config-types

Type aliases and validators for configuration settings.

This package provides aliases for types typically found in configuration files. When the pydantic versions are used, they come with validators that transform the strings to their intended type (and value).

When the standard types are used, they all resolve to string and it's your job to transform them in your application code.

The rest of this document assumes usage with pydantic or more specifically with pydantic-settings.

Quick start

Install with: pip install config-types[pydantic]

Create your application settings:

from pydantic_settings import BaseSettings
from config_types.pydantic import IsTrue, DottedPathAttribute

class Config(BaseSettings):
    debug: IsTrue = False
    cache_backend: DottedPathAttribute

settings = Config()

Populate your environment, for example using a .env file:

DEBUG=yes
CACHE_BACKEND=django_redis.cache.RedisCache

Now import your settings singleton where needed.

Reference

DottedPath

A reference to a python module that is imported and returned if found.

Example:

from config_types.pydantic import DottedPath
from pydantic_settings import BaseSettings


class AppConfig(BaseSettings):
    logging_module: DottedPath

# In `.env`:
# LOGGING_MODULE=loguru

ModuleAttributeRef

A reference to a module attribute, typically a class name, singleton or callable. The module path is a DottedPath and the attribute to be imported is separated by a colon.

Example:

from config_types.pydantic import ModuleAttributeRef
from pydantic_settings import BaseSettings

class AppConfig(BaseSettings):
    asgi_application: ModuleAttributeRef

# In `.env`:
# ASGI_APPLICATiON=app.main:application

DottedPathAttribute

The same as a ModuleAttributeRef, but done with all dots. The last part is the entity reference.

Example:

from config_types.pydantic import DottedPathAttribute
from pydantic_settings import BaseSettings

class AppConfig(BaseSettings):
    CACHE_BACKEND: DottedPathAttribute

# In `.env`:
# CACHED_BACKEND=django_redis.cache.RedisCache

IsTrue

A switch that defaults to false. The following values are recognised as true (case-insensitive):

• "yes"
• "on"
• "1"
• "true"

Example:

from config_types.pydantic import IsTrue
from pydantic_settings import BaseSettings

class AppConfig(BaseSettings):
    debug: IsTrue = False

# In your development `.env`:
# DEBUG=yes

IsFalse

A switch that defaults to true. The following values are recognised as false (case-insensitive):

• "no"
• "off"
• "0"
• "false"

Example:

from config_types.pydantic import IsFalse
from pydantic_settings import BaseSettings

class AppConfig(BaseSettings):
    use_https: IsFalse = True

# In your development `.env`:
# USE_HTTPS=off

RelativeUrlRef

A relative URL reference, in the strict sense. While the URL specs may classify a URL starting with '/' as relative, this does not. It is meant to be used as the final part of a URL, where the first part is different for each environment. The classic case, is an asset reference part of static resources, which is then prepended with the STATIC_URL.

The validator does not strip a leading slash, but throws an error, to prevent users from providing references to the root of the server, which a typical application does not allow.

Example:

from config_types.pydantic import RelativeUrlRef
from pydantic_settings import BaseSettings

class AppConfig(BaseSettings):
    favicon: RelativeUrlRef = 'favicons/16x16.ico'

# In `.env`:
# FAVICON=branding/logo-32.png

TODO

The following are planned:

• containers
• support for other packages that can support a similar interface, if any

Note: numbers aren't planned as pydantic-settings (and typing) have plenty support for it.