chore(elaborator): Ensure that push_location and push_type cannot be forgotten

[aakoshh]

Description

Problem*

Resolves a potential for forgetting to set the location or type of an Expression or Statement. Makes it easier to review the code knowing that it cannot happen.

Summary*

Adds a PushedExpr construct which wrap an ExprId and show the status of what has been pushed in the type as well as with flags. If not all pieces are present when the variable goes out of scope, it panics. The fact that we have to unwrap or dereference means the compiler tells us when we need to push the type information.

The new type also makes it possible to demand that location has been pushed, which is used in at least one method that produces the Type: if we ask for PushedExpr<HasLocation> then the compiler ensures that there shouldn't be a panic when querying the interner for the location of the ID.

The only place I found the type was missing was in make_error.

Additional Context

Originally I added support for pushing the type first, and then the location, or vice versa, but in practice location is always immediately available, so I changed all call sites to use either push_expr_full, or intern_expr followed by a new intern_expr_type. Since we never do it the other way, I removed the option to start with push_type, thinking that since all of these are pub, should we add another necessary pushable thing later, I don't want to support 3*2 paths to completion, if it can be restricted to go in a linear fashion.

For statements it turns out that the location was always available, and so I just used a new push_stmt_full everywhere, but still added support for pushing separately. Originally a PushedStmt was added, but it was removed as it wasn't used, along with push_stmt.

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.

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Test Suite Duration'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: 687015a	Previous: 27c9f67	Ratio
test_report_zkpassport_noir-ecdsa_	3 s	1 s	3

This comment was automatically generated by workflow using github-action-benchmark.

CC: @TomAFrench

[jfecher]

LGTM. Some nits

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Execution Time'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: 5e9af29	Previous: dec5ef2	Ratio
rollup-block-root-single-tx	0.003 s	0.002 s	1.50

This comment was automatically generated by workflow using github-action-benchmark.

CC: @TomAFrench