Query Information
PPL Command/Query:
source=bounty-types | where int_field NOT IN (42, -1, 0) | fields int_field
Expected Result:
NOT IN should exclude documents where int_field is null/missing.
Actual Result:
Result includes both 2147483647 and null rows.
Dataset Information
Dataset/Schema Type
Index Mapping
{
"mappings": {
"properties": {
"int_field": { "type": "integer" }
}
}
}
Sample Data
[
{ "int_field": 42 },
{ "int_field": -1 },
{ "int_field": 0 },
{ "int_field": 2147483647 },
{ "int_field": null }
]
Bug Description
Issue Summary:
NOT IN incorrectly matches null/missing field rows.
Steps to Reproduce:
- Create index with
int_field.
- Insert sample rows including one
null value.
- Run the query above.
- Observe
null row in output.
Impact:
Incorrect filter semantics and wrong result sets in production queries.
Query Information
PPL Command/Query:
Expected Result:
NOT INshould exclude documents whereint_fieldisnull/missing.Actual Result:
Result includes both
2147483647andnullrows.Dataset Information
Dataset/Schema Type
Index Mapping
{ "mappings": { "properties": { "int_field": { "type": "integer" } } } }Sample Data
[ { "int_field": 42 }, { "int_field": -1 }, { "int_field": 0 }, { "int_field": 2147483647 }, { "int_field": null } ]Bug Description
Issue Summary:
NOT INincorrectly matches null/missing field rows.Steps to Reproduce:
int_field.nullvalue.nullrow in output.Impact:
Incorrect filter semantics and wrong result sets in production queries.