Previously, all imports were available in the header, so we could start
processing dependencies as soon as we parsed it. However, the new imports
are treated as defs, so we have to parse the whole module to find them.
This commit essentially moves the dependency resolution from the `LoadHeader`
phase to the `Parse` phase, and it updates canonicalization to introduce
module symbols into scope when a `ValueDef::ModuleImport` is encountered.
NOTE:
- The `imports` header still parses, but it's no longer wired up. I will remove
it in an upcoming commit.
- Ingested files and imports that appear in nested expressions are not
yet supported by load
This new flag determines whether we should introduce a new kind to
represent lambda sets, or whether lambdas should be erased. The latter
is not yet implemented.
Prior to this commit, if you had a module structure like
```
| - A.roc
| - Dep
| - B.roc
```
where `B.roc` was defined as
```
interface B exposes [] imports []
```
and `A.roc` was defined as
```
interface A exposes [] imports [Dep.B]
```
The compiler would hang on you. The reason is that even though we expect
`B` to be named `Dep.B` relative to `A`, that would not be enforced.
With this patch, we now enforce such naming schemes - a module must have
the namespaced name it is referenced by. Currently, we determine the
expected namespaced name by looking at how transitive dependencies of the
root module reference the module. In the future, once we have a package
ecosystem and a solid idea of "package roots", we can use the "package
root" to determine how a module should be named.
Closes#4094
This fixed a bug where bindgen was providing cwd() for src_dir,
but actually the src_dir should have been based on the filename.
This prevents that problem from happening in the future!
