1. Add `invalid` value to `SimpleUnaryOperationKind`, `UnaryOperandKind`, `UnaryOperationKind`,
   `SimpleBinaryOperationKind`, `BinaryOperandsKind` and `BinaryOperationKind`.
2. Make BinaryOperationKind/UnaryOperationKind property return Invalid kind instead of throwing
   exception when handling code with errors.

Revert "Merge pull request #8641 from CyrusNajmabadi/addUsingInteractive"

This reverts commit 18d27544, reversing
changes made to 5925aaef.

The change being reverted was hte one to speed up add-using in the interactive
window. However, i'm concerned about potential perf problems this may cause in
the mainline add-using case. I'm reverting this for now until i can come up
with a fix i think is suitable for both.

address more PR feedback.

IOperation API review

Moving the Glyph enum back to the 'Features' dll.

Ensure custom modifiers are copied for types of overriding properties and events even in the case of dynamic/object mismatch.

Improve add-using perf in interactive window.

Fixes #8637

Correctly localize the "Failure" and "Warning" strings

Ensure that VB XML doc comment completion returns correct tag names

Ensure custom modifiers are copied for types of overriding properties and events even in the case of dynamic/object mismatch.

Fixes #7674.

Rename ContinueAsync to RunFromAsync and make it public

Ensure DeclarationTreeBuilder.GetModifiers doesn't throw on unexpected tokens that have errors.

1. When creating symbol tree infos, lazily produce BKtrees.  That way we don't
   take the perf hit for features that want to search but don't need fuzzy matching

2. Don't perform fuzzy matching interactive.  The problem here is that interactive
   Doesn't have a 'serialization service'.  As such, when VS loads and we run
   interactive, we end up having to produce all the same BKtrees for metadata again.
   In normal scenarios we can store these in the .vs file, so we only pay the price
   the very first time.  In Interactive we always pay the price the first time
   after VS loads.

Handle default parameter values for event accessors in CreateBinderForNodeAndUsage.

Prevent NaN from slipping into .Measure calls.

Fixes #8514

Normalize paths when generating pipe names

Harden determinism job

Fix Code Model race for real this time

Misc renames and minor cleanup

Add CompilationOutputFiles test helper.

I thought I'd fixed this subtle race with PR #8607. After all, the check-in tests were passing and it passed locally on my machine.
However, it was still failing consistently on the build server. Finally, I noticed that the only test passes that were failing were
on x86 *Release* (insert sound of forehead being smacked). After compiling Release, I could reproduce the problem locally.

It turns out that this was a race with GC. In a previous fix, when adding an element to FileCodeModel's element cache, we would first
check to see if the element was already in the cache. If it was in the cache, we'd remove it before adding it. However, the test to
see if the element was in the cache called `CleanableWeakComHandleTable.TryGetValue(TKey key, out TValue value)`, which would only return
true if `key` was in the cache and `value` was not set to null. In essence, it's defined like so:

```C#
public bool TryGetValue(TKey key, out TValue value)
{
    this.AssertIsForeground();

    WeakComHandle<TValue, TValue> handle;
    if (_table.TryGetValue(key, out handle))
    {
        value = handle.ComAggregateObject;
        return value != null;
    }

    value = null;
    return false;
}
```

However, when running in Release, the GC can run a bit more aggresively and `handle.ComAggregatedObject` (a weak reference) could easily be null.

The fix is to add a simpler `CleanableWeakComHandleTable.ContainsKey(TKey key)` and use that instead.

```C#
public bool ContainsKey(TKey key)
{
    this.AssertIsForeground();

    return _table.ContainsKey(key);
}
```

Enable Glyph's on code actions.