Datacraft

------

* Addition of zfillN casters
* Added --cast-list and --format-list command line args to dump registered casters and formatters
* Added `csv-with-header` (`csvh` for short) formatters, i.e. `--format csvh`
* Various bug fixes

------
* Bug fixes and removal of `datacraft.spec_builder` function and tooling from API.

------
* Added initial built in type help system with command line --type-list and --type-help command line arguments

--------
* Switched to lru_cache for pre python 3.9 compatibility

------

* Fix for custom types not loading when datacraft used programmatically.
* Addition of `datacraft.entries` function for quickly generating records from a given spec

python
import datacraft

spec = {
"super_power": {
"type": "values",
"config": { "sample": True },
"data": {
"fast reader": 0.5,
"wordle expert": 0.4,
"super strength": 0.05,
"super speed": 0.01,
"invisibility": 0.01,
"indestructible": 0.01,
"laser eyes": 0.01,
"teleportation": 0.00000000001
}
}
}
records = datacraft.entries(spec, 5)

------
* Introduced `datacraft.custom_type_loader` entrypoint for discovering and loading `datacraft.registry.*` functions.

Add to setup.cfg:

yaml
[options.entry_points]
datacraft.custom_type_loader =
mycustomstuff = mypackage:load_custom


See [custom-types-entry-point](https://datacraft.readthedocs.io/en/latest/usage.html#custom-types-entry-point) for
details