TF prefix helper for diff drive controller

[kuralme]

Used new tf frame prefix helper in differential drive controller to test out & showcase the helper. This is the first part of the changes, the other controllers will come next:

• mecanum_drive_controller
• omni_wheel_drive_controller

Related issue

[saikishor]

I'm not sure if it makes sense to have it in the control_toolbox package. What about controllers_interfaces/helpers.hpp existing file?

What do you think @christophfroehlich ?

[christophfroehlich]

I'm not sure if it makes sense to have it in the control_toolbox package. What about controllers_interfaces/helpers.hpp existing file?

OK, maybe you are right and we should put it here.

[kuralme]

I'm not sure if it makes sense to have it in the control_toolbox package. What about controllers_interfaces/helpers.hpp existing file?

OK, maybe you are right and we should put it here.

I will start working on it

[christophfroehlich]

There is a behavior-breaking change now, right?
Can we do that gracefully? e.g.,

• Deprecate the tf_frame_prefix_enable parameter in the yaml, and log a warning message if it is set to a value different than default. And we should maintain the behavior and still use the value.
• Please update the description of the tf_frame_prefix parameter
  https://github.com/ros-controls/ros2_controllers/blob/master/diff_drive_controller/src/diff_drive_controller_parameter.yaml#L49-L53
• Please add release and migration notes.

[codecov]

Codecov Report

❌ Patch coverage is 77.77778% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 84.88%. Comparing base (36cfca8) to head (6c0ac1f).

Files with missing lines	Patch %	Lines
...iff_drive_controller/src/diff_drive_controller.cpp	50.00%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1997      +/-   ##
==========================================
- Coverage   84.91%   84.88%   -0.03%     
==========================================
  Files         148      148              
  Lines       14367    14353      -14     
  Branches     1230     1228       -2     
==========================================
- Hits        12200    12184      -16     
- Misses       1740     1741       +1     
- Partials      427      428       +1     

Flag	Coverage Δ
unittests	84.88% <77.77%> (-0.03%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...ive_controller/test/test_diff_drive_controller.cpp	93.99% <100.00%> (-0.20%)	⬇️
...iff_drive_controller/src/diff_drive_controller.cpp	82.17% <50.00%> (-0.83%)	⬇️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[kuralme]

A note about my last commit: This is for the deprecations. I am aware of Bence's request, I will apply if a change needed to the helper source.

[christophfroehlich]

tf_frame_prefix_enable was not unused before? Instead, during the deprecation phase, we still should use this parameter before we remove it.

I'm not sure how we can gracefully transition the usage of node namespace? Currently, if tf_frame_prefix is not set, then automatically the node namespace is used.

[christophfroehlich]

(failing binary build are fine, I've released ros2_control but rosdistro is on hold for the upcoming sync).

[christophfroehlich]

We have the following failing test:

     [  FAILED  ] TestDiffDriveController.configure_succeeds_tf_blank_prefix_no_namespace
    
     1 FAILED TEST
    -- run_test.py: return code 1
    -- run_test.py: inject classname prefix into gtest result file '/home/runner/work/ros2_controllers/ros2_controllers/.work/target_ws/build/diff_drive_controller/test_results/diff_drive_controller/test_diff_drive_controller.gtest.xml'
    -- run_test.py: verify result file '/home/runner/work/ros2_controllers/ros2_controllers/.work/target_ws/build/diff_drive_controller/test_results/diff_drive_controller/test_diff_drive_controller.gtest.xml'
  >>>
build/diff_drive_controller/test_results/diff_drive_controller/test_diff_drive_controller.gtest.xml: 25 tests, 0 errors, 1 failure, 0 skipped
- diff_drive_controller.TestDiffDriveController configure_succeeds_tf_blank_prefix_no_namespace (/home/runner/work/ros2_controllers/ros2_controllers/.work/target_ws/src/ros2_controllers/diff_drive_controller/test/test_diff_drive_controller.cpp:333)
  <<< failure message
    /home/runner/work/ros2_controllers/ros2_controllers/.work/target_ws/src/ros2_controllers/diff_drive_controller/test/test_diff_drive_controller.cpp:351
    Expected equality of these values:
      controller_->odometry_message_.header.frame_id
        Which is: "/odom"
      odom_id
        Which is: "odom"
  >>>

I think this is related to #1999 and should be discussed.

[kuralme]

This inner if block is wrong imo. It skips slash normalization so the test fail, and the new tilde-replace-namespace not applied as well. That should also address #1999 with the tilde usage.

[christophfroehlich]

But this still breaks current behavior, right? That's what I'm concerned about. Any idea how to deprecate that gracefully? We could add a new parameter to opt-in the new behavior, together with a warning when it is not set. This can then be fully removed one release after next.