Community
Participate
Working Groups
Hi, Would you please consider adding support for Trove (http://trove.starlight-systems.com/) as it is a popular collections library. I wouldn't mind looking into this if you could point me towards the right direction. Thanks
The collections code is in the org.eclipse.mat.api plug-in, in the org.eclipse.mat.inspections.collections package. The CollectionsUtil class abstracts some of the differences between the Oracle and IBM collection classes, so that is the first place to look. It's not an extension point, so you would need to modify the code, rather than just add a new plug-in. If the Trove collections are significantly different in implementation it may be hard to fit them into the framework, but we already have the problem of IdentityHashMap which doesn't fit. You might then need to modify the individual queries in that package. If we have further collection libraries to support then we might need to abstract the operations on collections in a more general manner. It could also be nice to allow access to collection objects from other queries or perhaps OQL. If you would like to contribute an enhancement to Memory Analyzer for Trove then that would be great. One thing to be careful with is to ensure the code is your own, and not derived from Trove, which is under the GPL and so is not compatible with the Eclipse Public License.
There hasn't been any change to the issue for over 10 years. I am closing now.