[
https://issues.apache.org/jira/browse/TINKERPOP-2302?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16941629#comment-16941629
]
Dan LaRocque commented on TINKERPOP-2302:
-----------------------------------------
I've opened some PRs with the narrow approach. I started from tp34, then
merged forward to master. Please let me know if the PRs need changes.
> Add isOnGraphComputer() field accessor to ElementMapStep
> --------------------------------------------------------
>
> Key: TINKERPOP-2302
> URL: https://issues.apache.org/jira/browse/TINKERPOP-2302
> Project: TinkerPop
> Issue Type: Improvement
> Components: process
> Affects Versions: 3.5.0, 3.4.4
> Reporter: Dan LaRocque
> Priority: Minor
>
> TINKERPOP-2284 added {{ElementMapStep}}.
> This step implements {{GraphComputing}}, so it defines that interface's
> {{onGraphComputer()}} method. The method sets a private boolean field.
> For any implementation that wants to alter {{ElementMapStep}} behavior by
> replacing instances of the step with an alternative implementation (maybe a
> subclass), it could be convenient to make the private {{onGraphComputer}}
> state accessible. I think the rest of {{ElementMapStep}}'s public state is
> already accessible; it's just this field that is private.
> There are a few ways to go about this.
> * We could add an {{isOnGraphComputer()}} method to just {{ElementMapStep}},
> or make {{onGraphComputer}} protected. This is the narrowest version of the
> change.
> * We could add an {{isOnGraphComputer()}} method to the {{GraphComputing}}
> interface. This would affect a bunch of implementing steps. Most
> implementations on tp34 have an internal boolean field storing this state, so
> it would be easy on those. However, [{{GraphStep}} does not have a boolean
> onGraphComputer
> field|https://github.com/apache/tinkerpop/blob/tp34/gremlin-core/src/main/java/org/apache/tinkerpop/gremlin/process/traversal/step/map/GraphStep.java#L131-L135]:
> {code}
> @Override
> public void onGraphComputer() {
> this.iteratorSupplier = Collections::emptyIterator;
> convertElementsToIds();
> }
> {code}
> We could maybe implement this without adding another field, but it would
> probably be simpler to add a field than to infer it from existing state.
> *Altering this interface would also break any third-party implementations of
> {{GraphComputing}}.*
> * We could create a {{GraphComputing}} subinterface that adds an
> {{isOnGraphComputer()}}, and only implement it where a preexisting boolean
> field makes the job easy. This somewhere in between the first two approaches
> in terms of scope.
> I'm going to open a PR taking the first, narrowest possible approach, where
> I'm only touching {{ElementMapStep}}. I'd be happy to rework it to one of
> the alternatives, if desired.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
