feat: SSA parser

[asterite]

Description

Problem

Resolves #6475

Summary

This PR introduces the Ssa::from_src method to parse SSA from text into Ssa. The code is first parsed into an intermediate (new) AST, then from it an SSA is created by using the existing FunctionBuilder.

Then, some existing tests are updated to use this form of creating the SSA instead of manually doing it with FunctionBuilder.

Some notes:

• In the tests that I updated I also updated how the final SSA is checked: now we check that the string representation matches the one we expect, instead of manually checking some things. I think this is a bit more robust and shorter. And then, if we trust SSA parsing and SSA printing is correct (and I think we do because when we debug things we rely on the SSA output we get) then I think it should be fine to rely on this.
• I had to change how some SSA are printed, specifically to include some types that otherwise can't be determined when parsing. For example, arrays flatten their values, so if we have [Field 1, Field 1] it could actually be an array of (Field, Field) but we can't know that. So, now arrays include their type, like this: [Field 1, Field 1] of Field (or whatever type they are). Other instructions like call, allocate and load specify what type they return with -> types... at the end. I chose that syntax so that the types come at the end and one can still focus on the important bits on the left side.
• The IfElse instruction is still not parsed in this PR. I couldn't find an example of it. But we can include it in a follow-up PR.
• The optional assertion message in constrain and range_check isn't parsed (yet). This could be done in a follow-up PR, but my thinking here was that it's not important for tests, and it seems no tests actually use these strings (but it could be good to have for completeness)
• I didn't update all tests to avoid making this PR bigger. I updated some tests to verify that the parser is working and that most, if not all, needed features are there. But we can update them in follow-up PRs, and even add more tests to existing passes that aren't tested.

Additional Context

Documentation

Check one:

• No documentation needed.
• Documentation included in this PR.
• [For Experimental Features] Documentation to be submitted in a separate PR.

PR Checklist

• I have tested the changes locally.
• I have formatted the changes with Prettier and/or cargo fmt on default settings.

[asterite]

I'll merge this now but try to improve the parsing error message in a follow-up PR.

[asterite]

I pushed a commit that will show errors in a better way. For example, here's one output where there's a typo:

called `Result::unwrap()` on an `Err` value:
            acir(inline) fn main f0 {
              b0(v0: Field):

                v1 = ad v0, Field 1
                     ^^
                     Expected an instruction or terminator, found ad

                v2 = mul v1, Field 3
                return v2
            }

[asterite]

I also renamed assert_ssa_equals to assert_normalized_ssa_equals because I figured it might be confusing if you print an SSA and get one value, that that value isn't equal to what that assertion is producing...