Cache camera geometries by camera name

[maxnoe]

Just using lru_cache is not possible, as numpy arrays and dicts are not hashable.

We could use cachetools but here is a solution by hand.

[kosack]

Just using lru_cache is not possible, as numpy arrays and dicts are not hashable.

Yes, I think that's why I originally added a custom __hash__ function to CameraGeometry, but it was kind of a hack.

[codecov]

Codecov Report

Merging #1054 into master will increase coverage by 0.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #1054      +/-   ##
==========================================
+ Coverage   83.37%   83.39%   +0.01%     
==========================================
  Files         186      186              
  Lines       10583    10595      +12     
==========================================
+ Hits         8824     8836      +12     
  Misses       1759     1759

Impacted Files	Coverage Δ
ctapipe/io/simteleventsource.py	98.93% <100%> (+0.04%)	⬆️
ctapipe/io/tests/test_simteleventsource.py	100% <100%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 87273b3...859c3a9. Read the comment docs.

[maxnoe]

Yes, I think that's why I originally added a custom hash function to CameraGeometry, but it was kind of a hack.

The CameraGeometry is hashable, the inputs you need to create one are not. This is a different problem.