🐛 Vectorisation Formatting (MACOS support only - Temporary)

Summary

• Static wrapper functions around float & doubles processing per platform.
• Formats being supported (In progress & open to other libraries):
  • Apple Accelerate & VDSP (In Progress).
  • Windows Intel ipp (Supporting in the near future)

Why use a functional Style?

Functional allows for function overloading to extend support for any platform, without having to rewrite algorithms per platform. So, Even if I want to move to embedded or linux, or even support a platform with no inbuilt vectorisation. It's completely open.

Interface Implementation details

• Implement functions with seperate output pointer to allowing for the most amount of customisation.
• Stides or intervals are customisable per buffer (⚠️ At the moment there is no bounds checks).
• Functions are separated in nested namespaces to be more explicit.
• Future changes: std::span will be replacing the raw pointers at some point (🛟 It's a lot saver).

Including, Linking & Building

• Wrappers are included inside VIVI DSP at the moment as I write the VIVI_DSP library.
• External Dependencies:
  • Catch2 is being used for unit testing wrapper functions (Common sense testing).
  • Accelerate (VDSP) for macOS.