feat: create function for devs to customize patterns

[richard-ramos]

Problem

All patterns from the 35/WAKU2-NOISE RFC are available https://github.com/waku-org/js-noise/blob/master/src/patterns.ts

However, the API has not been reviewed and there are no examples to demonstrate their usage.

Acceptance Criteria

For each handshake:

• Review the API
• Include a test
• Provide a usage example
• Provide some docs

Priority: do it on a need basis

Check list:

• Noise_XK1_25519_ChaChaPoly_SHA256: Should be sorted with WebRTC for Waku js-waku#1181. Improve the API, ensure there is test coverage as it's being used.
• Noise_WakuPairing_25519_ChaChaPoly_SHA256: Should already be covered by the library, to double check (API and coverage) once above is done so we are familiar with API
• Noise_XX_25519_ChaChaPoly_SHA256
• Noise_XXpsk0_25519_ChaChaPoly_SHA256
• Noise_K1K1_25519_ChaChaPoly_SHA256

Notes

The hashing, cipher and curve are currently hardcoded (SHA-256, ChaChaPoly and ed25519). This is as per the RFC so seems find to leave as is.

[fryorcraken]

@richard-ramos can we have more details please? :)

[richard-ramos]

We currently have these patterns available https://github.com/waku-org/js-noise/blob/master/src/patterns.ts with those specific protocolIDs

- Noise_K1K1_25519_ChaChaPoly_SHA256: 10,
- Noise_XK1_25519_ChaChaPoly_SHA256: 11,
- Noise_XX_25519_ChaChaPoly_SHA256: 12,
- Noise_XXpsk0_25519_ChaChaPoly_SHA256: 13,
- Noise_WakuPairing_25519_ChaChaPoly_SHA256: 14,

These are the handshake patterns defined in the RFC, but maybe we should refactor the code so it's easy for a developer to be able to create new patterns, and allow them to specify the hashing and cipher to be used instead of hardcoding the sha256 and chachapoly, as well as specifying the type of curve to be used (instead of only allowing 25519).

[fryorcraken]

cool, do have tests for those?

The hardcoding seems fine because the RFC does not allow alternative hash or encryption scheme AFAIK

[fryorcraken]

@richard-ramos does the new description seem accurate?