Fix bug Delta column operations erases its properties on uppercase names

Also, show column properties correctly when column name contains uppercase characters and stores the exact column name in column statistics. Co-Authored-By: Slawomir Pajak <slawomir.pajak@starburstdata.com>
trinodb · Aug 7, 2023 · df9140a · df9140a
1 parent 6728ac8
commit df9140a
Show file tree

Hide file tree

Showing 21 changed files with 437 additions and 96 deletions.
diff --git a/...n/trino-delta-lake/src/main/java/io/trino/plugin/deltalake/AbstractDeltaLakePageSink.java b/...n/trino-delta-lake/src/main/java/io/trino/plugin/deltalake/AbstractDeltaLakePageSink.java
@@ -62,7 +62,6 @@
 import static io.trino.plugin.deltalake.DeltaLakeSessionProperties.getParquetWriterBlockSize;
 import static io.trino.plugin.deltalake.DeltaLakeSessionProperties.getParquetWriterPageSize;
 import static io.trino.plugin.deltalake.DeltaLakeTypes.toParquetType;
-import static io.trino.plugin.deltalake.transactionlog.TransactionLogAccess.canonicalizeColumnName;
 import static io.trino.plugin.hive.util.HiveUtil.escapePathName;
 import static java.lang.Math.min;
 import static java.lang.String.format;
@@ -146,21 +145,18 @@ public AbstractDeltaLakePageSink(
         ImmutableList.Builder<Type> dataColumnTypes = ImmutableList.builder();
         ImmutableList.Builder<String> dataColumnNames = ImmutableList.builder();
 
-        Map<String, String> canonicalToOriginalPartitionColumns = new HashMap<>();
-        Map<String, Integer> canonicalToOriginalPartitionPositions = new HashMap<>();
+        Map<String, Integer> toOriginalPartitionPositions = new HashMap<>();
         int partitionColumnPosition = 0;
         for (String partitionColumnName : originalPartitionColumns) {
-            String canonicalizeColumnName = canonicalizeColumnName(partitionColumnName);
-            canonicalToOriginalPartitionColumns.put(canonicalizeColumnName, partitionColumnName);
-            canonicalToOriginalPartitionPositions.put(canonicalizeColumnName, partitionColumnPosition++);
+            toOriginalPartitionPositions.put(partitionColumnName, partitionColumnPosition++);
         }
         for (int inputIndex = 0; inputIndex < inputColumns.size(); inputIndex++) {
             DeltaLakeColumnHandle column = inputColumns.get(inputIndex);
             switch (column.getColumnType()) {
                 case PARTITION_KEY:
-                    int partitionPosition = canonicalToOriginalPartitionPositions.get(column.getColumnName());
+                    int partitionPosition = toOriginalPartitionPositions.get(column.getColumnName());
                     partitionColumnInputIndex[partitionPosition] = inputIndex;
-                    originalPartitionColumnNames[partitionPosition] = canonicalToOriginalPartitionColumns.get(column.getColumnName());
+                    originalPartitionColumnNames[partitionPosition] = column.getColumnName();
                     partitionColumnTypes[partitionPosition] = column.getBaseType();
                     break;
                 case REGULAR:

diff --git a/plugin/trino-delta-lake/src/main/java/io/trino/plugin/deltalake/DeltaLakeColumnMetadata.java b/plugin/trino-delta-lake/src/main/java/io/trino/plugin/deltalake/DeltaLakeColumnMetadata.java
@@ -20,21 +20,22 @@
 import java.util.OptionalInt;
 
 import static com.google.common.base.MoreObjects.toStringHelper;
-import static java.util.Locale.ENGLISH;
 import static java.util.Objects.requireNonNull;
 
 public class DeltaLakeColumnMetadata
 {
     private final ColumnMetadata columnMetadata;
+    private final String name;
     private final OptionalInt fieldId;
     private final String physicalName;
     private final Type physicalColumnType;
 
-    public DeltaLakeColumnMetadata(ColumnMetadata columnMetadata, OptionalInt fieldId, String physicalName, Type physicalColumnType)
+    public DeltaLakeColumnMetadata(ColumnMetadata columnMetadata, String name, OptionalInt fieldId, String physicalName, Type physicalColumnType)
     {
         this.columnMetadata = requireNonNull(columnMetadata, "columnMetadata is null");
+        this.name = requireNonNull(name, "name is null");
         this.fieldId = requireNonNull(fieldId, "fieldId is null");
-        this.physicalName = physicalName.toLowerCase(ENGLISH);
+        this.physicalName = requireNonNull(physicalName, "physicalName is null");
         this.physicalColumnType = requireNonNull(physicalColumnType, "physicalColumnType is null");
     }
 
@@ -50,7 +51,7 @@ public OptionalInt getFieldId()
 
     public String getName()
     {
-        return columnMetadata.getName();
+        return name;
     }
 
     public Type getType()
@@ -73,6 +74,7 @@ public String toString()
     {
         return toStringHelper(this)
                 .add("columnMetadata", columnMetadata)
+                .add("name", name)
                 .add("fieldId", fieldId)
                 .add("physicalName", physicalName)
                 .add("physicalColumnType", physicalColumnType)
@@ -90,6 +92,7 @@ public boolean equals(Object o)
         }
         DeltaLakeColumnMetadata that = (DeltaLakeColumnMetadata) o;
         return Objects.equals(columnMetadata, that.columnMetadata) &&
+                Objects.equals(name, that.name) &&
                 Objects.equals(fieldId, that.fieldId) &&
                 Objects.equals(physicalName, that.physicalName) &&
                 Objects.equals(physicalColumnType, that.physicalColumnType);
@@ -98,6 +101,6 @@ public boolean equals(Object o)
     @Override
     public int hashCode()
     {
-        return Objects.hash(columnMetadata, fieldId, physicalName, physicalColumnType);
+        return Objects.hash(columnMetadata, name, fieldId, physicalName, physicalColumnType);
     }
 }