Prune the result of `Package.getAllDependenciesRange`

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.

Remove abandoned registry mirrors. ...

Both of these have not been working for a long time now.

purge changelog

update version to v1.37.0

Merge pull request #2892 from ibuclaw/merge_stable ...

update version to v1.37.0-rc.1

Merge remote-tracking branch 'upstream/stable' into merge_stable

update version to v1.37.0-rc.1

Change all references to std.file in commandline to FQN ...

This makes it more visible where std.file is used, as we want
to reduce the number of places we do I/O, and as we have
two ways to do I/O in Dub.

Trivial: Simplify ensureDirectory ...

mkdirRecurse will not do anything if the directory already exists,
so there is no need to call existsDirectory.

Merge pull request #2888 from ibuclaw/merge_stable ...

merge stable

Merge remote-tracking branch 'upstream/stable' into merge_stable

Let unittest use PackageManager.store ...

No need to mock, all filesystem functions have been mocked
by the PackageManager now.

VFS: Implement file attributes ...

To more accurately reflect the real filesystem,
we now support the access and modification times
for FSEntry.

update version to v1.37.0-beta.1

Merge remote-tracking branch 'upstream/master' into stable

CLI: Simplify control flow by bundling two operations together

Test that `add-path` packages take priority ...

This was not currently tested explicitly, and is currently
the reason why we still have the scanning behavior in the
PackageManager.
In order to remove the scanning behavior and have the PM
lazily scan packages, we first must add test to ensure
there will be no further regression.

Test: Make NativePath import public ...

So that tests can use 'NativePath' and 'toNativeString'.

Add the ability to re-invoke Dub with the same vfs ...

This will prove handy when writting tests that require multiple
invocation of Dub, such as idempotence tests, or things like
add-path which currently don't propagate their change through
Dub as they are one shot commands.

Mock IO in writeLocalPackageList / writeOverrides ...

Mocking IO is required for writeLocalPackageList,
as we want to test the 'add-path' behavior.
While it is not a big concern for overrides, as they
are deprecated and we are unlikely to add new tests,
it is easy to do, and already done for reading them.

Remove the need to mock PackageManager.refresh() ...

We can finally test the full features of the base function,
as we are now mocking the IO, not the logic.

Remove direct IO from Location.scanLocalPackages ...

This is one more step towards removing our mocking of refresh_.

PackageManager: Mock readText, do not mock load ...

This allows us to test the behavior of the base class much better,
by overriding the last IO function, readText.

Move Package.findPackageFile to PackageManager, mock ...

Moving it to PackageManager allows us to call PackageManager.existsFile,
allowing us to mock a lot more things.

Remove the need for intermediate `.zip` file in PackageSupplier ...

By using a different overload of `retryDownload`,
we return the data instead of writing to disk,
ensuring that we don't have a lingering file
in memory and reducing IO.

This follow the same approach for deprecations as previous
`PackageSupplier` deprecations: it's an inevitable, but
easily fixable, breaking change for classes that implement
`PackageSupplier`, but it's a proper deprecation for client
code, based on the assumption that all the classes implementing
`PackageSupplier` are in Dub.

Fix ubyte[] download overload by avoiding a cast ...

We were using the default template parameter, `char`,
and as the documentation for `get` states:
> If asking for char, content will be converted from
> the connection character set (specified in HTTP
> response headers or FTP connection properties,
> both ISO-8859-1 by default) to UTF-8.

However we are downloading a binary so converting
it to UTF-8 will fail. Especially if we intend
to cast it to `ubyte[]` afterwards.

Trivial: Print the Exception message in retryDownload ...

This makes the source of the issue more obvious when using
the verbose flag.

Trivial: Upgrade remaining actions/checkout to v4 ...

Avoid deprecated Node 16.

Test: Directly mock writeFile/existsFile from PackageManager ...

This allows us to reduce the number of functions we override,
as we would like to test writeSelections and refresh directly,
not the copy of them we have.