Thanks to Joseph Adler's contribution, we have a new AvroStorage ready.
Although there are additional requests that we would like to address, I
think we can implement them incrementally after we commit the current
patches. As of now,
- The core features are fully implemented.
- All the unit tests are passing with both Hadoop 1.x and 2.x.
- The documentation is added:
Since I modified the patches several times, I cannot give +1 by myself. So
I am wondering if another committer can review it. The latest patches are
In fact, Dmitriy asked in the jira whether or not we should add AvroStorage
to the core Pig. Although I prefer to add it to the core Pig, I am open to
Please let me know what you think.