Core: Add setter for firstRowId in V4 metadata by gaborkaszab · Pull Request #16849 · apache/iceberg

gaborkaszab · 2026-06-17T15:58:56Z

No description provided.

gaborkaszab · 2026-06-17T16:05:26Z

Currently, there is no mechanism to set Tracking.firstRowId. Following the PR for Tracking builder, it was commented that setting firstRowId shouldn't be through the builder but via inheritance.
This PR adds a simple setter function that can be used for this. If I understand correctly, it wouldn't suite inheritFrom(manifestTracking) because it's not a direct inheritance, but based calculated based on other manifest entries that has null firstRowId + the manifest's firstRowId.

gaborkaszab · 2026-06-17T16:06:01Z

cc @stevenzwu @rdblue @anoopj @amogh-jahagirdar

anoopj · 2026-06-17T20:47:33Z

  }

+  void setFirstRowId(Long newFirstRowId) {
+    Preconditions.checkArgument(newFirstRowId != null, "Invalid first row ID: null");


Why not just make the parameter long since it's required anyway?

+1 to make the parameter to long.

Good point! done

anoopj · 2026-06-17T20:51:57Z


+  void setFirstRowId(Long newFirstRowId) {
+    Preconditions.checkArgument(newFirstRowId != null, "Invalid first row ID: null");
+    Preconditions.checkArgument(newFirstRowId >= 0, "Invalid first row ID: negative");


Suggested change

Preconditions.checkArgument(newFirstRowId >= 0, "Invalid first row ID: negative");

Preconditions.checkArgument(

newFirstRowId >= 0, "Invalid first row ID: %s (must be >= 0)", newFirstRowId);

ebyhr · 2026-06-18T00:17:09Z

  }

+  void setFirstRowId(Long newFirstRowId) {
+    Preconditions.checkArgument(newFirstRowId != null, "Invalid first row ID: null");


+1 to make the parameter to long.

ebyhr · 2026-06-18T00:20:40Z

+  }
+
+  @Test
+  public void testSettingFirstRowId() {


How about adding a test case for setFirstRowId(0L) to demonstrate the boundary of the >= 0 check?

I changed one of the tests to use zero for input. Thx

gaborkaszab

Thanks for taking a look! I addressed all the comments. Additionally, there are 2 extra modifications I figured would make sense:

Not allow to set first row id for ADDED entries. I think for these null is the desired value
I also added a setter method to TrackedFileStruct for convenience. With this the user doesn't have to be aware of the actual representation being in Tracking and can simply invoke TrackedFileStruct.setFirstRowId() without knowing how it is represented internally.

Let me know what you think!

gaborkaszab · 2026-06-18T07:53:29Z

  }

+  void setFirstRowId(Long newFirstRowId) {
+    Preconditions.checkArgument(newFirstRowId != null, "Invalid first row ID: null");


Good point! done

gaborkaszab · 2026-06-18T08:15:35Z


+  void setFirstRowId(Long newFirstRowId) {
+    Preconditions.checkArgument(newFirstRowId != null, "Invalid first row ID: null");
+    Preconditions.checkArgument(newFirstRowId >= 0, "Invalid first row ID: negative");


gaborkaszab · 2026-06-18T08:18:46Z

+  }
+
+  @Test
+  public void testSettingFirstRowId() {


I changed one of the tests to use zero for input. Thx

github-actions Bot added the core label Jun 17, 2026

anoopj approved these changes Jun 17, 2026

View reviewed changes

ebyhr approved these changes Jun 18, 2026

View reviewed changes

gaborkaszab force-pushed the main_first_row_id_setter branch from fe964f0 to a0b0ddd Compare June 18, 2026 09:50

gaborkaszab commented Jun 18, 2026

View reviewed changes

Core: Add setter for firstRowId in V4 metadata

ab3b98a

gaborkaszab force-pushed the main_first_row_id_setter branch from a0b0ddd to ab3b98a Compare June 18, 2026 10:02

gaborkaszab changed the title ~~Core: Add setter for TrackingStruct.firstRowId~~ Core: Add setter for firstRowId in V4 metadata Jun 18, 2026

gaborkaszab requested a review from anoopj June 18, 2026 14:17

	Preconditions.checkArgument(newFirstRowId >= 0, "Invalid first row ID: negative");
	Preconditions.checkArgument(
	newFirstRowId >= 0, "Invalid first row ID: %s (must be >= 0)", newFirstRowId);

Conversation

gaborkaszab commented Jun 17, 2026

Uh oh!

gaborkaszab commented Jun 17, 2026

Uh oh!

gaborkaszab commented Jun 17, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

gaborkaszab left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants