Spark: Supports creating a branch on an empty table #8072
Conversation
|
Hi @rdblue @jackye1995, I see you have discussed this at here #7075, could you help to review this when you are free? Thanks a lot. |
|
+1 for this change. This will solve this issue #7593 at its root. |
.palantir/revapi.yml
Outdated
| "1.3.0": | ||
| org.apache.iceberg:iceberg-api: | ||
| - code: "java.method.addedToInterface" | ||
| new: "method org.apache.iceberg.ManageSnapshots org.apache.iceberg.ManageSnapshots::createBranch(java.lang.String)" |
There was a problem hiding this comment.
I don't think we want to break the API just by adding a new method. Rather than breaking the API, we typically add a default implementation that throws an UOE, similar to how it's done in Snapshot#removedDeleteFiles
ConeyLiu
force-pushed
the
create-empty-branch
branch
from
e267ac3 to
d55339f
Compare
|
@nastra would you mind taking a look at this again? Thanks a lot. |
|
@amogh-jahagirdar and @jackye1995, FYI |
| * | ||
| * @param name branch name | ||
| * @return this for method chaining | ||
| * @throws IllegalArgumentException if a branch with the given name already exists |
There was a problem hiding this comment.
Would be nice to have a test for this.
There was a problem hiding this comment.
Let me do a follow-up.
|
Thanks, @ConeyLiu! |
|
Thanks @rdblue for merging this. And also thanks @nastra @zhangbutao for reviewing. |
This change backports PR #8072 to Spark 3.1, 3.2, 3.3. Co-authored-by: xianyangliu <[email protected]>
4 participants
Currently, we only support creating branches with a given snapshot ID. However, there is no snapshot for an empty table. This PR achieves this by creating a newly empty snapshot for an empty table. And binding the branch to the new snapshot.