[Kernel] Update maxFieldId when updating the schema (delta-io#4365)

<!--
Thanks for sending a pull request!  Here are some tips for you:
1. If this is your first time, please read our contributor guidelines:
https://github.com/delta-io/delta/blob/master/CONTRIBUTING.md
2. If the PR is unfinished, add '[WIP]' in your PR title, e.g., '[WIP]
Your PR title ...'.
  3. Be sure to keep the PR description updated to reflect all changes.
  4. Please write your PR title to summarize what this PR proposes.
5. If possible, provide a concise example to reproduce the issue for a
faster review.
6. If applicable, include the corresponding issue number in the PR title
and link it in the body.
-->

#### Which Delta project/connector is this regarding?
<!--
Please add the component selected below to the beginning of the pull
request title
For example: [Spark] Title of my pull request
-->

- [ ] Spark
- [ ] Standalone
- [ ] Flink
- [ ] Kernel
- [ ] Other (fill in here)

## Description

<!--
- Describe what this PR changes.
- Describe why we need the change.
 
If this PR resolves an issue be sure to include "Resolves #XXX" to
correctly link and close the issue upon merge.
-->

## How was this patch tested?

<!--
If tests were added, say they were added here. Please make sure to test
the changes thoroughly including negative and positive cases if
possible.
If the changes were tested in any way other than unit tests, please
clarify how you tested step by step (ideally copy and paste-able, so
that other reviewers can test and check, and descendants can verify in
the future).
If the changes were not tested, please explain why.
-->

## Does this PR introduce _any_ user-facing changes?

<!--
If yes, please clarify the previous behavior and the change this PR
proposes - provide the console output, description and/or an example to
show the behavior difference if possible.
If possible, please also clarify if this is a user-facing change
compared to the released Delta Lake versions or within the unreleased
branches such as master.
If no, write 'No'.
-->

---------

Co-authored-by: Micah Kornfield <[email protected]>

Author: allisonport-db

kernel/kernel-api/src/main/java/io/delta/kernel/internal/util/ColumnMapping.java

@@ -326,6 +326,10 @@ private static Optional<Metadata> assignColumnIdAndPhysicalName(
       Metadata metadata, boolean isNewTable) {
     StructType oldSchema = metadata.getSchema();
 
+    // This is the maxColumnId to use when assigning any new field-ids; we update this as we
+    // traverse the schema and after traversal this is the value that should be stored in the
+    // metadata. Note - this could be greater than the current value stored in the metadata if
+    // the connector has added new fields with field-ids
     AtomicInteger maxColumnId =
         new AtomicInteger(
             Math.max(
@@ -335,8 +339,6 @@ private static Optional<Metadata> assignColumnIdAndPhysicalName(
                         .getOrDefault(COLUMN_MAPPING_MAX_COLUMN_ID_KEY, "0")),
                 findMaxColumnId(oldSchema)));
 
-    int oldMaxColumnId = maxColumnId.get();
-
     StructType newSchema = new StructType();
     for (StructField field : oldSchema.fields()) {
       newSchema =
@@ -354,17 +356,16 @@ private static Optional<Metadata> assignColumnIdAndPhysicalName(
       newSchema = rewriteFieldIdsForIceberg(newSchema, maxColumnId);
     }
 
+    // The maxColumnId in the metadata may be out-of-date either due to field-id assignment
+    // performed in this fx, or due to connector adding new fields
+    boolean shouldUpdateMaxId =
+        TableConfig.COLUMN_MAPPING_MAX_COLUMN_ID.fromMetadata(metadata) != maxColumnId.get();
+
     // We are comparing the old schema with the new schema to determine if the schema has changed.
     // If this becomes hotspot, we can consider updating the methods to pass around AtomicBoolean
     // to track if the schema has changed. It is a bit convoluted to pass around and update the
     // AtomicBoolean in the recursive and multiple methods.
-    if (oldSchema.equals(newSchema)) {
-      checkArgument(
-          oldMaxColumnId == maxColumnId.get(),
-          "The schema hasn't changed but the max column id has changed from %s to %s",
-          oldMaxColumnId,
-          maxColumnId.get());
-
+    if (oldSchema.equals(newSchema) && !shouldUpdateMaxId) {
       return Optional.empty();
     }
 

kernel/kernel-api/src/test/scala/io/delta/kernel/internal/util/ColumnMappingSuite.scala

@@ -17,6 +17,8 @@ package io.delta.kernel.internal.util
 
 import java.util
 
+import scala.collection.JavaConverters.mapAsJavaMapConverter
+
 import io.delta.kernel.exceptions.KernelException
 import io.delta.kernel.expressions.Column
 import io.delta.kernel.internal.actions.Metadata
@@ -607,6 +609,12 @@ class ColumnMappingSuite extends AnyFunSuite with ColumnMappingSuiteBase {
     if (enableIcebergCompatV2) {
       metadata = metadata.withIcebergCompatV2Enabled
     }
+    if (!metadata.getConfiguration.containsKey(COLUMN_MAPPING_MAX_COLUMN_ID_KEY)) {
+      // A hack, if the metadata doesn't have max column ID in it,
+      // then new metadata is always returned.
+      metadata =
+        metadata.withMergedConfiguration(Map(COLUMN_MAPPING_MAX_COLUMN_ID_KEY -> "100").asJava)
+    }
 
     assertThat(updateColumnMappingMetadataIfNeeded(metadata, isNewTable)).isEmpty
   }

kernel/kernel-defaults/src/test/scala/io/delta/kernel/defaults/DeltaColumnMappingSuite.scala

