Fix globbing top level buckets and

[pjbull]

#304 introduced a regression when globbing at the top level (CloudPath("s3://bucket").glob("*")) where the bucket was duplicated in the paths that glob yielded (CloudPath("s3://bucket/bucket/file1.txt") is retruned), which was reported in #311. Additionally, #311 mentions that CloudPath("s3://").glob("*") does not work as expected.

This PR adds two fixes:

• Fix the path construction logic in glob to not duplicate the bucket at the top level.
• Raise a not implemented error if globbing above a bucket level. Because of the way boto3 works, we would need to repeat our calls for each of the buckets to make globbing above a bucket work. This adds too much complexity. Instead, we give a friendly error that buckets can be listed with iterdir and .glob works within buckets.

Additionally we add a regression test and update setup.py and changelog to make a patch release to fix this bug.

[github-actions]

🚀 Deployed on https://deploy-preview-312--gallant-agnesi-5f7bb3.netlify.app

[codecov-commenter]

Codecov Report

Merging #312 (f780e22) into master (2556df0) will increase coverage by 0.0%.
The diff coverage is 100.0%.

Additional details and impacted files

@@          Coverage Diff           @@
##           master    #312   +/-   ##
======================================
  Coverage    94.1%   94.1%           
======================================
  Files          21      21           
  Lines        1393    1395    +2     
======================================
+ Hits         1312    1314    +2     
  Misses         81      81           

Impacted Files	Coverage Δ
cloudpathlib/cloudpath.py	92.5% <100.0%> (+<0.1%)	⬆️