maurosilber / datastruct

A small but useful package to load, validate and use typed data structures, including configuration files.

datastruct

A small but useful package to load, validate and use typed data structures, including configuration files.

You get:

• An easy way to define a typed hierarchical data structure.
• Hassle free definition nested structures.
• Loading from a variety of formats (json, yaml and everything supported by Serialize),
• Error checking including: missing values, unexpected value, wrong type, wrong value.
• Easy to integrate in another app error reporting.

Installation

pip install datastruct

Usage

>>> from typing import List
>>> from datastruct import DataStruct
>>> class EmailServer(DataStruct):
...
...     host: str
...     port: int
...     username: str
...     password: str
>>>
>>> class Config(DataStruct):
...
...     download_path: str
...     email_servers: List[EmailServer]
...     wait_time: float
>>>
>>> cfg = Config.from_filename('settings.yaml')

When an invalid value is found, an exception will be raised.

If you want to accumulate all errors for inspection:

>>> cfg = Config.from_filename('settings.yaml', raise_on_error=False)
>>> print(cfg.get_errors())

You can then use the DataStruct object in your code:

>>> print(cfg.email_servers[0].host)

Other features

You can easily specify default values that

>>> class EmailServer(DataStruct):
...
...     host: str
...     port: int = 25
...     username: str
...     password: str

See AUTHORS for a list of the maintainers.

To review an ordered list of notable changes for each version of a project, see CHANGES