Fix parameter naming convention.

Avoid visiting packages more than once in checkPacksRec.

Avoid linear time edge list construction. ...

Uses a set to determine redundant edge entries instead of iterating through the list.

Reduce allocation overhead in determineDependencyConfigs.

Avoid linear lookup and name allocations in Project.getDependency.

Avoid linear time haveConfig.

Use package indices instead of names or Package instances to identify packages. ...

Avoids costly AA lookups and string comparisons.

Avoid re-allocating allconfigs_path in determineAllConfigs.

Avoid re-allocating the AA for isReachableByAllParentPacks for every call.

Eliminate edges in the configuration graph in-place. ...

Avoids reallocating the edges array every time the configuration resolution eliminates a configuration.

Cache various package information to reduce GC allocations in getPackageConfigs. ...

Package.name and Package.getAllDependencies both need to reallocate their return value for every call, which results in heavy GC pressure during the configuration resolution algorithm.

Return early from determineDependencyConfigs if possible ...

The only package that can have a config already created is the root
package, and only if the `config` parameter isn't an empty string.
Every other package is in the `configs` array iff it was added by
a previous call to this function. The exact same configurations
will be found if this function is called again for the same (p, c)
pair and it will do needless work that won't modify anything.

Move writing selections file to PackageManager ...

The PackageManager is the main source of IO in Dub,
and having it take responsibility for writing the
selections file reduces the number of special casing
we need to make when doing dependency injection.

Remove an unused JSON serialization call ...

This was completely unused in the function.

Use the proper constructor for PackageDependency ...

This code was using a deprecated constructor, but as it was in lambda,
it didn't trigger a deprecation message, however '-de' points us to
the issue.

Make 'Dependency.init' invalid and align Any/Invalid property ...

VersionRange has Invalid and Any while Dependency has invalid and any.
One has to go, and it's going to be Dependency's.
In addition, make sure that the 'init' state of Dependency is a known state.

Adapt PackageManager.get[Best]Package to PackageName API ...

This paves the way to be able to automatically return subpackages.

Adapt loadSCMPackage to PackageName API

Remove PackageName alias this to toString ...

The alias this was originally added to make the transition
less painful, however due to an overload resolution bug,
it makes it harder at the moment, as thje compiler prefers
calling the deprecated `string` overloads over the
`in PackageName` ones. So remove it and a few hacks along
the way.

Adapt SelectedVersions to PackageName API ...

Adapting to the PackageName API meant relaxing the input
a SelectedVersion method can accept, as we now handle
the subpackage part gracefully (ignoring it), instead
of potentially running into issue.

Expand selections file handling to handle new versions ...

Up until now, the selections file has had a `fileVersion` in it,
but that number was always stuck to 1.
In order to make it possible to handle more than one version,
a bit of scaffolding needs to happen in the config parser.
This lays the ground for a version 2 of the selections file.

Introduce PackageName, to fully type a package name ...

Package names being passed around in the Dub codebase are string,
which is problematic as it suffers from validation / semantic issues.
Validation because we cannot impose restrictions on characters
or length easily, and semantic because some functions expect a main
package while some would accept a subpackage. Introducing a type
will remedy both those issues.

Allow building deprecated modules without deprecation warnings ...

The compiler issues a deprecation warning only on a
"non-deprecated" -> "deprecated" transition.  As such, building or
testing Dub packages which have deprecated modules when a
dub_test_root.d file needs to be generated will cause deprecation
warnings.

This in turn makes it impossible to use `buildRequirements
"disallowDeprecations"` in this situation.

Fix this by simply making the auto-generated root
module (dub_test_root) deprecated. As it is the compiler's entry
point, the module itself will not cause a deprecation warning, and it
importing deprecating modules will not cause deprecation warnings (as
there is no longer a transition of deprecation state).

Deprecate Project.this(PackageManager, NativePath) ...

This overload was originally only used from Dub.loadPackage(NativePath)
before it was changed (although the Project should only ever be instantiated
from the Dub class). But it is doing IO and its soft failure approach
is a misfeature, hence we deprecated it.

Trivial: Simplify Project.reinit by removing nested functions ...

When 95% of your function is nested, you're doing it wrong.
Move the function to be at the class level and private,
which makes the logic easier to follow and reduces indentation.

Trivial: Document Project.reinit ...

Arguably one of the most important method on the package management
front, as it is handling the logic of what dependencies will be used
in most cases (the DependencyResolver is only called when deps are
missing, but most projects have a dub.selections.json).

Deprecate Project.packageManager ...

There is no reason to expose the PackageManager through the Project.
In fact, the only two places it was used were in generators,
where the PackageManager was stored but never used.
The generators arguably should not have access to the PackageManager,
as all the Packages should have been loaded and handled to them
before the generation starts.

Fix #2696: Segfault with bad dub.sdl ...

This was a null pointer being passed down.
However, while looking at the code, it was obvious that similar
bugs would trigger for subpackages (as the code was not handling
`null` packages either), so test cases for those have been added.

Add unittest framework and some dependency tests ...

Dub has always been notoriously hard to test due to the amount of IO it does.
Over the past few years, I have progressively refactored it to allow
dependency injection to take place, which this finally put into action.
By overriding the `PackageManager`, `PackageSupplier`, and a few strategic
functions, we can start to unittest Dub's behavior solely in unittests.
This should hopefully tremendously helps with adding regression tests
for the package manager side of Dub (the build side still need work
to have similar capabilities).

Allow to load dub.selections.json before Project instantiation ...

The aim of this commit is to have at least one constructor
for Project that does not perform IO. By moving the IO to the
`Dub` class, we can keep the dependency injection in `Dub`
and `PackageManager` only, and don't have to specialize `Project`.