You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've been working on SJF4J and would be interested in adding it to the comparison. Before putting together a PR, I wanted to check whether new mapper implementations are still being accepted.
Yes, this benchmark is still being maintained. There hasn't been a new Java mapping library for a while, so it hasn't changed much.
I took a look at https://sjf4j.org/. I'm not sure that we could compare SJF4J with MapStruct, for example. Am I missing something?
I'm very glad to hear that. Thanks for the clarification.
You're not missing anything. Historically, SJF4J has focused on JSON-oriented data modeling and processing rather than object mapping itself. However, once you start working with JSON-semantic object graphs, data mapping becomes a fairly natural next step.
Since SJF4J is built around a rather unique Object-Based Node Tree (OBNT) model, existing mapping frameworks cannot fully leverage or understand that model. As a result, implementing a dedicated mapping solution within SJF4J became the most practical approach.
The mapping module is currently under development, so the design is still evolving.
If things go as planned, I may come back with a PR in the near future, and we can then see whether SJF4J fits the scope of the benchmark and makes for an interesting comparison alongside the existing mapping frameworks.
We could do what you suggested.
Yesterday, I upgraded to the minimum required version of Java for the benchmark: Java 17.
I hope to see your pull request in a few weeks.
Activity
arey commentedon Jun 1, 2026
Hi @hannyu
Yes, this benchmark is still being maintained. There hasn't been a new Java mapping library for a while, so it hasn't changed much.
I took a look at https://sjf4j.org/. I'm not sure that we could compare SJF4J with MapStruct, for example. Am I missing something?
Regards
Antoine
hannyu commentedon Jun 2, 2026
Hi @arey ,
I'm very glad to hear that. Thanks for the clarification.
You're not missing anything. Historically, SJF4J has focused on JSON-oriented data modeling and processing rather than object mapping itself. However, once you start working with JSON-semantic object graphs, data mapping becomes a fairly natural next step.
Since SJF4J is built around a rather unique Object-Based Node Tree (OBNT) model, existing mapping frameworks cannot fully leverage or understand that model. As a result, implementing a dedicated mapping solution within SJF4J became the most practical approach.
The mapping module is currently under development, so the design is still evolving.
If things go as planned, I may come back with a PR in the near future, and we can then see whether SJF4J fits the scope of the benchmark and makes for an interesting comparison alongside the existing mapping frameworks.
Thanks again!
arey commentedon Jun 4, 2026
Hi @hannyu
We could do what you suggested.
Yesterday, I upgraded to the minimum required version of Java for the benchmark: Java 17.
I hope to see your pull request in a few weeks.
Regards
Antoine
hannyu commentedon Jun 9, 2026
Hi @arey
I am back with #32 ✌️
hannyu commentedon Jun 13, 2026
Hi @arey
Thanks for merging the branch.
Does the main README still need to be updated to include SJF4J in the benchmark results, or is that handled automatically?
arey commentedon Jun 21, 2026
Hi @hannyu I had a little work in order the README.
That's done: https://github.com/arey/java-object-mapper-benchmark
Welcome SJF4J to the second rank 👍