It fixes main initial implementation flaw with not supported types. Now if a database doesn't support type it uses a string fallback. As a result now we support all specs-v1 types (added to tests) and more databases (sqlite and mysql tested).
Also it includes some minor improvements and simplifies tests (schema/data in the same file instead of reading it from the disc for convenience).
It fixes main initial implementation flaw with not supported types. Now if a database doesn't support type it uses a
string
fallback. As a result now we support all specs-v1 types (added to tests) and more databases (sqlite
andmysql
tested).Also it includes some minor improvements and simplifies tests (schema/data in the same file instead of reading it from the disc for convenience).
Tests provide good overview how it works with types now - https://github.com/frictionlessdata/tableschema-sql-py/blob/more-databases/tests/test_storage.py