(iceberg) branch main updated: Core: Fix metadata table scans with useRef by preserving metadata schema (#15276)

[etudenhoefner]

This is an automated email from the ASF dual-hosted git repository.

etudenhoefner pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/iceberg.git


The following commit(s) were added to refs/heads/main by this push:
     new c75efa9c32 Core: Fix metadata table scans with useRef by preserving 
metadata schema (#15276)
c75efa9c32 is described below

commit c75efa9c32d85bc1c804a3909a28048fc70c6833
Author: Rui Li <li...@apache.org>
AuthorDate: Tue Feb 10 21:15:24 2026 +0800

    Core: Fix metadata table scans with useRef by preserving metadata schema 
(#15276)
---
 .../main/java/org/apache/iceberg/SnapshotScan.java |  3 +-
 .../org/apache/iceberg/TestMetadataTableScans.java | 76 ++++++++++++++++++++++
 2 files changed, 78 insertions(+), 1 deletion(-)

diff --git a/core/src/main/java/org/apache/iceberg/SnapshotScan.java 
b/core/src/main/java/org/apache/iceberg/SnapshotScan.java
index 8dd1e7df20..655c4c72e2 100644
--- a/core/src/main/java/org/apache/iceberg/SnapshotScan.java
+++ b/core/src/main/java/org/apache/iceberg/SnapshotScan.java
@@ -124,7 +124,8 @@ public abstract class SnapshotScan<ThisT, T extends 
ScanTask, G extends ScanTask
     Snapshot snapshot = table().snapshot(name);
     Preconditions.checkArgument(snapshot != null, "Cannot find ref %s", name);
     TableScanContext newContext = 
context().useSnapshotId(snapshot.snapshotId());
-    return newRefinedScan(table(), SnapshotUtil.schemaFor(table(), name), 
newContext);
+    Schema newSchema = useSnapshotSchema() ? SnapshotUtil.schemaFor(table(), 
name) : tableSchema();
+    return newRefinedScan(table(), newSchema, newContext);
   }
 
   public ThisT asOfTime(long timestampMillis) {
diff --git a/core/src/test/java/org/apache/iceberg/TestMetadataTableScans.java 
b/core/src/test/java/org/apache/iceberg/TestMetadataTableScans.java
index 4d2971c7f1..210dfb3683 100644
--- a/core/src/test/java/org/apache/iceberg/TestMetadataTableScans.java
+++ b/core/src/test/java/org/apache/iceberg/TestMetadataTableScans.java
@@ -1752,6 +1752,82 @@ public class TestMetadataTableScans extends 
MetadataTableScanTestBase {
     assertEstimatedRowCount(new AllEntriesTable(table), 4);
   }
 
+  @TestTemplate
+  public void testMetadataTableScansOnBranch() throws IOException {
+    table.newFastAppend().appendFile(FILE_A).commit();
+    table.manageSnapshots().createBranch("testBranch").commit();
+    table.newFastAppend().appendFile(FILE_B).commit();
+
+    Table entriesTable = new ManifestEntriesTable(table);
+    assertThat(
+            rowCount(
+                entriesTable
+                    .newScan()
+                    .useRef("testBranch")
+                    .select("status")
+                    
.filter(Expressions.lessThan("data_file.partition.data_bucket", 0))))
+        .as("ManifestEntriesTable on branch should have no entry satisfying 
the filter")
+        .isEqualTo(0);
+    assertThat(rowCount(entriesTable.newScan()))
+        .as("ManifestEntriesTable on main should have 2 entries")
+        .isEqualTo(2);
+
+    Table dataFilesTable = new DataFilesTable(table);
+    assertThat(rowCount(dataFilesTable.newScan().useRef("testBranch")))
+        .as("DataFilesTable on branch should have 1 file")
+        .isEqualTo(1);
+    assertThat(rowCount(dataFilesTable.newScan()))
+        .as("DataFilesTable on main should have 2 files")
+        .isEqualTo(2);
+
+    Table manifestsTable = new ManifestsTable(table);
+    assertThat(rowCount(manifestsTable.newScan().useRef("testBranch")))
+        .as("ManifestsTable on branch should have 1 manifest")
+        .isEqualTo(1);
+    assertThat(rowCount(manifestsTable.newScan()))
+        .as("ManifestsTable on main should have 2 manifests")
+        .isEqualTo(2);
+
+    Table filesTable = new FilesTable(table);
+    assertThat(rowCount(filesTable.newScan().useRef("testBranch")))
+        .as("FilesTable on branch should have 1 file")
+        .isEqualTo(1);
+    assertThat(rowCount(filesTable.newScan()))
+        .as("FilesTable on main should have 2 files")
+        .isEqualTo(2);
+  }
+
+  @TestTemplate
+  public void testDeleteFilesTableScanOnBranch() throws IOException {
+    assumeThat(formatVersion).as("Delete files are not supported by V1 
Tables").isGreaterThan(1);
+
+    table.newFastAppend().appendFile(FILE_A).commit();
+    table.newRowDelta().addDeletes(fileADeletes()).commit();
+    table.manageSnapshots().createBranch("testBranch").commit();
+    table.newFastAppend().appendFile(FILE_B).commit();
+    table.newRowDelta().addDeletes(fileBDeletes()).commit();
+
+    Table deleteFilesTable = new DeleteFilesTable(table);
+    assertThat(rowCount(deleteFilesTable.newScan().useRef("testBranch")))
+        .as("DeleteFilesTable on branch should have 1 delete file")
+        .isEqualTo(1);
+    assertThat(rowCount(deleteFilesTable.newScan()))
+        .as("DeleteFilesTable on main should have 2 delete files")
+        .isEqualTo(2);
+  }
+
+  private int rowCount(TableScan scan) throws IOException {
+    int count = 0;
+    try (CloseableIterable<FileScanTask> tasks = scan.planFiles()) {
+      for (FileScanTask task : tasks) {
+        try (CloseableIterable<StructLike> rows = task.asDataTask().rows()) {
+          count += Iterators.size(rows.iterator());
+        }
+      }
+    }
+    return count;
+  }
+
   private void assertEstimatedRowCount(Table metadataTable, int size) throws 
Exception {
     TableScan scan = metadataTable.newScan();