Github user dimas-b commented on a diff in the pull request:
https://github.com/apache/tinkerpop/pull/899#discussion_r206547429
--- Diff:
gremlin-server/src/main/java/org/apache/tinkerpop/gremlin/server/op/AbstractOpProcessor.java
---
@@ -252,15 +264,29 @@ protected boolean isForceFlushed(final
ChannelHandlerContext ctx, final RequestM
/**
* @deprecated As of release 3.2.2, replaced by {@link
#makeFrame(ChannelHandlerContext, RequestMessage, MessageSerializer, boolean,
List, ResponseStatusCode, Map)}.
*/
+ @Deprecated
protected static Frame makeFrame(final ChannelHandlerContext ctx,
final RequestMessage msg,
final MessageSerializer serializer,
final boolean useBinary, final List<Object> aggregate,
final ResponseStatusCode code) throws
Exception {
return makeFrame(ctx, msg, serializer, useBinary, aggregate, code,
Collections.emptyMap());
}
+ /**
+ * Caution: {@link #makeFrame(ResponseHandlerContext, RequestMessage,
MessageSerializer, boolean, List, ResponseStatusCode, Map)}
+ * should be used instead of this method whenever a {@link
ResponseHandlerContext} is available.
+ */
protected static Frame makeFrame(final ChannelHandlerContext ctx,
final RequestMessage msg,
--- End diff --
I did not want to force other classes to use the variant with
`ResponseHandlerContext` because it is not strictly required unless the request
is handled by `AbstractEvalOpProcessor`.
Please also see my comment on `evalOpInternal`.
If `evalOpInternal` is `private` the new `makeFrame` might as well be
`private`, in which case I'd rather remove the "caution" comment.
---
