Reusable Django app for modelling provider/resource availability with weekly hours, breaks, one-off exceptions, and optional tenant-wide holidays. It answers “when can this provider be booked?” and stays out of booking/slot generation concerns.

Features

• Weekly recurring working windows per provider
• Recurring breaks inside working windows
• Exceptions for closed days and special hours
• Optional tenant holidays (enabled when a tenant model is configured)
• Timezone-safe selectors that return aware datetimes
• Admin UX tuned for day-of-week workflows

Installation

1. Add the package and Django to your project (Django 4.2+):

pip install django-availability

2. Configure settings:

• AVAILABILITY_PROVIDER_MODEL (required): the provider model as app_label.ModelName (default: "providers.Provider").
• AVAILABILITY_TENANT_MODEL (optional): tenant model as app_label.ModelName. When set, tenant FKs are added to models and TenantHoliday is enabled.

3. Add to INSTALLED_APPS:

INSTALLED_APPS = [
    # ...
    "availability",
]

4. Run migrations:

python manage.py migrate

Models

• WeeklyAvailability: recurring weekly working windows (weekday, start_time, end_time, is_active, sort_order).
• AvailabilityBreak: recurring breaks per weekday that subtract from working windows.
• AvailabilityException: one-off CLOSED or SPECIAL_HOURS per date; unique per provider+date.
• TenantHoliday (tenant mode only): tenant-wide closed dates.

Validation rules:

• end_time must be after start_time.
• Active weekly windows/breaks cannot overlap for a provider+weekday.
• SPECIAL_HOURS require start/end; CLOSED must not provide times.
• Tenant consistency enforced when AVAILABILITY_TENANT_MODEL is set.

Public API (selectors)

Import from availability.selectors:

• get_working_windows_for_date(provider, date, tz=None) -> list[(start_dt, end_dt)]
• get_working_windows(provider, start_date, end_date, tz=None) -> dict[date, list[(start_dt, end_dt)]]
• is_provider_available(provider, dt, tz=None) -> bool

Example:

from datetime import date, datetime
from django.utils import timezone
from availability.selectors import get_working_windows_for_date, is_provider_available

provider = ...
target_date = date(2024, 1, 1)
tz = timezone.get_current_timezone()

windows = get_working_windows_for_date(provider, target_date, tz=tz)
if is_provider_available(provider, datetime(2024, 1, 1, 10, 30), tz=tz):
    print("open")

The selectors:

• honor CLOSED/SPECIAL_HOURS exceptions
• subtract breaks from weekly or special-hour windows
• close days that fall on tenant holidays (if enabled)
• return timezone-aware datetimes; tz defaults to Django’s current timezone

Admin

Admin list filters cover weekday/kind/tenant and call full_clean() on save for safety. Ordering is optimized for weekday flows.

Testing locally

This repo ships a minimal test project. Run:

python test availability

The helper script pins DJANGO_SETTINGS_MODULE=tests.settings so the bundled test provider/tenant models are used. You can pass other labels, e.g. python test tests.

Releasing to PyPI

1. Bump version: update pyproject.toml (and availability/__init__.py if mirroring).
2. Clean dist: remove old artifacts: rm -rf dist/ build/ *.egg-info.
3. Build:

python -m pip install --upgrade build twine
python -m build

4. Check: python -m twine check dist/*.
5. Publish (replace token): python -m twine upload dist/* -u __token__ -p pypi-XXXXXXXXXXXXXXXXXXXX.
   • For TestPyPI: python -m twine upload --repository testpypi dist/*.
6. Verify install (optional): pip install --no-cache-dir django-availability.

Scope notes

• No booking/slot generation/payment/notification logic.
• Datetimes are expected/stored in UTC; local rules are weekday + local times.
• Overnight windows are not supported in v0.

• Download URL: django_availability-0.1.0.tar.gz
• Upload date: Dec 13, 2025
• Size: 12.6 kB
• Tags: Source
• Uploaded using Trusted Publishing? No
• Uploaded via: twine/6.2.0 CPython/3.13.3

See more details on using hashes here.

See more details on using hashes here.