Structures

A Python3.13+ library that wraps cattrs for a modular approach to constructing objects with the ability to string data through the process.

Install from pypi:

> python -m pip install strcs

Documentation at https://strcs.readthedocs.io/

Example

fromtypingimportAnnotated, NewTypeimportattrsimportstrcsreg=strcs.CreateRegister()
creator=reg.make_decorator()
Number=NewType("Number", int)
Word=NewType("Word", str)
@attrs.define(frozen=True)classMaths(strcs.MetaAnnotation):
multiply: intdefcalculate(self, val: int) ->Number:
returnNumber(val*self.multiply)
classThing:
pass@attrs.defineclassConfig:
thing: Annotated[Thing, strcs.FromMeta("thing")]
words: list[Word]
some_number: Annotated[Number, Maths(multiply=2)]
contrived1: strcontrived2: strsome_other_number: int=16@creator(Number)defcreate_number(val: object, /, annotation: Maths) ->strcs.ConvertResponse[Number]:
ifnotisinstance(val, int):
returnNonereturnannotation.calculate(val)
@creator(Word)defcreate_word(val: object, /, word_prefix: str="") ->strcs.ConvertResponse[Word]:
ifnotisinstance(val, str):
returnNonereturnWord(f"{word_prefix}{val}")
@creator(Config)defcreate_config(val: object, /) ->strcs.ConvertResponse[Config]:
ifnotisinstance(val, dict):
returnNoneresult=dict(val)
if"contrived"inresult:
contrived=result.pop("contrived")
result["contrived1"], result["contrived2"] =contrived.split("_")
returnresultthing=Thing()
meta=strcs.Meta({"thing": thing, "word_prefix": "the_prefix__"})
config=reg.create(
Config,
{"words": ["one", "two"], "some_number": 20, "contrived": "stephen_bob"},
meta=meta,
)
print(config)
assertisinstance(config, Config)
assertconfig.thingisthingassertconfig.words== ["the_prefix__one", "the_prefix__two"]
assertconfig.some_number==40assertconfig.some_other_number==16assertconfig.contrived1=="stephen"assertconfig.contrived2=="bob"

Development

To have a virtualenv that has everything needed in it:

> source run.sh activate

To run tests, linting, formatting, type checking:

> ./test.sh
> ./lint
> ./format
> ./types