tempestwx 0.0.2

Tempest SDK for Python

Tempest

Python SDK for the Tempest API.

Installation

Requires Python 3.13+.

• From PyPI:

  pip install tempestwx

Usage

Ensure TEMPEST_ACCESS_TOKEN is set (via environment or .env).

from tempestwx import Tempest

# Recommended: Use context manager for automatic resource cleanup
with Tempest() as twx:
    stations = twx.stations()
    print(stations)

# Alternative: Manual resource management
twx = Tempest()
try:
    stations = twx.stations()
    print(stations)
finally:
    twx.close()

The client returns Pydantic models. Use .model_dump() to convert to dict when needed.

Configuration & Settings

The client uses a deterministic cascade to resolve configuration:

1. Library defaults (built‑in constants)
2. config.json (first existing among: $TEMPEST_CONFIG_PATH, ./config.json)
3. .env file (loaded automatically once, without overriding already exported environment vars)
4. Live environment variables (take precedence over .env values)
5. Explicit parameters passed to Tempest(...) (highest precedence)

Resolved values are materialized into an immutable Settings instance:

from tempestwx.settings_loader import load_settings
settings = load_settings()
print(settings.api_uri, settings.token)

You can explicitly override when constructing the client:

from tempestwx import Tempest

with Tempest(token="OVERRIDE_TOKEN") as twx:
    stations = twx.stations()

Or derive a modified settings object:

from tempestwx.settings_loader import load_settings

base = load_settings()
custom = base.with_overrides(api_uri="https://example.test/api/")

with Tempest(settings=custom) as twx:
    stations = twx.stations()

Environment Variables

Supported variables:

• TEMPEST_ACCESS_TOKEN – API auth token
• TEMPEST_API_URI – Base API URI (defaults to https://swd.weatherflow.com/swd/rest/)
• TEMPEST_CONFIG_PATH – Optional path to a JSON config file (fallbacks to ./config.json)
• Unit overrides (optional):
  • TEMPEST_DEFAULT_UNIT_TEMPERATURE
  • TEMPEST_DEFAULT_UNIT_PRESSURE
  • TEMPEST_DEFAULT_UNIT_WIND
  • TEMPEST_DEFAULT_UNIT_DISTANCE
  • TEMPEST_DEFAULT_UNIT_PRECIP
  • TEMPEST_DEFAULT_UNIT_BRIGHTNESS
  • TEMPEST_DEFAULT_UNIT_SOLAR_RADIATION
  • TEMPEST_DEFAULT_UNIT_BUCKET_STEP_MINUTES

.env Support

If a .env file exists in the working directory, it is loaded automatically (without overriding already exported variables) the first time load_settings() runs.

Example .env:

TEMPEST_ACCESS_TOKEN=your-token-here

config.json example

If present, values provide defaults that can be overridden by environment variables:

{
  "api_uri": "https://swd.weatherflow.com/swd/rest/",
  "default_unit_temperature": "c",
  "default_unit_pressure": "mb",
  "default_unit_wind": "mps",
  "default_unit_distance": "km",
  "default_units_precip": "mm",
  "default_units_brightness": "lux",
  "default_units_solar_radiation": "w/m2",
  "default_units_bucket_step_minutes": 1
}

Note: tokens are not read from config.json. Use environment variables or .env for TEMPEST_ACCESS_TOKEN.

Reloading Settings

Caching avoids repeated disk & env parsing. To pick up changes at runtime:

from tempestwx.settings_loader import reload_settings
reload_settings()  # clears cache and re-evaluates cascade

Units Overrides

You can supply partial unit overrides via UnitsOverrides (only fields you specify are changed):

from tempestwx import Tempest
from tempestwx.settings import UnitsOverrides
from tempestwx.settings_loader import load_settings

base = load_settings()
custom = base.with_overrides(units_overrides=UnitsOverrides(temp="f", wind="mph"))

with Tempest(settings=custom) as twx:
    forecast = twx.better_forecast(station_id=12345)

Token Context Override

Temporarily swap tokens within a context:

from tempestwx import Tempest

with Tempest() as twx:
    # Use default token
    stations = twx.stations()

    # Temporarily use different token
    with twx.token_as("temporary-token"):
        other_stations = twx.stations()

    # Back to default token
    more_stations = twx.stations()

Async Usage

All endpoints support async when the client is created with asynchronous=True:

import asyncio
from tempestwx import Tempest

async def main():
    # Use async context manager for proper cleanup
    async with Tempest(asynchronous=True) as twx:
        stations = await twx.stations()
        print(stations)

asyncio.run(main())

Roadmap

• OAuth Authorization Code (with PKCE) grant types
• Additional Tempest APIs, e.g. TempestOne
• Standalone documentation with MkDocs