Introduce a way to parse byte array in specific encoding into Codepoints

[pepelev]

This may be ref struct that wraps span and allows to enumerate it.

This object also must support feeding of another span as continuation of original span and so forth.

It also may be useful to support non ref struct (over Memory<T>) to allow it to be stored in fields

[pepelev]

Also it is worth to add a way to parse char sequence into Codepoints