Application settings helper for Django apps.
Application settings helper for Django apps.
Why another app settings app? Because none of the other suited my needs!
This one is simple to use, and works with unit tests overriding settings.
pip install django-app-settings
To run all the tests: tox. See CONTRIBUTING.
# Define your settings class import appsettings class MySettings(appsettings.AppSettings): boolean_setting = appsettings.BooleanSetting(default=False) required_setting = appsettings.StringSetting(required=True) named_setting = appsettings.IntegerSetting(name='integer_setting') prefixed_setting = appsettings.ListSetting(prefix='my_app_') class Meta: setting_prefix = 'app_' # Related settings in settings.py APP_INTEGER_SETTING = -24 MY_APP_PREFIXED_SETTING =  # Instantiate your class wherever you need to appconf = MySettings() assert appconf.boolean_setting is False # True (default value) assert appconf.required_setting == 'hello' # raises AttributeError assert appconf.named_setting < 0 # True assert appconf.prefixed_setting # False (empty list) # Values are cached to avoid perf issues with override_settings(APP_REQUIRED_SETTING='hello', APP_INTEGER_SETTING=0): # ...but cache is cleaned on Django's setting_changed signal assert appconf.required_setting == 'hello' # True assert appconf.named_setting < 0 # False # You can still access settings through the class itself (values not cached) print(MySettings.boolean_setting.get_value()) # explicit call print(MySettings.boolean_setting.value) # with property # Run type checking and required presence on all settings at once MySettings.check() # raises Django's ImproperlyConfigured (missing required_setting) # MySettings.check() is best called in django.apps.AppConfig's ready method
You can easily create your own Setting classes for more complex settings.
import re import appsettings from django.core.exceptions import ValidationError class RegexSetting(appsettings.Setting): def validate(self, value): re_type = type(re.compile(r'^$')) if not isinstance(value, (re_type, str)): # Raise ValidationError raise ValidationError('Value must be a string or a compiled regex (use re.compile)') def transform(self, value): # ensure it always returns a compiled regex if isinstance(value, str): value = re.compile(value) return value
Please check the documentation to see even more advanced usage.
Software licensed under ISC license.
Deprecate loding setting values from environment (#98 by @stinovlas)
Fix optional nested dict setting (#92 by @rastytheamateur)
Fix typos in usage.rst (#102 by @oto-stefan)
Fix docs build (#100 by @stinovlas)
Ignore environment variables when using override_settings.
Don’t announce type annotations.
Read setting values from environment variables.
Fix bug causing NestedDictSetting to be always required.
Add support for python 3.8 and Django 3.0.
Drop support for python 2.7 and 3.4.
Drop deprecated type checkers.
Add type annotations.
Raise ImproperlyConfigured from Setting.check for all errors in a setting.
Move repository to https://github.com/pawamoy/django-appsettings.
Fix transform_default in NestedListSetting, by @stinovlas (see PR #61).
Fix default values for empty arguments.
Deprecate setting checkers in favor of validators, similarly to Django form fields.
Add NestedSetting for easy management of nested settings.
Going from alpha to beta status. Logic has been reworked.
An instance of a subclass of AppSettings will now dynamically get settings values from project settings, and cache them. This allows to use the instance the same way in code and tests, without performance loss. See issue #16.
Cache is invalidated when Django sends a setting_changed signal (i.e. when using TestCase or override_settings). See issue #16.
Setting main class now accepts callable as default value, and two new parameters to keep control on its behavior: call_default, which tells if the default value should be called (if callable) or not, and transform_default, which tells if the default value should be transformed as well by the transform method. See issue #17.
Settings type checkers now have custom parameters like max_length, empty or key_type, that can be passed directly through the settings classes as keyword arguments. Check the documentation for more information.
Settings classes have been rewritten more explicitly, using class inheritance instead of hard-to-debug generators. Composed types like float lists or boolean sets have been removed in favor of more flexible list, set and tuple types which now accept an optional item_type parameter.
ImportedObjectSetting has been renamed ObjectSetting, and now supports object paths down to arbitrary level of nesting. Before, it only supported object paths down to classes or functions, now you can for example give it the path to a constant in a class within a class, itself contained in a module within a package. It will work as long a the deepest module is importable through importlib.import_module and each object down to the last is obtainable through getattr method.
Many thanks to ziima for having shared good ideas and thoughts!
Add six dependency (now required).
Rename Int settings to Integer, and Bool ones to Boolean.
Remove metaclass generated getters and checkers.
Settings are not checked when they default to the provided default value.
Settings classes received better default values corresponding to their types.
Add full_name property to Setting class.
Add required parameter to Setting class (default False).
Import settings classes in main module to simplify imports.
Add PositiveInt and PositiveFloat settings.
Add support for Django 1.11.
Implement basic settings classes.
Implement basic Setting class.
Change distribution name to app-settings.
Alpha release on PyPI.
Release history Release notifications | RSS feed
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Hashes for django-app-settings-0.7.2.tar.gz
Hashes for django_app_settings-0.7.2-py2.py3-none-any.whl