ls_public_bucket.py depends on non-existent create_dataset function

[jdh-ama]

Expected behaviour

All of the imports used in ls_public_bucket.py exist somewhere.

Actual behaviour

from datacube.scripts.dataset import create_dataset, parse_match_rules_options

refers to a non existent function which was removed from datacube.scripts.dataset in commit 0f71e70f58a25f0e473fa872ff5938d8a95749b9

create_dataset first appears in release 1.6.0. It is removed in release 1.6.1 and then the ls_public_bucket.py script which imports create_dataset was added in release 1.7.0.

Steps to confirm the behaviour

git checkout tags/datacube-1.7
git log -G create_dataset

commits of interest are:
638ff76eb22c5f5c61549a58b54c59c901baa822
0f71e70f58a25f0e473fa872ff5938d8a95749b9

Environment information

• Which datacube --version are you using?

Open Data Cube core, version 1.7+247.g8d517db0

• What datacube deployment/enviornment are you running against?
  N/A, issue exists in the repository

[jdh-ama]

Issue opendatacube/datacube-core#511 is related to this. One proposed fix is to update the ls_public_bucket.py script to use the new Doc2Dataset class that replaces the create_dataset function.