lirui-apache commented on a change in pull request #12403:
URL: https://github.com/apache/flink/pull/12403#discussion_r435917242
##########
File path:
flink-connectors/flink-connector-hive/src/main/java/org/apache/flink/table/catalog/hive/util/HiveTypeUtil.java
##########
@@ -182,11 +182,17 @@ private static DataType
toFlinkPrimitiveType(PrimitiveTypeInfo hiveType) {
@Override
public TypeInfo visit(CharType charType) {
- if (charType.getLength() > HiveChar.MAX_CHAR_LENGTH) {
- throw new CatalogException(
- String.format("HiveCatalog
doesn't support char type with length of '%d'. " +
- "The
maximum length is %d",
+ // Flink treats string literal UDF parameters as CHAR.
Such types may have precisions not supported by
+ // Hive, e.g. CHAR(0). Promote it to STRING in such
case if we're told not to check precision.
+ if (charType.getLength() > HiveChar.MAX_CHAR_LENGTH ||
charType.getLength() < 1) {
Review comment:
I'm not sure whether that makes sense because currently we don't have a
use case to involve a `VARCHAR(0)`.
Even for CHAR, Flink also has some logic to [verify the
length](https://github.com/apache/flink/blob/master/flink-table/flink-table-common/src/main/java/org/apache/flink/table/types/logical/CharType.java#L67)
is > 0. So having this CHAR(0) as UDF parameter type seems a little
inconsistent in the first place.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
