danielcweeks commented on code in PR #13699:
URL: https://github.com/apache/iceberg/pull/13699#discussion_r2249002781
##########
arrow/src/main/java/org/apache/iceberg/arrow/ArrowSchemaUtil.java:
##########
@@ -51,94 +54,133 @@ private ArrowSchemaUtil() {}
public static Schema convert(final org.apache.iceberg.Schema schema) {
ImmutableList.Builder<Field> fields = ImmutableList.builder();
- for (NestedField f : schema.columns()) {
- fields.add(convert(f));
+ for (NestedField field : schema.columns()) {
+ fields.add(TypeUtil.visit(field.type(), new
IcebergToArrowTypeConverter(field)));
}
-
return new Schema(fields.build());
}
public static Field convert(final NestedField field) {
- final ArrowType arrowType;
-
- final List<Field> children = Lists.newArrayList();
- Map<String, String> metadata = null;
-
- switch (field.type().typeId()) {
- case BINARY:
- arrowType = ArrowType.Binary.INSTANCE;
- break;
- case FIXED:
- final Types.FixedType fixedType = (Types.FixedType) field.type();
- arrowType = new ArrowType.FixedSizeBinary(fixedType.length());
- break;
- case BOOLEAN:
- arrowType = ArrowType.Bool.INSTANCE;
- break;
- case INTEGER:
- arrowType = new ArrowType.Int(Integer.SIZE, true /* signed */);
- break;
- case LONG:
- arrowType = new ArrowType.Int(Long.SIZE, true /* signed */);
- break;
- case FLOAT:
- arrowType = new ArrowType.FloatingPoint(FloatingPointPrecision.SINGLE);
- break;
- case DOUBLE:
- arrowType = new ArrowType.FloatingPoint(FloatingPointPrecision.DOUBLE);
- break;
- case DECIMAL:
- final Types.DecimalType decimalType = (Types.DecimalType) field.type();
- arrowType = new ArrowType.Decimal(decimalType.precision(),
decimalType.scale(), 128);
- break;
- case STRING:
- arrowType = ArrowType.Utf8.INSTANCE;
- break;
- case TIME:
- arrowType = new ArrowType.Time(TimeUnit.MICROSECOND, Long.SIZE);
- break;
- case UUID:
- arrowType = new ArrowType.FixedSizeBinary(16);
- break;
- case TIMESTAMP:
- arrowType =
- new ArrowType.Timestamp(
- TimeUnit.MICROSECOND,
- ((Types.TimestampType) field.type()).shouldAdjustToUTC() ?
"UTC" : null);
- break;
- case DATE:
- arrowType = new ArrowType.Date(DateUnit.DAY);
- break;
- case STRUCT:
- final StructType struct = field.type().asStructType();
- arrowType = ArrowType.Struct.INSTANCE;
-
- for (NestedField nested : struct.fields()) {
- children.add(convert(nested));
- }
- break;
- case LIST:
- final ListType listType = field.type().asListType();
- arrowType = ArrowType.List.INSTANCE;
-
- for (NestedField nested : listType.fields()) {
- children.add(convert(nested));
- }
- break;
- case MAP:
- metadata = ImmutableMap.of(ORIGINAL_TYPE, MAP_TYPE);
- final MapType mapType = field.type().asMapType();
- arrowType = new ArrowType.Map(false);
- List<Field> entryFields = Lists.transform(mapType.fields(),
ArrowSchemaUtil::convert);
- Field entry =
- new Field("", new FieldType(field.isOptional(), arrowType, null),
entryFields);
- children.add(entry);
- break;
- default:
- throw new UnsupportedOperationException("Unsupported field type: " +
field);
+ return TypeUtil.visit(field.type(), new
IcebergToArrowTypeConverter(field));
+ }
+
+ private static class IcebergToArrowTypeConverter extends
TypeUtil.SchemaVisitor<Field> {
+ private final NestedField currentField;
+
+ IcebergToArrowTypeConverter(NestedField field) {
+ this.currentField = field;
+ }
+
+ @Override
+ public Field schema(org.apache.iceberg.Schema schema, Field structResult) {
+ return structResult;
+ }
+
+ @Override
+ public Field struct(StructType struct, List<Field> fieldResults) {
+ ArrowType arrowType = ArrowType.Struct.INSTANCE;
Review Comment:
I don't think we need to assign the type to a variable like this. It can be
inlined.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
