fix: dimos topic send

[paul-nechifor]

dimos topic send has been broken for certain messages since I removed __init__.py files.

Closes DIM-1124

[greptile-apps]

Greptile Summary

This PR fixes dimos topic send for namespace-package message modules. The main changes are:

• Message classes are discovered by walking message submodules.
• topic_send builds its eval context from discovered message classes.
• Tests cover eval-context discovery and LCM delivery for a Twist message.

Confidence Score: 4/5

This is close, but the import failure path should be fixed before merging.

• Happy-path message discovery now works for namespace packages.
• Discovered message modules with missing import-time dependencies are still skipped silently.
• Affected send commands can still fail with a misleading missing-name error.

dimos/robot/cli/topic.py

Important Files Changed

Filename	Overview
dimos/robot/cli/topic.py	Adds namespace-package message discovery for topic_send, but still suppresses import-time failures from discovered message modules.
dimos/robot/cli/test_topic.py	Adds tests for eval-context discovery and LCM delivery on the happy path.

_{Reviews (2): Last reviewed commit: "Merge branch 'main' into paul/fix/init-b..." | Re-trigger Greptile}

[greptile-apps]

Message Imports Disappear Silently

When a discovered message module has a missing import-time dependency, this branch skips it and removes that message class from the eval context. A valid command such as sending a class from that module then fails later with NameError instead of reporting the module import failure, so dimos topic send remains broken for those message types.

[codecov]

❌ 2 Tests Failed:

Tests completed	Failed	Passed	Skipped
2690	2	2688	71

View the top 1 failed test(s) by shortest run time

dimos.navigation.cmu_nav.modules.far_planner.test_far_planner_rosbag.TestFarPlannerRosbag::test_waypoint_accuracy

Stack Traces | 73.9s run time

self = <dimos.navigation.cmu_nav.modules.far_planner.test_far_planner_rosbag.TestFarPlannerRosbag object at 0x7755d4cda4e0>

    def test_waypoint_accuracy(self) -> None:
        """Feed identical inputs at original timing and compare waypoint output."""
        if not FAR_PLANNER_BIN.exists():
            pytest.skip(f"FAR planner binary not found: {FAR_PLANNER_BIN}")
    
        window = load_rosbag_window()
        ref_wp = window.way_point
        assert len(ref_wp) > 0, "No reference waypoints in fixture"
    
        lcm = lcmlib.LCM()
        wp_collector = LcmCollector(topic=WAYPOINT_OUT_LCM, msg_type=PointStamped)
        wp_collector.start(lcm)
    
        stop_event = threading.Event()
        handle_thread = threading.Thread(
            target=lcm_handle_loop, args=(lcm, stop_event), daemon=True
        )
        handle_thread.start()
    
        runner = NativeProcessRunner(binary_path=str(FAR_PLANNER_BIN), args=_far_planner_args())
    
        try:
            runner.start()
            assert runner.is_running, "FAR planner binary failed to start"
            time.sleep(_PROCESS_STARTUP_SEC)
    
            # Feed at original timing (1:1 with rosbag)
            feed_at_original_timing(
                lcm,
                window,
                topic_map={
                    "odom": ODOM_LCM,
                    "scan": SCAN_LCM,
                    "terrain": TERRAIN_LCM,
                    "terrain_ext": TERRAIN_EXT_LCM,
                    "goal": GOAL_LCM,
                },
            )
    
            time.sleep(_POST_FEED_DRAIN_SEC)
    
        finally:
            runner.stop()
            stop_event.set()
            handle_thread.join(timeout=DEFAULT_THREAD_JOIN_TIMEOUT)
            wp_collector.stop(lcm)
    
        our_wps = [(msg.x, msg.y) for msg in wp_collector.messages]
    
        # Compute deviation score
        score = _compute_waypoint_deviation(our_wps, ref_wp)
    
        # Log score for visibility
        logger.info(f"\n{'=' * 60}")
        logger.info("FAR PLANNER DEVIATION SCORE")
        logger.info(f"  Our waypoints:     {len(our_wps)}")
        logger.info(f"  Reference:         {len(ref_wp)}")
        logger.info(f"  Count ratio:       {score['count_ratio']:.3f}")
        logger.info(f"  Mean error:        {score['mean_error_m']:.3f} m")
        logger.info(f"  Max error:         {score['max_error_m']:.3f} m")
        logger.info(f"  Mean X diff:       {score['mean_x_diff']:.3f} m")
        logger.info(f"  Mean Y diff:       {score['mean_y_diff']:.3f} m")
        logger.info(f"{'=' * 60}\n")
    
        # Assertions — generous thresholds, the point is to measure
        assert len(our_wps) > 0, "FAR planner produced no waypoints"
>       assert score["mean_error_m"] < 5.0, (
            f"Mean waypoint error {score['mean_error_m']:.2f}m exceeds 5m threshold"
        )
E       AssertionError: Mean waypoint error 5.17m exceeds 5m threshold
E       assert 5.1664396514849225 < 5.0

