Datalink in pure python.
Read the docs at readthedocs
Datalink is a protocol for near-realtime transfer of seismic data, usually in miniseed, but has flexibility to carry any payload such as JSON. The Protocol is defined at https://iris-edu.github.io/libdali/datalink-protocol.html
See ringserver from IRIS for the most common datalink server instance. The public instance at rtserve.iris.washington.edu/ allows access to near-realtime streaming seismic data over web sockets at ws://rtserve.iris.washington.edu/datalink
For parsing for miniseed2 and miniseed3 see simplemseed, also in pure python.
Support for both regular sockets and websockets. For example:
import asyncio
import simpledali
async def main():
host = "localhost"
port = 16000
uri = f"ws://{host}:{port}/datalink"
verbose = True
programname = "simpleDali"
username = "dragrace"
processid = 0
architecture = "python"
# for regular socket (DataLinkPort)
async with simpledali.SocketDataLink(host, port, verbose=verbose) as dali:
serverId = await dali.id(programname, username, processid, architecture)
print(f"Connect to {host}:{port} via regular socket")
print(f"Socket Id: {serverId.message}")
# for web socket (ListenPort)
async with simpledali.WebSocketDataLink(uri, verbose=verbose) as dali:
serverId = await dali.id(programname, username, processid, architecture)
print(f"Connect to {uri} via websocket")
print(f"WebSocket Id: {serverId.message}")
asyncio.run(main())
The dali2jsonl script will archive '/JSON' packets as JSON Lines. This is a similar function to the MSeedWrite configuration on ringserver, but in a separate process and saves JSON packets instead of miniseed. See jsonlines.org for the file format, basically one JSON value per line.
dali2jsonl --help
usage: dali2jsonl [-h] [-v] -c CONF
Archive JSON datalink packets as JSON Lines.
optional arguments:
-h, --help show this help message and exit
-v, --verbose increase output verbosity
-c CONF, --conf CONF Configuration as TOML
The TOML configuration looks like:
[datalink]
# datalink host, defaults to localhost
host='localhost'
# datalink port, defaults to 18000
port=15004
# Match regular expression pattern on stream ids, ex '.*/JSON'
match='.*/JSON'
[jsonl]
# JSONL Write pattern, usage similar to MSeedWrite in ringserver
write='/data/scsn/www/jsonl/%n/%s/%Y/%j/%n.%s.%l.%c.%Y.%j.%H.jsonl'
There are examples of sending and receiving Datalink packets in the example directory.