@@ -21,8 +21,8 @@ import scala.collection.immutable.Seq
 import io.delta.kernel.Table
 import io.delta.kernel.exceptions.InvalidConfigurationValueException
 import io.delta.kernel.internal.TableConfig
-import io.delta.kernel.internal.util.ColumnMappingSuiteBase
-import io.delta.kernel.types.{IntegerType, StringType, StructType}
+import io.delta.kernel.internal.util.{ColumnMapping, ColumnMappingSuiteBase}
+import io.delta.kernel.types.{FieldMetadata, IntegerType, StringType, StructField, StructType}
 
 class DeltaColumnMappingSuite extends DeltaTableWriteSuiteBase with ColumnMappingSuiteBase {
 
@@ -86,11 +86,33 @@ class DeltaColumnMappingSuite extends DeltaTableWriteSuiteBase with ColumnMappin
       assertColumnMapping(structType.get("a"), 1)
       assertColumnMapping(structType.get("b"), 2)
 
+      assert(TableConfig.COLUMN_MAPPING_MAX_COLUMN_ID.fromMetadata(getMetadata(
+        engine,
+        tablePath)) == 2)
+
       val protocol = getProtocol(engine, tablePath)
       assert(protocol.getMinReaderVersion == 2 && protocol.getMinWriterVersion == 7)
     }
   }
 
+  test("new table with existing column mappings in schema writes COLUMN_MAPPING_MAX_COLUMN_ID") {
+    withTempDirAndEngine { (tablePath, engine) =>
+      val props = Map(TableConfig.COLUMN_MAPPING_MODE.getKey -> "id")
+      val fieldMetadata = FieldMetadata.builder()
+        .putLong(ColumnMapping.COLUMN_MAPPING_ID_KEY, 1)
+        .putString(ColumnMapping.COLUMN_MAPPING_PHYSICAL_NAME_KEY, "col-0").build()
+      val structField = new StructField("col_name", IntegerType.INTEGER, false, fieldMetadata)
+      val schema = new StructType(Seq(structField).asJava)
+      createEmptyTable(engine, tablePath, schema, tableProperties = props)
+
+      val structtype = getMetadata(engine, tablePath).getSchema
+      assertColumnMapping(structtype.get("col_name"), 1)
+      assert(TableConfig.COLUMN_MAPPING_MAX_COLUMN_ID.fromMetadata(getMetadata(
+        engine,
+        tablePath)) == 1)
+    }
+  }
+
   test("can update existing table to column mapping mode = name") {
     withTempDirAndEngine { (tablePath, engine) =>
       createEmptyTable(engine, tablePath, simpleTestSchema)

kernel/kernel-defaults/src/test/scala/io/delta/kernel/defaults/DeltaTableSchemaEvolutionSuite.scala

@@ -32,7 +32,7 @@ import io.delta.kernel.utils.CloseableIterable.emptyIterable
  */
 class DeltaTableSchemaEvolutionSuite extends DeltaTableWriteSuiteBase with ColumnMappingSuiteBase {
 
-  test("Add nullable column succeeds") {
+  test("Add nullable column succeeds and correctly updates maxFieldId") {
     withTempDirAndEngine { (tablePath, engine) =>
       val table = Table.forPath(engine, tablePath)
       val initialSchema = new StructType()
@@ -70,6 +70,7 @@ class DeltaTableSchemaEvolutionSuite extends DeltaTableWriteSuiteBase with Colum
       assertColumnMapping(innerStruct.get("d"), 4, "d")
       assertColumnMapping(innerStruct.get("e"), 5, "e")
       assertColumnMapping(structType.get("c"), 2)
+      assert(getMaxFieldId(engine, tablePath) == 5)
     }
   }
 
@@ -87,6 +88,7 @@ class DeltaTableSchemaEvolutionSuite extends DeltaTableWriteSuiteBase with Colum
         tableProperties = Map(
           TableConfig.COLUMN_MAPPING_MODE.getKey -> "id",
           TableConfig.ICEBERG_COMPAT_V2_ENABLED.getKey -> "true"))
+      assertColumnMapping(table.getLatestSnapshot(engine).getSchema.get("c"), 2)
 
       val currentSchema = table.getLatestSnapshot(engine).getSchema()
       val newSchema = new StructType()
@@ -98,6 +100,7 @@ class DeltaTableSchemaEvolutionSuite extends DeltaTableWriteSuiteBase with Colum
 
       val structType = table.getLatestSnapshot(engine).getSchema
       assertColumnMapping(structType.get("a"), 1)
+      assert(getMaxFieldId(engine, tablePath) == 2)
     }
   }
 
@@ -1209,4 +1212,9 @@ class DeltaTableSchemaEvolutionSuite extends DeltaTableWriteSuiteBase with Colum
     assert(e.isInstanceOf[T])
     assert(e.getMessage.contains(expectedMessageContained))
   }
+
+  private def getMaxFieldId(engine: Engine, tablePath: String): Long = {
+    TableConfig.COLUMN_MAPPING_MAX_COLUMN_ID
+      .fromMetadata(getMetadata(engine, tablePath))
+  }
 }