django-annoying

django-annoying is a Django application that smooths over a handful of common annoyances in the framework by providing small, focused decorators, fields, functions, and middleware. It bundles conveniences such as a render_to decorator for views, an ajax_request decorator that returns JSON, and a get_object_or_None helper that mirrors get_object_or_404 but returns None instead of raising.

Rather than a single big abstraction, it is a grab-bag of independently useful utilities — a signals-as-decorators helper, an autostrip form decorator, an AutoOneToOne field, a JSONField, a get_config settings helper, and a StaticServer middleware — each designed to remove a bit of repetitive Django code.

What You Get

• A render_to decorator that maps a view’s returned dict to a template
• An ajax_request decorator that serializes a returned dict to a JsonResponse
• A get_object_or_None function that returns None instead of raising when nothing matches
• Model helpers like AutoOneToOne and a JSONField
• Form and settings utilities including autostrip, get_config, and a StaticServer middleware

Common Use Cases

• Cutting repetitive render and JSON-response boilerplate in Django views
• Fetching an object that may not exist without wrapping the query in try/except
• Storing Python objects as JSON on a model field
• Reading optional settings with a fallback default via get_config

Under The Hood

Architecture - The annoying/ package is a flat set of single-responsibility modules: decorators.py holds the view and signal decorators (render_to, ajax_request, signals, autostrip), functions.py provides helpers like get_object_or_None and get_config, fields.py defines the AutoOneToOne and JSONField model fields, middlewares.py implements the StaticServer middleware, and exceptions.py/utils.py supply supporting pieces. Each utility is independent, so importing one does not pull in the others.

Tech Stack - Written in Python against the Django framework, packaged for PyPI with a bundled py.typed marker for type-checking support. It has no dependencies beyond Django itself and tracks compatibility across Django versions.

Code Quality - The codebase is small, readable, and organized by concern, and the project is mature with 44 contributors over a long history. Recent development activity is low — it is maintained under Code Shelter — which fits a stable utility library whose surface rarely needs to change.

API Design - The utilities are intentionally familiar and low-ceremony: get_object_or_None mirrors Django’s own get_object_or_404, and the decorators read naturally atop existing views. Each helper does one obvious thing, so the learning curve is minimal and adoption is incremental — you import only the shortcut you need.