[jtc] Improve trajectory sampling efficiency

[RobertWilbrandt]

This PR addresses the increasing effort required in Trajectory::sample() raised in #1293 by exploiting the monotonically increasing access time for any trajectory. Instead of scanning through the complete trajectory to find the relevant trajectory points for the current sampling time, the Trajectory class keeps track of the last accessed index and only searches from there on. This supersedes the proposed binary search in #1294 and is based on #474.

Based on the requests in #474, i added checks for the new trajectory state to existing tests and added a new test to make sure that the index is correctly reset when calling Trajectory::update(). I verified that this indeed solves the problem by re-running the trajectory from #1293 and again plotting the periods between write() calls in my hardware interface:

The upper plot shows the progression of the period before this change, the lower one the behavior wtih this PR applied. The large spikes at the beginning and end are the start and end of the trajectory, the spikes in the middle are probably a result of me testing this on my normal laptop without RT kernel and doing some things on the side.

[christophfroehlich]

LGTM, thanks!

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 80.63%. Comparing base (7c89c17) to head (aa5aff5).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1297      +/-   ##
==========================================
+ Coverage   80.57%   80.63%   +0.06%     
==========================================
  Files         109      109              
  Lines        9553     9594      +41     
  Branches      832      834       +2     
==========================================
+ Hits         7697     7736      +39     
  Misses       1578     1578              
- Partials      278      280       +2     

Flag	Coverage Δ
unittests	80.63% <100.00%> (+0.06%)	⬆️

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

Files with missing lines	Coverage Δ
...include/joint_trajectory_controller/trajectory.hpp	75.00% <100.00%> (+1.66%)	⬆️
joint_trajectory_controller/src/trajectory.cpp	91.30% <100.00%> (+0.14%)	⬆️
...int_trajectory_controller/test/test_trajectory.cpp	100.00% <100.00%> (ø)

... and 3 files with indirect coverage changes