UnboundLocalError with Trigdat._reconcile_timescales.mask

[derekocallaghan]

When loading a Trigdat file with Trigdat.open():

File ~/gitrepos/gifts-6u/gifts-ground-software/gdt/missions/fermi/gbm/trigdat.py:171, in Trigdat.open(cls, file_path, **kwargs)
    168 obj._rates = obj._data['RATE'].reshape(-1, 14, 8)
    170 # store the position history
--> 171 idx, dt = obj._time_indices(1024)
    172 eic = obj._data['EIC'][idx]
    173 quat = obj._data['SCATTITD'][idx]

File ~/gitrepos/gifts-6u/gifts-ground-software/gdt/missions/fermi/gbm/trigdat.py:664, in Trigdat._time_indices(self, time_res)
    662 cnt = len(idx)
    663 # reconcile 8 s and 1 s data
--> 664 idx = self._reconcile_timescales(back_idx, idx)
    666 # reconcile 8 s + 1 s and 256 ms data
    667 if time_res <= 256:

File ~/gitrepos/gifts-6u/gifts-ground-software/gdt/missions/fermi/gbm/trigdat.py:703, in Trigdat._reconcile_timescales(self, idx1, idx2)
    700 end_times2 = self._data['ENDTIME'][idx2]
    702 # find where bracketing timescale ends and inserted timescale begins
--> 703 if mask.sum():
    704     mask = end_times1 >= start_times2[0]
    705     start_idx =  (np.where(mask))[0][0]

UnboundLocalError: cannot access local variable 'mask' where it is not associated with a value

This appears to be related to the changes in https://github.com/USRA-STI/gdt-fermi/pull/22, where mask isn't declared prior to usage:

https://github.com/USRA-STI/gdt-fermi/blob/ad6c3c3995efe8138ef170504ecf68a6786dc67e/src/gdt/missions/fermi/gbm/trigdat.py#L700-L712

It looks like moving the mask initialization to before the first if mask.sum() statement fixes it, i.e.:

diff --git a/src/gdt/missions/fermi/gbm/trigdat.py b/src/gdt/missions/fermi/gbm/trigdat.py
index f23b9bc..348d08f 100644
--- a/src/gdt/missions/fermi/gbm/trigdat.py
+++ b/src/gdt/missions/fermi/gbm/trigdat.py
@@ -700,8 +700,8 @@ class Trigdat(FitsFileContextManager):
         end_times2 = self._data['ENDTIME'][idx2]

         # find where bracketing timescale ends and inserted timescale begins
+        mask = end_times1 >= start_times2[0]
         if mask.sum():
-            mask = end_times1 >= start_times2[0]
             start_idx =  (np.where(mask))[0][0]
             idx = np.concatenate((idx1[0:start_idx], idx2))

[BillCleveland-USRA]

This issue was fixed with pull request #28.