apilloud commented on a change in pull request #12348:
URL: https://github.com/apache/beam/pull/12348#discussion_r465373933
##########
File path:
sdks/java/extensions/sql/zetasql/src/main/java/org/apache/beam/sdk/extensions/sql/zetasql/translation/ExpressionConverter.java
##########
@@ -850,6 +848,25 @@ private RexNode convertSimpleValueToRexNode(TypeKind kind,
Value value) {
// TODO: Doing micro to mills truncation, need to throw exception.
ret = rexBuilder().makeLiteral(convertTimeValueToTimeString(value),
timeType, false);
break;
+ case TYPE_DATETIME:
+ // Cannot simply call makeTimestampWithLocalTimeZoneLiteral() for
ZetaSQL DATETIME type
+ // because later it will be unparsed to the string representation of
timestamp (e.g. "SELECT
+ // DATETIME '2008-12-25 15:30:00'" will be unparsed to "SELECT
TIMESTAMP '2008-12-25
+ // 15:30:00:000000'"). So we create a wrapper function here such that
we can later recognize
+ // it and customize its unparsing in BeamBigQuerySqlDialect.
+ ret =
+ rexBuilder()
+ .makeCall(
+ SqlOperators.createZetaSqlFunction(
+ BeamBigQuerySqlDialect.DATETIME_LITERAL_FUNCTION,
Review comment:
Instead of wrapping with a function here, it looks like you can do the
same thing by creating a class overloading `SqlTimestampLiteral.toString()` and
wrapping it in `BeamSqlUnparseContext`? This would reduce the complexity at the
Rel layer.
https://github.com/apache/beam/blob/6fdde4f4eab72b49b10a8bb1cb3be263c5c416b5/sdks/java/extensions/sql/src/main/java/org/apache/beam/sdk/extensions/sql/meta/provider/bigquery/BeamSqlUnparseContext.java#L70
----------------------------------------------------------------
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:
us...@infra.apache.org
