Rjd 1333/previous following lanelets

[gmajrobotec]

Description

Abstract

Function getFollowingLanelets has confusing functionality or imprecise naming. Consider the following invocation: getFollowingLanelets(120660, {120660, <random-lanelets>}, 10000, true), then will be added to the returned vector unconditionally.
Function getPreviousLanelets has a mistake. The function is supposed to process one lanelet in every iteration, but every iteration starts from the lanelet passed as the argument. The result will be the lanelet passed as the argument and many copies of the one - previous lanelet. The mistake was introduced in change tier4@3fc8c0a#diff-da44510bdbbba766d1ba47318640cfd8bcff2e350eafe3d77d364bfbf70e25cdL745-L770. The previous implementation seems to have been right.

Details

Function getFollowingLanelets was changed to check if candidates lanelets are actually following each other.
Function getPreviousLanelets was rewritten to match previous implementation.

References

Jira ticket: internal link

Destructive Changes

There are no destructive changes.

[robomic]

HdMapUtils::getFollowingLanelets is not readable as it is for me. You should provide a more detailed explanation of its functionality and what candidate_lanelet_ids is and can/cannot contain

[robomic]

I think working with indices instead of iterators in getFollowingLanelets would be more convenient at this point, but this is good as it is