arne-bdt commented on issue #1279:
URL: https://github.com/apache/jena/issues/1279#issuecomment-1562406230
There's a straightforward way to improve the speed of `GraphMem` for
`Graph#add` by nearly 50% and `Graph#delete` by 10-35%. This change has almost
no impact on any other operation. However, it uses an additional array for the
hashCodes in `HashCommon`. As a result, the memory consumption of the indexing
structures (excluding the space for the triples) would increase by 31-51%.
I propose not to implement this for the deprecated `GraphMem`, but instead
to create a successor `GraphMem2` with the following changes:
- Speed up `Graph#add` and `Graph#delete` accepting the given memory cost
- Remove support for Iterator#remove to reduce complexity
(other implementations like `DatasetGraphInMemory` don't support this
either)
- Switch to term-equality as proposed in
[issue#1867](https://github.com/apache/jena/issues/1867#issuecomment-1546931793)
mostly by:
- Removing calls to `org.apache.jena.graph.Node#sameValueAs`
- Replacing usages of `org.apache.jena.graph.Node#getIndexingValue` with
the Node itself
What do you say?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
