[jog_arm] Pass planning scene monitor into cpp interface #1849
Conversation
davetcoleman
force-pushed
the
jog_arm_psm
branch
2 times, most recently
from
37a9fbe
to
2363b4b
Compare
davetcoleman
changed the title
- Separate LOGNAME into each file - Add unit test
davetcoleman
force-pushed
the
jog_arm_psm
branch
from
2363b4b
to
b7dca01
Compare
was this comment meant for moveit/moveit.ros.org#401? |
moveit_experimental/moveit_jog_arm/include/moveit_jog_arm/jog_interface_base.h
Show resolved
Hide resolved
|
|
||
| double velocity_scale_coefficient = -log(0.001) / parameters_.collision_proximity_threshold; | ||
| // Copy the planning scene's version of current state into new memory | ||
| robot_state::RobotState current_state(getLockedPlanningSceneRO()->getCurrentState()); |
There was a problem hiding this comment.
Great! Better decoupling of PSM.
| planning_scene_monitor::PlanningSceneMonitorPtr planning_scene_monitor_; | ||
| }; // class TestJogCppInterface | ||
|
|
||
| TEST_F(TestJogCppInterface, PlanningSceneTest) |
There was a problem hiding this comment.
These tests are not very informative yet, but they are a start 😉
There was a problem hiding this comment.
I realized need to do huge refactors of jog_arm to make it testable (separate ROS and simplify threading)
|
|
||
| TEST_F(TestJogCppInterface, InitTest) | ||
| { | ||
| moveit_jog_arm::JogCppApi jog_cpp_interface(planning_scene_monitor_); |
There was a problem hiding this comment.
What about inserting a ROS sleep for a second or so to give the threads some time to actually process data?
There was a problem hiding this comment.
I don't think we should add arbitrary sleeps until its proven needed.
There was a problem hiding this comment.
Currently, the thread is immediately stopped due to immediate destruction of JogCppApi.
Adding a sleep would give the thread same time to cycle.
moveit_experimental/moveit_jog_arm/test/jog_cpp_interface_test.test
Outdated
Show resolved
Hide resolved
moveit_experimental/moveit_jog_arm/include/moveit_jog_arm/collision_check_thread.h
Outdated
Show resolved
Hide resolved
davetcoleman
force-pushed
the
jog_arm_psm
branch
from
f60a1ea
to
5741c88
Compare
|
@rhaschke i force pushed changes to my feature branch, but just noticed you had some kind of formatting changes that got squashed, and i can't restore. what were they? |
davetcoleman
force-pushed
the
jog_arm_psm
branch
from
5741c88
to
167605c
Compare
That's why one should always use |
|
That's a cool trick I hadn't heard of! |
* [jog_arm] Pass planning scene monitor into cpp interface - Separate LOGNAME into each file - Add unit test * Add PR feedback
* [jog_arm] Pass planning scene monitor into cpp interface - Separate LOGNAME into each file - Add unit test * Add PR feedback
* [jog_arm] Pass planning scene monitor into cpp interface - Separate LOGNAME into each file - Add unit test * Add PR feedback
I need to pass in the PSM from a larger application. Previously it required creating a new PSM even in the same process