BUG: Sparse indexing with bool sparse may be incorrect

[sinhrks]

• tests added / passed
• passes git diff upstream/master | flake8 --diff
• whatsnew entry

indexing with SparseArray with bool dtype may return incorrect result because of some internal conversions.

arr = pd.SparseArray([1, 2, 3])
indexer = pd.SparseArray([True, False, True], fill_value=False, dtype=bool)

# NG
arr[indexer]
# [2.0, 2.0]
# Fill: nan
# IntIndex
# Indices: array([0, 1], dtype=int32)

# Expected (after this PR)
arr[indexer]
# [1.0, 3.0]
# Fill: nan
# IntIndex
# Indices: array([0, 1], dtype=int32)

[codecov-io]

Current coverage is 85.28% (diff: 83.33%)

Merging #13985 into master will decrease coverage by <.01%

@@             master     #13985   diff @@
==========================================
  Files           139        139          
  Lines         50206      50211     +5   
  Methods           0          0          
  Messages          0          0          
  Branches          0          0          
==========================================
+ Hits          42818      42822     +4   
- Misses         7388       7389     +1   
  Partials          0          0          

Powered by Codecov. Last update 29d9e24...0909fa8

[jreback]

thanks!