python/iot3/mobility: ETSI time fixes, expose more GNSS details, extend support for CAM

[ymorin-orange]

Changes

• IoT3 SDK:
  • mobility:
    • fix ETSI zero-timestamp (python/mobility: ETSI EPOCH zero-timestamp can't be covnerted back to UNIX timestamp #585)
    • add ETSI time conversion tests
    • expose list of GNSS devices (python/iot3/mobility: expose GNSS devices #582)
    • expose type of GNSS fix (python/iot3/mobility: expose GNSS fix status #583)
    • add length and width to CAM
    • fix station-ID vs station-UUID (python/mobility: fix station_id() to enforce and recognise the three-part station-UUID #587)
• applications:
  • vehicle: add length and width in CAM

Close #582
Close #583
Close #585
Close #587

Test

How to test

Note: in the following, lines starting with $ are to be executed on your
machine, as a non-root user;
lines starting with (docker)$ are to be executed in the Docker container;
lines starting with (docker)🐍 $ are to be executed in the Docker container,
in the python venv.

1. Prepare a test environment:
   1. be sure to have unrestricted access to test.mosquitto.org (IPv4 and IPv6)
   2. be sure to have an MQTT broker that listens locally on port 1883,
      with no credentials and no ACL; if not, run your own:

      $ mosquitto
      

   3. in another terminal, prepare a collector implementing
      the OpenTelemetry API, on localhost. If you don't have one,
      you may use an existing one, like:

      $ docker container run \
          --rm \
          -p 16686:16686 \
          -p 4318:4318 \
          jaegertracing/all-in-one:1.58
      

      then open a browser on the Jaegger UI
      (or that of your own collector if you have one):

      http://localhost:16686/
      

   4. in another terminal, start a fake gpsd process
      (no need for an actual GNSS device):

      $ TMPDIR=/tmp gpsfake -q -P 2948 -u -n -c 0.1 tests/data/NMEA.log
      

   5. in another terminal, start a container with Python 3.11
      and the necessary packages:

      $ docker container run \
          --detach \
          --name iot3 \
          --rm \
          -ti \
          --network host \
          -e http_proxy \
          -e https_proxy \
          -e no_proxy \
          --user $(id -u):$(id -u) \
          --mount type=bind,source=$(pwd),destination=$(pwd) \
          --workdir $(pwd) \
          python:3.11.15-slim-trixie \
          /bin/bash -il
      
      $ docker container exec -u 0:0 iot3 apt update
      $ docker container exec -u 0:0 iot3 apt install -y git build-essential curl socat
      $ docker container exec -u 0:0 iot3 install -d -m 0755 /run/zoneinfo
      $ docker container exec -u 0:0 iot3 curl -o /run/zoneinfo/leap-seconds.list 'https://data.iana.org/time-zones/data/leap-seconds.list' 
      $ docker container exec -d iot3 socat UNIX-LISTEN:/tmp/mqtt.socket,fork TCP4:localhost:1883
      
      $ docker container attach iot3
      

      Note: the socat command exposes the MQTT broker listening
      on localhost:1883 (see 1.2., above), inside the
      container listening on the UNIX socket /tmp/mqtt.socket.
   6. prepare a Python 3.11 environment, with tests dependencies
      and packages' dependencies:

      (docker)$ python3.11 -m venv /tmp/venv
      (docker)$ . /tmp/venv/bin/activate
      (docker)🐍 $ pip --disable-pip-version-check --no-cache-dir install \
                      -r python/requirements-tests.txt \
                      $(sed -r -e '/.*"(.+(==|>=|<=).+)".*/!d; s//\1/; s/ //g;' \
                        python/*/pyproject.toml
                       )
      

      Note: we install the packages dependencies manually,
      and do not rely on pip to do so, because our Python packages depend
      one on the others by git hash, as they are not published on PyPi yet,
      so installing one of our packages may overwrite another.
2. Build the Python packages:

   (docker)🐍 $ for pkg in python/*/; do
                   python -m build -w "${pkg}" || break
                done
   

