Fixed ffi of sliced arrays

[jorgecarleitao]

This PR fixes FFIying of sliced arrays. I thought that this was fixed, but #538 (thanks @lissahyacinth!) showed that it was not.

Draft because I think that this is still not correct, but I need to check the arrow spec for more details.

[codecov]

Codecov Report

Merging #540 (5510fe5) into main (b62184d) will increase coverage by 0.28%.
The diff coverage is 61.26%.

@@            Coverage Diff             @@
##             main     #540      +/-   ##
==========================================
+ Coverage   81.27%   81.55%   +0.28%     
==========================================
  Files         380      382       +2     
  Lines       23386    23429      +43     
==========================================
+ Hits        19006    19108     +102     
+ Misses       4380     4321      -59     

Impacted Files	Coverage Δ
src/array/binary/mod.rs	81.01% <ø> (+2.26%)	⬆️
src/array/boolean/ffi.rs	0.00% <0.00%> (ø)
src/array/dictionary/mod.rs	56.89% <ø> (+10.22%)	⬆️
src/array/ffi.rs	70.00% <ø> (+5.00%)	⬆️
src/array/fixed_size_binary/ffi.rs	0.00% <0.00%> (ø)
src/array/fixed_size_binary/mod.rs	50.70% <ø> (+3.26%)	⬆️
src/array/fixed_size_list/ffi.rs	0.00% <0.00%> (ø)
src/array/fixed_size_list/mod.rs	60.00% <ø> (+4.44%)	⬆️
src/array/list/mod.rs	75.00% <ø> (-0.30%)	⬇️
src/array/map/ffi.rs	0.00% <0.00%> (ø)
... and 32 more

---

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 b62184d...5510fe5. Read the comment docs.