Issue #12

Issue #13

Previously only lists were supported. Sets support was added by
converting to a naturally sorted set and then converting to a list for
ease of use.

Issue was likely caused by the change to load the DAO at runtime instead
of static_init. This meant that the wrapping transactions code was never
applied, and would not allow proper updates.

Flyaway package name changed, so we want to be sure that we don't apply
this to packages that don't have the upgraded path.

Should be non breaking, but this is a notable change and should be
released for use

Switched to provider to enable runtime switching capability of
Persistence DAO without having to recompile the common package. Added
config interface, though there seems to be an issue in loading it in the
implementing app.

param

Cache recorders were using pagination params as part of the unique key
generation but this is incorrect. Using those keys led to bad matching
as pagination params would not affect any of the current cache recorder
results. This also meant some cases where defaults were set would
generate keys that don't match outside of the cache.

Also added a ton of redundancy to referencing the response as it isn't
always present.

This can be used to provide your own datasource to the generator when
not using the default datasource/dao.

Current standard CSRF functionality is default, with the distributed
CSRF tokens able to be enabled through distributed mode flag.
Additionally, adding persistence DTO to your default datasource and
adding the subsequent table to your database is required. Some extra
code is required to enable this functionality on a non-default
datasource.

Ported some code for centralized date time functionalities to
standardize access to zoned datetimes.

Current code should still support limit, but takes pagesize as the
primary source of information now.

Updated SQL parameter ordinal injection to account for growing offset
(multiple ordinals in the same clause would render incorrectly). Added
join type override to allow for left or inner joins.

Imported and improved API middleware service into core lib. Will allow
either retrieval of full set of results or single page with passthrough
of pagination attributes for link header generation. Added Flat
(non-aware) request wrapper for when calls are made outside of request
context and context cannot be properly initialized.

Updated pagination default max to 100 to better accommodate batch
fetching.

CacheRecorder allows for recording of side effects and reapplication of
said effects to the response for cached calls. Tidied up pagination to
not slice assets that aren't larger than the max size (fixes
pre-paginated result sets)