Python Carson Living is a library written in Python that exposes the carson.live devices as Python objects.

This walks through installing the library, logging in, and opening a unit door end to end.

Carson Living Python requires Python 3.11 or newer.

pip install carson-living-electric-boogaloo

from carson_living import Carson

carson = Carson("[email protected]", "your password")
print(carson.user)
# >> Martin
print(carson.token)
# >> ey...

Carson Living issues long-lived JWT tokens. Copy the printed carson.token value now; Reuse a saved token below covers skipping the login request on future runs.

for door in carson.first_building.doors:
    if door.is_unit_door:
        print("Opening Unit Door {}".format(door.name))
        door.open()

Pass a saved JWT token during initialization to skip the login request:

carson = Carson("[email protected]", "your password", "ey....")
print(carson.token)
# >> ey...

for camera in building.cameras:
    with open("image_{}.jpeg".format(camera.entity_id), "wb") as file:
        camera.get_image(file)

for camera in building.cameras:
    with open("video_{}.flv".format(camera.entity_id), "wb") as file:
        camera.get_video(file, timedelta(seconds=10))

three_hours_ago = datetime.utcnow() - timedelta(hours=3)
for camera in building.cameras:
    with open("image_{}.jpeg".format(camera.entity_id), "wb") as file:
        camera.get_image(file, three_hours_ago)

three_days_ago = datetime.utcnow() - timedelta(days=3)
for cam in building.cameras:
    with open("video_{}.flv".format(cam.entity_id), "wb") as file:
        cam.get_video(file, timedelta(seconds=5), three_days_ago)

camera.get_image_url() and camera.get_video_url() build a URL with an embedded auth_key (A=c000....) that something outside this library can fetch directly. Call building.eagleeye_api.update_session_auth_key() first if the building's key may be stale; get_image() and get_video() don't need this since they refresh internally.

building.eagleeye_api.update_session_auth_key()
for cam in building.cameras:
    img_url = cam.get_image_url(three_days_ago)
    print(img_url)
    # >> https://cXXX.eagleeyenetworks.com/asset/prev/image.jpeg?id=c0&timestamp=20200122211442.575&asset_class=pre&A=c000~...
    response = requests.get(img_url)
    with open("image_{}_with_url.jpeg".format(cam.entity_id), "wb") as file:
        file.write(response.content)
    break  # only fetch one camera in this example

Use cam.get_video_url() the same way.

./scripts/carsoncli.py has further API usage examples.

• User (carson.user): read
• Building (carson.buildings): read
• Doors (building.doors): read, open
• Cameras (building.cameras): read, images, video

• Visitor functionality (/visitors)
• Thread / messaging functionality (/threads)
• Delivery functionality (/deliveries)
• Dashboard functionality (/dashboard)
• Service functionality (/service)
• Twilio integration (twilio/access-token/)
• A separate EagleEye API package

pip install git+https://github.com/lowlydba/python-carson-living@main

Carson Living issues JWT tokens with a long validity window, so this library treats carson.token as reusable across process restarts rather than something to re-fetch on every run. It also handles expired tokens and the resulting 401 responses internally, so a caller doesn't need to check token validity before making a request.

get_image() and get_video() refresh the Eagle Eye auth_key on demand, but a pre-generated URL from get_image_url()/get_video_url() embeds whatever key was current at call time and stops working once that key expires. The key is scoped to a building rather than a single camera, so one update_session_auth_key() call covers every camera in that building.

Docstrings follow the Google Python Style Guide.

This project uses gitflow as its branching model.

This project is a fork of pbrink231/python-carson-living, itself forked from Martin Riedel's original rado0x54/python-carson-living.

Project setup and the API object design were inspired by, and partly launched off, python-ring-doorbell, which saved a lot of headaches with tox, setuptools, and Travis.