Datacraft

------
* 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

------
* Added typing support to ``csv_select`` field definitions: i.e.
json
{
"multiple_csv_fields": {
"type": "csv_select",
"data": {
"one:float": 1,
"tre": 3,
"svn": {"col": 7, "cast": "int"}
},
"config": {"datafile": "{{ input_csv }}"}
}
}

* Fixes for GeoJSON example documentation

------

* Added new `templated` type
(see [templated type](https://datacraft.readthedocs.io/en/latest/coretypes.html#templated)) e.g:
json
{"filled_in": {"type": "templated", "data": "{{one}} and {{two}} and a ...", "refs": ["one", "two"]}}

* Expanded API for suppliers module. Most core types have an analog function in the suppliers module to create a value
supplier. e.g:
python
from datacraft import suppliers
supplier_map = { 'char': suppliers.values(['a', 'b', 'c']), 'num': suppliers.values([1, 2, 3]) }
template_str = 'letter {{ char }}, number {{ num }}'
supplier = suppliers.templated(supplier_map, template_str)
supplier.next(0) -> letter a, number 1

* Records now returned as list by default, so `--records-per-file` or `-r` dictates the size of this list e.g:
shell
outputs 10 total records with two per file as a list
datacraft --inline '{foo:date.iso: {}}' --format json-pretty -x --log-level off -i 10 -r 2 -o . -p demo -e .json
ls
demo-0.json demo-1.json demo-2.json demo-3.json demo-4.json
cat demo-0.json
[
{
"foo": "2022-02-01T17:17:58"
},
{
"foo": "2022-02-14T19:56:45"
}
]

* Migrated from travis-ci to circle-ci for builds
* Lots of bug fixes and code cleanup