SDK for XRTC API - the next generation TCP streaming protocol
Project description
XRTC API
XRTC is the next generation ultra-low latency TCP streaming protocol. It is 30-50x faster than LL-HLS and RTMP, and it is 2-5x faster than WebRTC. Unlike UDP-based WebRTC, XRTC uses a pure TCP/HTTP for ease of cross-firewall deployment and security.
This is an SDK for XRTC API in Python. The SDK implements the following convenience features:
- non-async context manager with requests package, error management
- async context manager with asyncio/aiohttp for handling parallel HTTP requests, error management
- login and connection configurations loading from .env file or from the environment
- configurations, serialized and deserialized request bodies and response data models and parser with Pydantic
To start using XRTC, please obtain your free API token at XRTC web site
This project is maintained by Delta Cygni Labs Ltd with the headquarters in Tampere, Finland.
Installation
Installation from Pypi:
pip install xrtc
Update from Pypi if you have already installed the package:
pip install xrtc --upgrade
Installation from source (advanced users only):
pip install .
Installation from source if you want the package to be editable (advanced users only):
pip install . -e
Login credentials and connection URLs
Login credentials you can provide as arguments to the contexts or specify
in a dotenv file (by default xrtc.env
) placed to the work directory.
Example of xrtc.env
file content:
# XRTC credentials
ACCOUNT_ID=AC0987654321012345
API_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Usage examples
See more on GitHub, examples directory.
Simple set and get:
from xrtc import XRTC
# Get your free account and API key from https://xrtc.org
with XRTC(account_id="AC0987654321012345", api_key="xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx") as xrtc:
# Upload an item
xrtc.set_item(items=[{"portalid": "exampleportal", "payload": "examplepayload"}])
# Download items and iterate through them
for item in xrtc.get_item(portals=[{"portalid": "exampleportal"}]):
print(item.dict())
The same example with the async context manager:
from xrtc import AXRTC
async def main():
"""Async function that enables the use of async context manager."""
# Get your free account and API key from https://xrtc.org
async with AXRTC(account_id="AC0987654321012345", api_key="xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx") as xrtc:
# Upload an item
await xrtc.set_item(items=[{"portalid": "exampleportal", "payload": "examplepayload"}])
# Download items and iterate through them
async for item in xrtc.get_item(portals=[{"portalid": "exampleportal"}]):
print(item.dict())
AXRTC.run(main())
A more sophisticated example for continuous setting and getting with XRTC and async context manager.
Measures end-to-end latency in ms. Note different get item modes (watch, probe) as well as cutoff
parameter to discard the items from previous runs. Two set of credentials are used for setting
and getting: the accountid
is the same, but the apikey
are different (request them twice from
XRTC web site), and the credentials are loaded from the respective
dotenv files.
import asyncio
from time import time
import json
from xrtc import AXRTC
class LatencyTest:
def __init__(self):
self.test_running = True
async def setting(self):
"""Set time co-routine."""
async with AXRTC(env_file_credentials="xrtc_set.env") as xrtc:
# Keep uploading items
for counter in range(0, 100):
payload = json.dumps({"time": str(time())})
await xrtc.set_item(items=[{"portalid": "latency", "payload": payload}])
await asyncio.sleep(0.1)
# Uploading finished, sleep to let all items arrive
await asyncio.sleep(1)
self.test_running = False
async def getting(self):
"""Get time co-routine."""
mean = 0
iteration = 0
async with AXRTC(env_file_credentials="xrtc_get.env") as xrtc:
# Keep polling for items
while self.test_running:
# mode="watch" means wait until there is fresh item. Compare to mode="probe"
# cutoff=500 discards the items older than 500ms, e.g. from the previous run
# iterate through the items (a single request may bring several items)
async for item in xrtc.get_item(
portals=[{"portalid": "latency"}], mode="stream", cutoff=500
):
received_time = json.loads(item.payload)["time"]
latency = round((time() - float(received_time)) * 1000, 1)
# Recurring sample mean
mean = round(1 / (iteration + 1) * (mean * iteration + latency), 1)
iteration += 1
print(f"{iteration = }: {latency = } ms, {mean = } ms")
async def execute(self):
"""Launch parallel setting and getting tasks."""
await asyncio.gather(self.setting(), self.getting())
latency_test = LatencyTest()
asyncio.run(latency_test.execute())
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.