Back to report index

guardrails-seg3d 9f189cc: AI3D-379 Review fixes: sync __version__ to 0.2.1, correct --set docstring

Miroslav Simko <ms@iolabs.ch> 2026-09-02T08:58:10+02:00

Commit #62 ยท 3 snippets

 guardrails/__init__.py | 2 +-
 guardrails/config.py   | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
Importance #1: guardrails/config.py @@ -3,9 +3,9 @@
3Mirrors the config convention used by the iolabs point-cloud packages3Mirrors the config convention used by the iolabs point-cloud packages
4(``iolabs_point_cloud_segmentation_trajectory`` etc.): the package owns a4(``iolabs_point_cloud_segmentation_trajectory`` etc.): the package owns a
5``guardrails.default.json`` algorithm config, and a typed params object5``guardrails.default.json`` algorithm config, and a typed params object
6(:class:`DetectorConfig`) is loaded from it at CLI start. Runtime overrides are6(:class:`DetectorConfig`) is loaded from it at CLI start. Runtime overrides are
7applied through repeatable ``--set PATH=VALUE`` flags, never repo-local JSON.7applied through repeatable ``--set KEY=VALUE`` flags, never repo-local JSON.
8``config.py`` is the loader/schema: the pydantic model field set is the schema8``config.py`` is the loader/schema: the pydantic model field set is the schema
9and every field default is kept identical to ``guardrails.default.json``9and every field default is kept identical to ``guardrails.default.json``
10(guarded by a unit test), so ``DetectorConfig()`` and ``load_config()`` agree.10(guarded by a unit test), so ``DetectorConfig()`` and ``load_config()`` agree.
1111
Importance #2: guardrails/__init__.py @@ -1,4 +1,4 @@
1"""Classical geometric guardrail detection for MLS point clouds."""1"""Classical geometric guardrail detection for MLS point clouds."""
22
3__version__ = "0.1.0"3__version__ = "0.2.1"
44
Importance #3: guardrails/config.py @@ -3,9 +3,9 @@
3Mirrors the config convention used by the iolabs point-cloud packages3Mirrors the config convention used by the iolabs point-cloud packages
4(``iolabs_point_cloud_segmentation_trajectory`` etc.): the package owns a4(``iolabs_point_cloud_segmentation_trajectory`` etc.): the package owns a
5``guardrails.default.json`` algorithm config, and a typed params object5``guardrails.default.json`` algorithm config, and a typed params object
6(:class:`DetectorConfig`) is loaded from it at CLI start. Runtime overrides are6(:class:`DetectorConfig`) is loaded from it at CLI start. Runtime overrides are
7applied through repeatable ``--set PATH=VALUE`` flags, never repo-local JSON.7applied through repeatable ``--set KEY=VALUE`` flags, never repo-local JSON.
8``config.py`` is the loader/schema: the pydantic model field set is the schema8``config.py`` is the loader/schema: the pydantic model field set is the schema
9and every field default is kept identical to ``guardrails.default.json``9and every field default is kept identical to ``guardrails.default.json``
10(guarded by a unit test), so ``DetectorConfig()`` and ``load_config()`` agree.10(guarded by a unit test), so ``DetectorConfig()`` and ``load_config()`` agree.
1111