Feat: Add SHA-256 hashing option

[mdambski]

Description

This change set introduces the sha256 hashing option as an alternative to md5. By default, the library will use md5 unless this hashing algorithm is unavailable (which would effectively break the library for the user).

Rationale

At the moment, the library uses md5 hashing by default and does not allow the option to choose an alternative.

In some distributions of Python that are compliant with FIPS Regulation, md5 will either be missing or will raise an error when used. These concerns are brought up in the Python docs related to hash algorithms.

In such environments, using a library that relies solely on md5 forces users to either abandon usage of the library or create and maintain their own fork with the needed patches. Both options are far from optimal.

cc: @Pouyanpi

Related Issue(s)

• Fixes feature: Add FIPS compliant hashing algorithm SHA256 #909

Checklist

• I've read the CONTRIBUTING guidelines.
• I've updated the documentation if applicable.
• I've added tests if applicable.
• @mentions of the person or team responsible for reviewing proposed changes.

[Pouyanpi]

Thank you very much @mdambski, the PR looks great, but I think we can make it simpler as the hashing in kb is not that critical.

Also, I think when we allow the users to configure the caching there is no need to set the default dynamically.

please have a look at my comments and let me know your thoughts, It could be that I've missed the actual purpose. Thanks!

[github-actions]

Documentation preview

https://nvidia.github.io/NeMo-Guardrails/review/pr-988

[mdambski]

Hi @Pouyanpi, all change requests are addressed. Please let me know if we can move forward 🙏

[Pouyanpi]

Thank you @mdambski you just need to remove the unused config folder

[mdambski]

@Pouyanpi the config was removed

[Pouyanpi]

Thanks, looks good to me.

[Pouyanpi]

Tests are failing for Python 3.9 (test_utils.py)

@mdambski would you please have a look?

[mdambski]

@Pouyanpi the tests are fixed 👍

[Pouyanpi]

@mdambski Thank you very much. I'll merge this PR later as it is decided to be part of 0.13.0 release. Probably Python 3.12 also needs to wait for 0.13.0, I'll let you know.