stevenzwu commented on code in PR #9606:
URL: https://github.com/apache/iceberg/pull/9606#discussion_r1509461051
##########
flink/v1.18/flink/src/main/java/org/apache/iceberg/flink/FlinkSchemaUtil.java:
##########
@@ -64,26 +68,75 @@ public static Schema convert(TableSchema schema) {
RowType root = (RowType) schemaType;
Type converted = root.accept(new FlinkTypeToType(root));
- Schema iSchema = new Schema(converted.asStructType().fields());
- return freshIdentifierFieldIds(iSchema, schema);
+ Schema icebergSchema = new Schema(converted.asStructType().fields());
+ return freshIdentifierFieldIds(icebergSchema, schema);
}
- private static Schema freshIdentifierFieldIds(Schema iSchema, TableSchema
schema) {
+ /** Convert the flink table schema to apache iceberg schema with column
comment. */
+ public static Schema convert(ResolvedSchema flinkSchema) {
+ List<Column> tableColumns = flinkSchema.getColumns();
+ // copy from org.apache.flink.table.api.Schema#toRowDataType
+ DataTypes.Field[] fields =
+ tableColumns.stream()
+ .map(
+ column -> {
+ if (column.getComment().isPresent()) {
+ return DataTypes.FIELD(
+ column.getName(), column.getDataType(),
column.getComment().get());
+ } else {
+ return DataTypes.FIELD(column.getName(),
column.getDataType());
+ }
+ })
+ .toArray(DataTypes.Field[]::new);
+
+ LogicalType schemaType = DataTypes.ROW(fields).notNull().getLogicalType();
+ Preconditions.checkArgument(
+ schemaType instanceof RowType, "Schema logical type should be
RowType.");
+
+ RowType root = (RowType) schemaType;
+ Type converted = root.accept(new FlinkTypeToType(root));
+ Schema icebergSchema = new Schema(converted.asStructType().fields());
+ return freshIdentifierFieldIds(icebergSchema, flinkSchema);
+ }
+
+ /** @deprecated Use {@link #freshIdentifierFieldIds(Schema, ResolvedSchema)}
instead. */
+ @Deprecated
+ private static Schema freshIdentifierFieldIds(Schema icebergSchema,
TableSchema flinkSchema) {
+ // Locate the identifier field id list.
+ Set<Integer> identifierFieldIds = Sets.newHashSet();
+ if (flinkSchema.getPrimaryKey().isPresent()) {
+ for (String column : flinkSchema.getPrimaryKey().get().getColumns()) {
+ Types.NestedField field = icebergSchema.findField(column);
+ Preconditions.checkNotNull(
+ field,
+ "Cannot find field ID for the primary key column %s in flinkSchema
%s",
+ column,
+ icebergSchema);
+ identifierFieldIds.add(field.fieldId());
+ }
+ }
+
+ return new Schema(
+ icebergSchema.schemaId(), icebergSchema.asStruct().fields(),
identifierFieldIds);
+ }
+
+ private static Schema freshIdentifierFieldIds(Schema icebergSchema,
ResolvedSchema flinkSchema) {
// Locate the identifier field id list.
Set<Integer> identifierFieldIds = Sets.newHashSet();
- if (schema.getPrimaryKey().isPresent()) {
- for (String column : schema.getPrimaryKey().get().getColumns()) {
- Types.NestedField field = iSchema.findField(column);
+ if (flinkSchema.getPrimaryKey().isPresent()) {
+ for (String column : flinkSchema.getPrimaryKey().get().getColumns()) {
+ Types.NestedField field = icebergSchema.findField(column);
Preconditions.checkNotNull(
field,
- "Cannot find field ID for the primary key column %s in schema %s",
+ "Cannot find field ID for the primary key column %s in flinkSchema
%s",
Review Comment:
this seems incorrect. I think the old comment is correct, `schema` implied
`Iceberg table schema`.
also style is to avoid variable name in error message, which is not user
friendly. hence it would have been `Flink schema`. But this is irrelevant with
the above line.
--
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: issues-unsubscr...@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For additional commands, e-mail: issues-h...@iceberg.apache.org
