Added null check for Offline LOI

[anandwana001]

Fixes #2773

Earlier app crash when any offline LOI is not available.
Added Null check of such LOI access logic

@shobhitagarwal1612 @sufyanAbbasi PTAL?

[codecov]

Codecov Report

Attention: Patch coverage is 21.87500% with 25 lines in your changes missing coverage. Please review.

Project coverage is 61.89%. Comparing base (c3ad147) to head (7819f36).
Report is 3 commits behind head on master.

Files with missing lines	Patch %	Lines
.../android/ground/repository/SubmissionRepository.kt	0.00%	15 Missing ⚠️
.../ground/repository/LocationOfInterestRepository.kt	37.50%	2 Missing and 3 partials ⚠️
...ndroid/ground/ui/syncstatus/SyncStatusViewModel.kt	20.00%	4 Missing ⚠️
...round/ui/datacollection/DataCollectionViewModel.kt	75.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##             master    #2788      +/-   ##
============================================
- Coverage     61.96%   61.89%   -0.08%     
  Complexity     1166     1166              
============================================
  Files           265      265              
  Lines          6223     6235      +12     
  Branches        877      883       +6     
============================================
+ Hits           3856     3859       +3     
- Misses         1840     1847       +7     
- Partials        527      529       +2     

Files with missing lines	Coverage Δ
...round/ui/datacollection/DataCollectionViewModel.kt	71.57% <75.00%> (-0.37%)	⬇️
...ndroid/ground/ui/syncstatus/SyncStatusViewModel.kt	45.45% <20.00%> (-7.18%)	⬇️
.../ground/repository/LocationOfInterestRepository.kt	50.87% <37.50%> (-0.11%)	⬇️
.../android/ground/repository/SubmissionRepository.kt	19.51% <0.00%> (-1.01%)	⬇️

[gino-m]

Earlier app crash when any offline LOI is not available.

Do you mean "when no LOIs are available offline"?

[gino-m]

Any idea why this might be happening? Was the LOI removed remotely, or perhaps the ID in the local db changed?

[gino-m]

Nvm, I see the LocationOfInterestMutation references the LOI by UUID, not local DB primary key.

If this happens, shouldn't we show the mutation as "failed" in the sync status screen? To do that, we may need to introduce a new type of error state in SyncStatus, UNRECOVERABLE_ERROR for ex. We could also rename FAILED to PENDING_RETRY or something clearer and update its current incorrect ktdoc in SyncStatus.

[kenstershiro]

@sufyanAbbasi to try to take a look at the PR approval in @gino-m 's absence

[sufyanAbbasi]

Sorry for the delay, I think the code looks good! As @gino-m pointed out, we will definitely want to indicate irreversibly failing mutations, maybe by taking the diff between the map and mapNotNull and marking those in the UI as broken.