p4core

What the P4 puck tools — p4gui, puckutility and pucktuner — share: the CAN session, flashing, configuration CSVs, calibration, and the scaffolding that lets every operation run from the command line or from a GUI.

Each app is split model / controller / view:

• model — Session (and the app's subclass of it): the bus, the selected puck, SYNC. Nothing but the Session starts SYNC.
• controllers — plain functions (p4core.ops, and each app's own controllers module) that take a Session, node IDs, options and a Reporter, and return results. No printing, no dialogs.
• shared algorithms — tuning (pucktuner's current / velocity / position gain design and PWM selection), ezero, calibrate.
• views — cli (argparse subcommands, ConsoleReporter) and the wx GUI (WxReporter), both calling the same controllers.

Shared commands

p4core.cli.add_common_commands gives every app:

scan                         node IDs on the bus
info [--all] [--flashloader] firmware, model, live readings
flash FIRMWARE [--all]       .bin / .ebin over CAN (GOLDEN images refused)
config load CSV [--all]      download, Save.All, reboot
config export TEMPLATE OUT   a template filled with the puck's values
config check CSV             validate without a puck
config templates             the shipped templates
set-id NEW                   NetCfg, reboot, PDO COB-ID remap
reset [--all]                NMT reset and wait
calibrate [STEP...]          encoder-test, bias, gain, ezero (motor turns)

with --can PORT (default $P4_CAN or can0), --id N (default: the only puck on the bus), --fd/--classic, -y, -q, -v and --json, before or after the command. Exit codes: 0 ok, 1 failed, 2 usage, 3 no bus/puck, 130 interrupted.

Layout

The repository root is the package (as in p4sysid): add it to an app as a submodule named p4core, or pip install it (package data included).

Tests

python3 -m unittest discover -s tests

tests/fake.py is a fake node over the real puck4.eds; the calibration tests drive a small simulated puck.