Overview
During the review of PR #27 (adding test_ddl_with_json_columns), the question was raised whether the dialect should also support and test CrateDB's column_policy table option (e.g. STRICT, DYNAMIC, IGNORED) in DDL generation.
Background
column_policy is a CrateDB table-level option that controls how columns are handled when inserting data that does not match the defined schema. It is not currently implemented in the dialect.
Currently implemented CrateDB-specific dialect options are:
crate_number_of_shards
crate_clustered_by
crate_number_of_replicas
crate_partitioned_by
crate_index
crate_columnstore
The ObjectType implementation does not currently handle any DDL-specific extras beyond what is defined in src/sqlalchemy_cratedb/type/object.py (lines 64–79).
Tasks
References
Overview
During the review of PR #27 (adding
test_ddl_with_json_columns), the question was raised whether the dialect should also support and test CrateDB'scolumn_policytable option (e.g.STRICT,DYNAMIC,IGNORED) in DDL generation.Background
column_policyis a CrateDB table-level option that controls how columns are handled when inserting data that does not match the defined schema. It is not currently implemented in the dialect.Currently implemented CrateDB-specific dialect options are:
crate_number_of_shardscrate_clustered_bycrate_number_of_replicascrate_partitioned_bycrate_indexcrate_columnstoreThe
ObjectTypeimplementation does not currently handle any DDL-specific extras beyond what is defined insrc/sqlalchemy_cratedb/type/object.py(lines 64–79).Tasks
column_policyshould be added as a supported table option in the CrateDB dialectcolumn_policy = 'strict',column_policy = 'dynamic',column_policy = 'ignored')References
JSON(B)types using CrateDB'sOBJECT#27JSON(B)types using CrateDB'sOBJECT#27 (comment)