Eiger support

[gfabbris]

The setup was heavily based on the NSLS2 CHX, and the handler was based on the area_detector_handlers.eiger.EigerHandler.

Some of the most important aspects:

• This was tested at the beamline, but without x-rays.

• The trigger is based on EPICS updating the image count (eiger.cam.array_counter), which is probably a bit slow.

• The stage/unstage process has a bunch of blocking calls that are not managed by the RunEngine. As far as I can tell this is also done by other area detector devices setup at NSLS2, but I think this is not great. This can get ugly if you try to "control+c" while these blocking calls are going on.

• The image file naming used here is probably debatable. It is controlled by eiger.file.base_name, and will automatically create a new folder using that base name. It will automatically try to add the eiger.file.seq_id in the same so that the folders are unique:

  # if:
  In  [1]: eiger.file.read_name = 'my_test_{}'
  In  [2]: eiger.file.seq_id.get()
  Out [2]: 675
  # Then file name used for this scan will be "my_test_675"

  https://github.com/APS-4ID-POLAR/ipython-polar/blob/1050815165d1dbc1e79ec5431789b8ed66ae23cb/profile_bluesky/startup/instrument/devices/ad_eiger.py#L144 but it will raise an OSError if the folder already exists! And note that our local scans will change the file names to be scan{}.format(RE.md["scan_id"]+1), which tracks the scan number. https://github.com/APS-4ID-POLAR/ipython-polar/blob/1050815165d1dbc1e79ec5431789b8ed66ae23cb/profile_bluesky/startup/instrument/plans/local_scans.py#L186

• The MyEigerHandler needs to be registered with the local db while reading. So it might be better to move it to polartools after this pull request is settled.

• This also adds the nanopositioner motors and local versions of the grid_scan and rel_grid_scan (with all the 4-ID-D specific pre-processing).

Closes #130.

[lgtm-com[bot]]

This pull request introduces 1 alert when merging 776c459a12e4f5357bdb8aea9a022f2464100222 into e9ac549a33e158efc2f321b2333a76a1c0de0c1c - view on LGTM.com

new alerts:

• 1 for Suspicious unused loop iteration variable

[lgtm-com[bot]]

This pull request introduces 1 alert when merging 12bd01b85215ddb5b5c2fc8b369e7802e5972e95 into e9ac549a33e158efc2f321b2333a76a1c0de0c1c - view on LGTM.com

new alerts:

• 1 for Suspicious unused loop iteration variable

[gfabbris]

The current trigger process can have a pretty bad dead time, because the PV being tracked is not optimal (the detector is ready to take another image before that). I'm exploring a change that might help in the eiger_new_mechanism banch.

[gfabbris]

@prjemian and @strempfer: This is ready for review.

I didn't test the latest version with the detector, but I'm willing to just merge and fix it later if needed (not sure when will have access to the detector).

Note that this creates two options of trigger, time-based and new-image-based, that can be selected during load.