Initial version of dynamic ONVIF camera discovery with DLS pipeline startup

[jmotow]

Initial version of dynamic ONVIF camera discovery with DLS pipeline startup

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Introduces an asyncio-based ONVIF discovery sample that continuously detects cameras, queries ONVIF media profiles, and starts/stops GStreamer (DL Streamer) pipelines dynamically based on a JSON configuration.

Changes:

• Replaced the old one-shot discovery/launcher code with an async discovery engine + camera registry + pipeline lifecycle manager.
• Introduced a new config.json schema (named camera entries with hostname, port, definition) and verbose controls.
• Added wheel packaging assets and updated documentation for building/running the sample.

Reviewed changes

Copilot reviewed 16 out of 16 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
samples/gstreamer/python/onvif_cameras_discovery/requirements.txt	Replace invalid urllib dependency with urllib3
samples/gstreamer/python/onvif_cameras_discovery/misc.py	Add helper for printing discovered/removed cameras and profiles in a table
samples/gstreamer/python/onvif_cameras_discovery/dls_onvif_sample.py	New asyncio entrypoint wiring the discovery engine + graceful shutdown
samples/gstreamer/python/onvif_cameras_discovery/dls_onvif_discovery_utils.py	Remove legacy synchronous utilities (replaced by new engine)
samples/gstreamer/python/onvif_cameras_discovery/dls_onvif_discovery_thread.py	Add GStreamer pipeline launcher/stopper using a GLib loop thread
samples/gstreamer/python/onvif_cameras_discovery/dls_onvif_discovery_engine.py	Add unified WS-Discovery + ONVIF profile retrieval + async orchestration
samples/gstreamer/python/onvif_cameras_discovery/dls_onvif_data.py	Extend ONVIFProfile with camera connection metadata fields/properties
samples/gstreamer/python/onvif_cameras_discovery/dls_onvif_config_manager.py	Add config loader/refresh and pipeline lookup by ip:port
samples/gstreamer/python/onvif_cameras_discovery/dls_onvif_camera_entry.py	Add registry + entry model to bind cameras, profiles, pipelines, lifecycle state
samples/gstreamer/python/onvif_cameras_discovery/config.json	Update config to new structured format with named cameras and verbose flag
samples/gstreamer/python/onvif_cameras_discovery/build_whl/setup.py	Add setuptools packaging definition for building a wheel
samples/gstreamer/python/onvif_cameras_discovery/build_whl/pyproject.toml	Add build-system metadata for wheel builds
samples/gstreamer/python/onvif_cameras_discovery/build_whl/build_dls_onvif_sample_whl.sh	Add helper script to build/uninstall the wheel
samples/gstreamer/python/onvif_cameras_discovery/build_whl/README.md	Add instructions for building and validating the wheel
samples/gstreamer/python/onvif_cameras_discovery/build_whl/MANIFEST.in	Include README/requirements/config/assets in packaging inputs
samples/gstreamer/python/onvif_cameras_discovery/README.md	Rewrite docs to reflect async engine, new config format, and wheel build

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[oonyshch]

no unit tests are included for the new async discovery engine. given the lifecycle complexity (cameras appearing/disappearing, pipeline start/stop, rollback on failure), consider adding tests for at least: (1) camera discovered triggers pipeline start, (2) camera lost triggers pipeline stop, (3) startup failure triggers _cleanup_on_failed_start correctly. mock-based tests would significantly improve confidence here.

[oonyshch]

overall, good. please take attention before merge:

• two required ci checks: license header scan and pylint. run pylint locally on all new python files and fix issues.
• verify license header format matches existing files, also for CI check
• rebase branch with main.

additional recommendations:

• add unit tests for the async discovery lifecycle (camera discovered/lost, pipeline start/stop, failure rollback).
• see inline comments for potential deadlock issue and path traversal security concern.