org/jira/browse/S2GRAPH-121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15670784#comment-15670784
]
ASF GitHub Bot commented on S2GRAPH-121:
----------------------------------------
Github user daewon commented on the issue:
https://github.com/apache/incubator-s2graph/pull/96
It works well. +1
> Create `Result` class to hold traverse result edges.
> ----------------------------------------------------
>
> Key: S2GRAPH-121
> URL: https://issues.apache.org/jira/browse/S2GRAPH-121
> Project: S2Graph
> Issue Type: Improvement
> Affects Versions: 0.2.0
> Reporter: DOYUNG YOON
> Assignee: DOYUNG YOON
> Labels: refactoring
> Fix For: 0.2.0
>
> Original Estimate: 168h
> Remaining Estimate: 168h
>
> Purpose of suggestion.
> GroupBy, OrderBy, Filtering implementation of fetched result edges are spread
> out multiple parts(Storage, AsynchbaseStorage, PostProcess). It would be much
> better we can separate post process logic(OrderBy, GroupBy, FilterOut) in
> `Result` class so any caller simply can benefit single implementation.
> In general, I am suggesting changing PostProcess only call method of Result
> class, not actually implement logic.
> I was asked if we can provide binary format(not JSON) such as protobuf,
> thrift, etc and I think it is reasonable demand. While I was tapping out
> possibility of this, I found out moving post process implementation into
> `Result` class can make this really easy, I guess we can simply focus on
> actual `format` not post process logic itself.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
