Design data sample management

[jrobinAV]

Taipy uses data sets for many purposes (documentation, templates, demos, tests, etc...).
With the blueprint functionality, the usage of data samples will increase.

Taipy must store, manage, expose, document, and maintain many datasets.

This issue aims to define, describe, and design this functionality into the following document:
https://docs.google.com/document/d/1NE5FJc3un1OCqHCzpmb9tYiPXBve-VXatm55WsmSivs/edit

[trgiangdo]

We can store the dataset in Azure, along side its MD5.

When a dataset is pulled, the code needs to check if the dataset has existed, and its MD5 matches the one on cloud or not.