Beamlines that end in the middle of a branch should only include devices up to their beamline.
Current Behavior
Currently we expect beamlines to be located at the end of a branch. This works in most cases, but some beamlines (XPP) live in the middle of a branch, so their active paths end up including a variety of irrelevant devices.
Possible Solution
This will probably require a slightly different way of specifying where a beamline sits. Perhaps we also allow a device name to designate a beamline.
Steps to Reproduce (for bugs)
initialize the test database (path.json) and look at the active paths for XPP:
Note how the first beam path includes all the L0 devices, even though XPP should really end before xpp_lodcm
Expected Behavior
Beamlines that end in the middle of a branch should only include devices up to their beamline.
Current Behavior
Currently we expect beamlines to be located at the end of a branch. This works in most cases, but some beamlines (XPP) live in the middle of a branch, so their active paths end up including a variety of irrelevant devices.
Possible Solution
This will probably require a slightly different way of specifying where a beamline sits. Perhaps we also allow a device name to designate a beamline.
Steps to Reproduce (for bugs)
initialize the test database (
path.json
) and look at the active paths forXPP
:Note how the first beam path includes all the
L0
devices, even though XPP should really end beforexpp_lodcm
Context
leftover from #133
Your Environment
pcds-5.4.1 + #133