twalthr commented on code in PR #23829:
URL: https://github.com/apache/flink/pull/23829#discussion_r1410385087
##########
flink-table/flink-table-common/src/main/java/org/apache/flink/table/expressions/ValueLiteralExpression.java:
##########
@@ -219,6 +222,83 @@ public String asSummaryString() {
return stringifyValue(value);
}
+ @Override
+ public String asSerializableString() {
+ if (value == null &&
!dataType.getLogicalType().is(LogicalTypeRoot.NULL)) {
+ return String.format(
+ "CAST(NULL AS %s)",
+ // casting does not support nullability
+
dataType.getLogicalType().copy(true).asSerializableString());
+ }
+ final LogicalType logicalType = dataType.getLogicalType();
+ switch (logicalType.getTypeRoot()) {
+ case TINYINT:
+ return String.format("CAST(%s AS TINYINT)", value);
+ case SMALLINT:
+ return String.format("CAST(%s AS SMALLINT)", value);
+ case BIGINT:
+ return String.format("CAST(%s AS BIGINT)", value);
+ case FLOAT:
+ return String.format("CAST(%s AS FLOAT)", value);
+ case DOUBLE:
+ return String.format("CAST(%s AS DOUBLE)", value);
+ case CHAR:
+ case VARCHAR:
+ case DECIMAL:
+ case INTEGER:
+ return stringifyValue(value);
+ case BOOLEAN:
+ case SYMBOL:
Review Comment:
Would it make sense to throw an exception for symbols instead? And let the
call handle those cases instead of potentially serializing invalid SQL.
--
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]
