Background
Two polyfill functions in mlflow_cratedb/patch/mlflow/model.py carry TODO notes to submit the patches upstream to crate-python:
polyfill_uniqueness_constraints (line 10):
# TODO: Submit patch to `crate-python`, to be enabled by a
# dialect parameter `crate_polyfill_unique` or such.
polyfill_refresh_after_dml (line 49):
# TODO: Submit patch to `crate-python`, to be enabled by a
# dialect parameter `crate_dml_refresh` or such.
Both polyfills compensate for CrateDB dialect limitations (no native UNIQUE constraints, eventual consistency for reads after writes) that ideally belong in the sqlalchemy-cratedb SQLAlchemy dialect as opt-in dialect parameters.
Tasks
- Open issues / PRs in the crate/sqlalchemy-cratedb repository for:
- A dialect parameter
crate_polyfill_unique (or similar) to enable automatic uniqueness polyfilling.
- A dialect parameter
crate_dml_refresh (or similar) to enable automatic REFRESH TABLE after DML operations.
- Once accepted and released in
sqlalchemy-cratedb, replace the local polyfill implementations in this repository with dialect parameter configuration.
References
/cc @amotl
Background
Two polyfill functions in
mlflow_cratedb/patch/mlflow/model.pycarry TODO notes to submit the patches upstream tocrate-python:polyfill_uniqueness_constraints(line 10):polyfill_refresh_after_dml(line 49):Both polyfills compensate for CrateDB dialect limitations (no native UNIQUE constraints, eventual consistency for reads after writes) that ideally belong in the sqlalchemy-cratedb SQLAlchemy dialect as opt-in dialect parameters.
Tasks
crate_polyfill_unique(or similar) to enable automatic uniqueness polyfilling.crate_dml_refresh(or similar) to enable automaticREFRESH TABLEafter DML operations.sqlalchemy-cratedb, replace the local polyfill implementations in this repository with dialect parameter configuration.References
mlflow_cratedb/patch/mlflow/model.py, lines 10 and 49/cc @amotl