ehddnr301 commented on issue #74551: URL: https://github.com/apache/airflow/issues/74551#issuecomment-6142148108
The same issue reproduces in several other `*_prefix_pattern` filters. Environment: Breeze, PostgreSQL 16.15, database default collation `en_US.utf8`. Stored value: `카페매출집계`; search term: `카페`. Result counts: | Endpoint | `*_pattern` | `*_prefix_pattern` | |---|---|---| | `GET /api/v2/dagTags` (`tag_name_…`) | 1 | **0** | | `GET /api/v2/dags/~/dagRuns` (`triggering_user_name_…`) | 1 | **0** | | `GET /api/v2/eventLogs` (`owner_display_name_…`, `owner_…`) | 1 | **0** | | `GET /api/v2/assets` (`name_…`) | 1 | **0** | Additional column-level checks, without going through the corresponding APIs, reproduced this for asset group, asset alias name, task display name, rendered map index, `dag_id`, `task_id`, operator name, partition key, connection ID, variable key, pool name, XCom key, and import error filename. In the cases tested on PostgreSQL, matching rows were missed when the search prefix ended in a non-ASCII character (e.g. `카페`, `김`, `café`), whereas `München`, which contains `ü` but ends in ASCII `n`, matched correctly. SQLite was unaffected in these tests. --- Drafted-by: Claude Code (Opus 5.5); reviewed by ehddnr301 before posting -- 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]
