pytest-embedded-wireskein

日本語 README

A pytest-embedded plugin that gives each test a WireSkein recorder, ws_run. The test records the commands it sent, the logic-analyzer captures, and what each step should look like on the wire. When the test body ends, the plugin checks the captures against these expectations and fails the test if a check fails.

Status: beta.

Install

pip install pytest-embedded-wireskein

Python 3.13 or newer. This installs wireskein and pytest-embedded.

Use

from wireskein.runlog import square, level, only_moving

def test_pwm(dut, ws_run, probe):              # `probe` is whatever drives your logic analyzer
    with ws_run.section(1, "pwm"):
        for duty in (64, 128, 0):
            want = [square("PA1", 1000, duty / 255), only_moving(["PA1"])] if duty else [level("PA1", 0)]
            with ws_run.section(2, f"duty={duty}", expect=want):
                ws_run.command(f"PWM {duty}")
                dut.write(f"PWM {duty}")
                ws_run.reply(dut.expect(r"PWM duty=\d+").group(0).decode())
                t = ws_run.armed()               # time.monotonic() right after arming the capture
                data, rate = probe.capture()     # bytes, one sample per byte, bit k = pin k
                ws_run.capture(t, rate, interleaved=data, names=["PA1", "PA0"])

When test_pwm returns, the plugin closes the run and verifies it. If a check fails, the test fails in its call phase (FAILED, not ERROR):

FAILED test_pwm.py::test_pwm - wireskein: 1 NG (5 ok, 1 ng, 0 unchecked (4 segments, 3 captures))
NG  pwm/duty=128  square  c0002.bin  duty 0.6999 vs 0.5020
report: /tmp/pytest-embedded/2026-09-29_12-00-00-000000/test_pwm/wireskein/report.json

The checks (square, level, starts, ends, only_moving, pulses, i2c, spi, uart) and the heading rules are described in the WireSkein README.

Where the run is written

ws_run writes to <test_case_tempdir>/wireskein/, next to pytest-embedded's dut.log: <root-logdir>/pytest-embedded/<time>/<test name>/wireskein/.

The same directory can be checked again later with wireskein verify <dir>. The test's user_properties carry wireskein_report (the path of report.json), and the result lines are added to the test report as a wireskein section (shown with -rA or on failure).

When the plugin verifies

• The run is verified after the test body, only if the test recorded a capture or an expectation.
• A check that fails makes the test fail.
• A check whose pins were not captured is unchecked and does not fail the test.
• If the test body already failed, the run is still recorded and verified, but the test's own failure is kept.

Options

Connecting a probe

This plugin knows nothing about probes or targets. A fixture that drives the logic analyzer (for example the one of a board-family plugin) connects to ws_run when both are installed:

• right after arming a capture: t = ws_run.armed() (time.monotonic(); a capture client's own stamp of the same clock works too)
• when the samples are read: ws_run.capture(t, rate, interleaved=data, names=[...], width=8, positions=None, start_us=..., time_base_slipped=True). data is the probe's sample stream (width bits per sample, channel k at bit positions[k]), names the target's pin names in channel order. Pass time_base_slipped only when the probe reports it.
• channels at different rates: ws_run.capture(t, tick_hz, channels=[wireskein.fileformat.Channel(name, bits, n, step=...)])
• anything else about the capture: attachments={"probe.json": {...}}
• the console traffic: ws_run.command(text), ws_run.reply(text)

Development

uv sync
uv run pytest

Release

The release works the same way as in pytest-embedded-arduino-cli. Update ## Unreleased in CHANGELOG.md, then run the Release workflow with the version (for example 0.0.2). PyPI publishing uses Trusted Publishing.

License

MIT