docs(checkpoint,checkpoint-postgres): clarify Store value and namespace semantics#8001
docs(checkpoint,checkpoint-postgres): clarify Store value and namespace semantics#8001Mason Daugherty (mdrxy) wants to merge 1 commit into
Store value and namespace semantics#8001Conversation
![open-swe[bot]](https://avatars.githubusercontent.com/in/1395372?s=60&v=4){kind=small}
| prefix: Filter namespaces that start with this path. Matched exactly and | ||
| case-sensitively. | ||
| suffix: Filter namespaces that end with this path. Matched exactly and | ||
| case-sensitively. |
There was a problem hiding this comment.
🔵 Docs hide namespace wildcard matching
This newly added text says prefix/suffix are matched “exactly”, but NamespacePath explicitly allows * and both the in-memory matcher and Postgres query builder treat * as a wildcard for one namespace segment. Users reading this reference page would be told that wildcard namespace filters are not supported/exact-only even though list_namespaces(prefix=("users", "*")) is valid behavior. The same wording is repeated in alist_namespaces and the Postgres store notes, so those copies should be updated consistently to say matching is case-sensitive while preserving the documented * wildcard semantics.
(Refers to lines 963-966)
Was this helpful? React with 👍 or 👎 to provide feedback.
1 participant
Resolves a customer escalation (DOC-1115) where the
put/aput/alist_namespacessemantics forAsyncPostgresStoreweren't documented anywhere reachable from its reference page. Two confusions drove it: the empty value/Noneand case-sensitivity behavior was never written down, andAsyncBatchedBaseStoreshadowedBaseStore's documented methods with empty docstrings.