Add support for Autoware compatible point types XYZIRCAEDT and XYZIRC

[marioney]

This PR adds native support for PointXYZIRC and PointXYZIRCAEDT point types.

These point types are strictly required by Autoware across its entire perception and localization pipeline. By supporting these natively, we eliminate the need for an external, CPU-intensive point-cloud adaptation node which I've been using so far.

I cherry-picked some initial changes from jkk-research/ouster-ros@672b960 for the XYZIRC implementation and expanded upon them to build full native support for the extended XYZIRCAEDT format.

Related Issues & PRs

This issue has been requested by the community before in:

• Not able to cooperate with Autoware under the namespace "sensing" #383
• ROS ouster Point struct meaning, and Autoware implmentation for enhancing data field of Point  #217

It has also been actively discussed within the Autoware foundation:

• https://github.com/orgs/autowarefoundation/discussions/5293

Summary of Changes

• Point Type Definitions: Added PointXYZIRC and PointXYZIRCAEDT custom PCL structs to common_point_types.h.
• Factory Registration: Registered the xyzirc and xyzircaedt profiles in point_cloud_processor_factory.h for use in parameter configurations.
• Transformation Logic (point_transform.h):
  • channel: Mapped to native ring but inverted the ordering (top-to-bottom) to retain Velodyne compatibility as expected by Autoware.
  • intensity: Included Autoware-specific scaling (reflectivity / 64).
  • return_type: Mapped native return IDs (e.g., 1 for strongest, 2 for last).
  • azimuth, elevation, distance: Integrated calculations natively within the driver's point-filling loop to save external compute.
  • time_stamp: Mapped the native per-point nanosecond timestamp offset.
• Documentation: Updated README.md and configuration files (driver_params.yaml, os_sensor_cloud_image_params.yaml) to describe the new point type parameters.

Validation

• Compilation & Tests: The package builds cleanly and all internal ouster_ros unit tests pass successfully (16/16).
• **Execution **: Launched the driver specifying point_type: xyzircaedt and confirmed that the output sensor_msgs/PointCloud2 correctly encodes all 10 fields, matching the Autoware binary format specification.
• Autoware Compatibility: Integrated it directly into an active Autoware deployment. The point clouds were successfully ingested by Autoware's sensing, preprocessing, and localization pipelines without any compatibility warnings.