3. Install the Python packages:

   (docker)🐍 $ pip --disable-pip-version-check --no-cache-dir \
                       install --no-deps python/*/dist/*.whl
   

4. Run the IoT3 Core SDK tests:

   (docker)🐍 ./python/iot3/tests/test-iot3-core
   (docker)🐍 ./python/iot3/tests/test-iot3-core-all
   (docker)🐍 ./python/iot3/tests/test-iot3-core-mqtt
   (docker)🐍 ./python/iot3/tests/test-iot3-core-otel
   (docker)🐍 ./python/iot3/tests/test-iot3-mobility
   (docker)🐍 ./python/iot3/tests/test-iot3-mobility-gnss
   (docker)🐍 ./python/iot3/tests/test-iot3-mobility-message
   

   Note: if test.mosquitto.org is unreachable, skip the test
   ./python/iot3/tests/test-iot3-core-mqtt.
5. Test CAM emitedby its_vehicle:
   1. start an MQQT client listening to CAM:

      $ mosquitto_sub -t '#' -C 1 -F %J |jq .
      

   2. create a config file:

      (docker)🐍 cat >/tmp/its-vehicle.cfg <<-_EOF_
      [general]
      station-uuid = ora_car_1234
      station-type = passengerCar
      length = 4.335
      width = 2.257
      type = CAM
      depth = 26
      messages = cam cpm denm
      depth-sub-cam = 18
      depth-sub-cpm = 18
      depth-sub-denm = 15
      mirror-self = False
      report-freq = 5.0
      topic-pub-prefix = default/inQueue/v2x/
      topic-sub-prefix = default/outQueue/v2x/
      speed-thresholds = 8.33 19.44 27.78
      
      [broker.main]
      host = localhost
      port = 1883
      tls = false
      
      [gpsd]
      port=2948
      persistence = 1.0
      _EOF_
      

   3. start its-vehicle:

      (docker)🐍 its-vehicle -c /tmp/its-vehicle.cfg
      

   4. when the MQTT client dumps a message, terminate its-vehicle with Ctrl-C

Expected results

1. The test environment is ready
2. The packages were all built successfully
3. The packages were all installed successfully
4. The tests all succeed (no exception is raised)
5. The CAM are version 2.4.0; they have dimensions and station type, and the
   station_id is the third part of the source_uuid:

     "tst": "2026-08-20T12:48:14.862572+0200",
     "topic": "default/inQueue/v2x/cam/ora_car_1234/0/3/1/3/1/0/3/1/2/3/2/3/1/3/3/1/3/0/3/3/0/2/0/2/2/2",
     "qos": 0,
     "retain": 0,
     "payloadlen": 853,
     "payload": {
       "message_type": "cam",
       "message_format": "json/raw",
       "source_uuid": "ora_car_1234",
       "timestamp": 1787222894862,
       "version": "2.4.0",
       "message": {
         "protocol_version": 1,
         "station_id": 1234,
         "generation_delta_time": 33942,
         "basic_container": {
           "station_type": 5,
           "reference_position": {
             "latitude": 533613367,
             "longitude": -65056200,
             "altitude": {
               "value": 11690,
               "confidence": 11
             },
             "position_confidence_ellipse": {
               "semi_major": 1957,
               "semi_minor": 1957,
               "semi_major_orientation": 0
             }
           }
         },
         "high_frequency_container": {
           "basic_vehicle_container_high_frequency": {
             "heading": {
               "value": 317,
               "confidence": 127
             },
             "speed": {
               "value": 1,
               "confidence": 127
             },
             "drive_direction": 2,
             "vehicle_length": {
               "value": 43,
               "confidence": 4
             },
             "vehicle_width": 23,
             "longitudinal_acceleration": {
               "value": 161,
               "confidence": 102
             },
             "curvature": {
               "value": 1023,
               "confidence": 7
             },
             "curvature_calculation_mode": 2,
             "yaw_rate": {
               "value": 32767,
               "confidence": 8
             }
           }
         }
       }
     }
   }
   

[ymorin-orange]

@mathieu1fb Just because you were so curious this morning, here's the PR! ;-)