NIRSpec IFU associations missing imprints for some exposures

[stscijgbot-jp]

Issue JP-3853 was created on JIRA by Christian Hayes:

The associations for IFU exposures with multiple dithers, dedicated background exposures, and separate imprint/leakcals for the science and background exposures are missing imprints in the association files for some of the science exposures.  The associations should contain the science exposure, the appropriate imprint/leakcal for that exposure, all of the dedicated backgrounds, and all of the background imprints/leakcals.  In the asn for the first dither position everything appears to be there, but for subsequent dither positions, the imprint for the science exposure is missing.

I see this in PID 2162 obs 1 and PID 1554 obs 1 which both have one imprint for mulitple science exposures (1554 has two separate imprints for the two science targets) and one imprint for the multiple background exposures.  For the exposures with dither point 2+ the asn does not include the science imprint (i.e., the imprint with meta.observation.bkgdtarg=False).  This appears to affect observations that have one imprint each for science and background.  It's not clear if this is also affecting observations with imprints for each dither point in both science and background dithers, since many (all?) of these have backgrounds in a separate observation that are linked via background candidate associations.

[stscijgbot-jp]

Comment by David Law on JIRA:

Adding another example case: PID 1192 Obs 16