Skip to content

Testing with Kelvin Publisher

kelvin-publisher is a CLI tool for testing applications during development. It's not meant to be imported in your code, but used as a standalone testing utility.

Note

You can read more on how to install the Publisher in the Kelvin SDK App overview here.

The publisher has three modes for simulating data:

1. Simulator - Random Data

Generate random data to your application's inputs:

Random Data Publisher Example
1
2
kelvin-publisher simulator --help
kelvin-publisher simulator

This automatically discovers your application's inputs and publishes random data.

2. CSV - File-based Data

Publish data from a CSV file:

CSV Data Publisher Example
1
2
kelvin-publisher csv --help
kelvin-publisher csv --csv data.csv

Replays test data from CSV files.

3. Generator - Custom Data

Use a custom Python class to generate data:

Generator Publisher Example
1
2
kelvin-publisher generator --help
kelvin-publisher generator --entrypoint action_generator.py:CustomActionGenerator

Example generator script:

Example Generator for Publisher Python Example
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
import asyncio
from typing import AsyncGenerator

from kelvin.message import Number
from kelvin.krn import KRNAssetDataStream
from kelvin.publisher import DataGenerator

class OtherGenerator(DataGenerator):
    def __init__(self) -> None:
        print("Hello from OtherGenerator")

    async def run(self) -> AsyncGenerator[Number, None]:
        print("Running OtherGenerator")
        for i in range(20, 30):
            yield Number(
                resource=KRNAssetDataStream("test-asset-1", "input-number"),
                payload=i,
            )
            await asyncio.sleep(1)

This mode allows you to implement sophisticated test scenarios with custom logic.