feat: Better ByteArray manipulation and reading

[adambor]

Feature Request

Describe the Feature Request

Currently the ByteArray implementation has just a rudimentary set of operations and doesn't allow reading/writting primitive types other than u8. This means a project which wants to read i.e. u32 or u64 from an array is forced to create an own implementation which reads the data byte-by-byte from the underlying ByteArray, example here. Having a possibility to directly read/write primitive types (u8 - u256) from/to ByteArray can significantly improve developer experience and also result in a significant reduction of step count.

Describe Preferred Solution

Since the underlying data structure of the ByteArray is an array of bytes31, all of the primitive read/write operations (u8 - u256) can be done in at most 2 array accesses, this can significantly reduce step count of programs relying heavily on reading/writting binary data.

Describe Alternatives

The Alexandria library also offers a variety of functions to read/write primitive types from the ByteArray however all of those only read the values byte-by-byte and don't leverage the possibility to read the values as a whole in a fewer array accesses.

Related Code

An example code of how this can be accomplished is available here. This also provides a to_u32_array which is useful when trying to sha256 hash the ByteArray (currently the core library uses byte-by-byte reading when passing the ByteArray to the sha256 hash function).

If the feature request is approved, would you be willing to submit a PR?
(Help can be provided if you need assistance submitting a PR)

• Yes
• No

[orizi]

In the next release (coming soon) some update allowing for more efficient by byte iteration is added, as well as some ByteSpan support.

in general - not sure when extracting "word by word" would be sensible for us - as it somewhat means that we have to talk about endianity there - which i'd rather avoid unless we have some extremely efficient versions.