Yancy

Changes for version 1.024 - 2019-04-26

Added
- When reading a schema from the database, Yancy will now prefer "natural" keys to artificial ones. For example, if you have a "users" table with an auto-increment "id" column and a unique "username" column, Yancy will pick the "username" column as the ID field. This creates a nicer URL structure in Yancy controllers.
- Default values are now properly set from the database when reading the schema.
- For DBIx::Class backends, Yancy will now call the `yancy` method on the ResultSource object to get default configuration for the collection. Using this method you can quickly make Yancy ignore your table by returning `{ 'x-ignore' => 1 }`, you can improve Yancy's detection of field settings, or you can provide a default set of list columns.
- Virtual collections can now be created with `x-view`. A virtual collection can provide a read-only subset of the underlying collection's properties. In the future, virtual collections can become editable and will have column filters.
- Added a `yancy backend copy` command. Use this command to copy data from your app's backend to another backend. This can initialize a test database, migrate from one backend to another, or promote data from a staging server to production.
Fixed
- Yancy will now normalize the JSON schema passed-in so that consumers get a full schema. All the default values Yancy uses will be placed in the schema data structure returned by the `schema` helper.
- When reading a MySQL database, Yancy now correctly finds a unique column. Previously, it could pick a non-unique column that had an index.
Other
- Added tests for running search queries on various non-string types (integer, boolean)
- Added more tests for `order_by` in the backend list method
- Migrated the Yancy doc site example to use the new Yancy::Backend::Static module. This backend module is in its own distribution because it depends on some extra prerequisites.