[FLINK-3141] Design of NULL values handling in operation

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.1.0
• Component/s: Table SQL / API
• Assignee: Unassigned
• Reporter: Chengxiang Li
• Created: 08/Dec/15 06:27
• Updated: 18/May/16 13:43
• Resolved: 18/May/16 13:43

Description

We discuss and finalize how NULL value is handled in specified cases here. this is the first proposal:

1. NULL compare. In ascending order, NULL is smaller than any other value, and NULL == NULL return false.
2. NULL exists in GroupBy Key, all NULL values are grouped as a single group.
3. NULL exists in Aggregate columns, ignore NULL in aggregation function.
4. NULL exists in both side Join key, refer to #i, NULL == NULL return false, no output for NULL Join key.
5. NULL in Scalar expression, expression within NULL(eg. 1 + NULL) return NULL.
6. NULL in Boolean expression, add an extra result: UNKNOWN, more semantic for Boolean expression in reference #1.