[Enhancement]: Enable Uint construction from fixed bytes

[joe-p]

Problem

Currently Uint256 cannot be constructed directly from a bytes<32>. This means you must go from bytes<32> -> biguint -> Uint256. This works, but it increases opcode cost due to the usage of b| because we loose the length information when going to biguint. For example, here is the diff of the TEAL when switching a type from bytes<32> to Uint256:

Part of them problem here is also that the compiler does not seem to know the length of the output of bzero, which is should when the argument is a constant.

Solution

No response

Proposal

Allow a bytes<32> to be passed to the ctor of Uint256. This also applies to all other UintN types

Pros & Cons

No response

Dependencies

No response