davidradl commented on code in PR #27231:
URL: https://github.com/apache/flink/pull/27231#discussion_r2533399265
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/functions/inference/TypeInferenceOperandChecker.java:
##########
@@ -240,10 +246,28 @@ private void insertImplicitCasts(SqlCallBinding
callBinding, List<DataType> expe
/** Adopted from {@link
org.apache.calcite.sql.validate.implicit.AbstractTypeCoercion}. */
private SqlNode castTo(SqlNode node, RelDataType type) {
- return SqlStdOperatorTable.CAST.createCall(
- SqlParserPos.ZERO,
- node,
-
SqlTypeUtil.convertTypeToSpec(type).withNullable(type.isNullable()));
+ final SqlDataTypeSpec dataType;
+ if (type instanceof RawRelDataType) {
+ dataType = createRawDataTypeSpec((RawRelDataType) type);
+ } else {
+ dataType =
SqlTypeUtil.convertTypeToSpec(type).withNullable(type.isNullable());
Review Comment:
@ferenc-csaky I was looking at this again and wondered what you thought
[SqlTypeUtil](https://github.com/apache/flink/blob/7de3392b47096fb0f4163b9217a5d344b7361791/flink-table/flink-table-planner/src/main/java/org/apache/calcite/sql/type/SqlTypeUtil.java#L1059)
is a Flink class, that is in a Calcite package. I would assume we create Flink
versions of Calcite classes so we can add in Flink specifics. Putting the Raw
processing here would mean that it would be picked up by any other callers. Or
is there a convention to keep Flink dependancies out of these Calcite package
classes? Am I missing something?
--
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]