handle_thread = <Thread(Thread-157 (lcm_handle_loop), stopped daemon 131209660180160)>
lcm        = <LCM object at 0x7755e1099920>
our_wps    = [(6.297216415405273, -3.322401762008667), (6.306988716125488, -3.3080596923828125), (6.3008599281311035, -3.2252717018...81311035, -3.225271701812744), (6.3008599281311035, -3.225271701812744), (6.3008599281311035, -3.225271701812744), ...]
ref_wp     = array([[     5.5327,        -0.3,         2.5,        0.75],
       [     5.7342,        -0.3,         2.5,        0.7...,      7.8589,      1.4869,    0.095978],
       [     65.337,      7.8589,      1.4869,    0.095978]], shape=(300, 4))
runner     = NativeProcessRunner(binary_path='.../result/bin/far_planner_...hred', '4', '--new_point_counter', '5', '--obs_inflate_size', '1', '--visualize_ratio', '0.4', '--wp_churn_dist', '0'])
score      = {'count_ratio': 0.55, 'max_error_m': 8.713269220479598, 'mean_error_m': 5.1664396514849225, 'mean_x_diff': 4.151850025653839, ...}
self       = <dimos.navigation.cmu_nav.modules.far_planner.test_far_planner_rosbag.TestFarPlannerRosbag object at 0x7755d4cda4e0>
stop_event = <threading.Event at 0x7755d797dcd0: set>
window     = RosbagWindow(odom=array([[     5.3674,           0,           0, ...,           0,           0,           1],
       [...,           1],
       [     0.2975,           0,           0,           0,           0,           0,           1]]))])
wp_collector = LcmCollector(topic='/rbfp_wp#geometry_msgs.PointStamped', msg_type=<class 'dimos.msgs.geometry_msgs.PointStamped.Point...1417, 5877690.999504224, 5877691.196800777, 5877691.398209056, 5877691.59548678, 5877691.794883912, 5877691.994291641])

.../modules/far_planner/test_far_planner_rosbag.py:276: AssertionError

View the full list of 1 ❄️ flaky test(s)

dimos.perception.test_image_embedding.TestImageEmbedding::test_clip_embedding_initialization

Flake rate in main: 30.00% (Passed 7 times, Failed 3 times)

Stack Traces | 26.3s run time

request = <SubRequest 'monitor_threads' for <Function test_clip_embedding_initialization>>

    @pytest.fixture(autouse=True)
    def monitor_threads(request):
        # Capture threads before test runs
        test_name = request.node.nodeid
        with _seen_threads_lock:
            _before_test_threads[test_name] = {
                t.ident for t in threading.enumerate() if t.ident is not None
            }
    
        yield
    
        with _seen_threads_lock:
            before = _before_test_threads.get(test_name, set())
    
        # Threads intentionally left running for the whole process and cleaned up on
        # exit, so they don't count as per-test leaks.
        expected_persistent_thread_prefixes = [
            "Dask-Offload",
            # HuggingFace safetensors conversion thread - no user cleanup API
            # https://github..../transformers/issues/29513
            "Thread-auto_conversion",
        ]
    
        def live_new_threads():
            # Threads created during this test that are still running. A thread that
            # has already stopped is not a leak -- it's done, just not yet reaped
            # from threading's registry -- so we key on is_alive(), not presence.
            result = []
            for t in threading.enumerate():
                if t.ident is None or t.ident in before or t.name == "MainThread":
                    continue
                if any(t.name.startswith(prefix) for prefix in expected_persistent_thread_prefixes):
                    continue
                if t.is_alive():
                    result.append(t)
            return result
    
        # Some C extensions tear their callback threads down asynchronously, so a
        # thread can stay alive briefly after the test cleaned up its owner (notably
        # zenoh's pyo3-closure threads, freed shortly after the session is closed).
        # A single snapshot races that teardown and flags a thread that is about to
        # exit. Give new threads a grace period to drain; only ones that stay alive
        # are real leaks (a genuinely leaked thread never exits).
        deadline = time.monotonic() + 5.0
        leaked = live_new_threads()
        while leaked and time.monotonic() < deadline:
            time.sleep(0.02)
            leaked = live_new_threads()
    
        if not leaked:
            return
    
        with _seen_threads_lock:
            # Report each leaked thread only once across the session.
            truly_new = [t for t in leaked if t.ident not in _seen_threads]
            for t in leaked:
                _seen_threads.add(t.ident)
    
        if not truly_new:
            return
    
        thread_names = [t.name for t in truly_new]
    
>       pytest.fail(
            f"Non-closed threads created during this test. Thread names: {thread_names}. "
            "Please look at the first test that fails and fix that."
        )
E       Failed: Non-closed threads created during this test. Thread names: ['Thread-249']. Please look at the first test that fails and fix that.

before     = {131208072812224, 131208081204928, 131208097990336, 131208106383040, 131208123168448, 131208131561152, ...}
deadline   = 5878350.437798342
expected_persistent_thread_prefixes = ['Dask-Offload', 'Thread-auto_conversion']
leaked     = [<TMonitor(Thread-249, started daemon 131208056026816)>]
live_new_threads = <function monitor_threads.<locals>.live_new_threads at 0x7755a57a2980>
request    = <SubRequest 'monitor_threads' for <Function test_clip_embedding_initialization>>
t          = <TMonitor(Thread-249, started daemon 131208056026816)>
test_name  = 'dimos/perception/test_image_embedding.py::TestImageEmbedding::test_clip_embedding_initialization'
thread_names = ['Thread-249']
truly_new  = [<TMonitor(Thread-249, started daemon 131208056026816)>]

dimos/conftest.py:273: Failed

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[leshy]

needs to become proto agnostic now with zenoh

[greptile-apps]

Import errors disappear

This still hides import-time failures from discovered message modules. When a message file is found but one of its dependencies is missing, importlib.import_module() raises ImportError, this branch skips the module, and the message class never enters the eval context. A command like topic_send("/camera", "Image(...)") then reports name 'Image' is not defined instead of the real missing import, so valid message types can still fail with a misleading parse error.

Suggested change

	try:
	submodule = importlib.import_module(f"{package_name}.{name}")
	except ImportError:
	continue
	submodule = importlib.import_module(f"{package_name}.{name}")