/// <summary>
/// Returns the full path for find results, based on the initially provided path.
/// </summary>
public string ToSpecifiedFullPath()
{
// We want to provide the enumerated segment of the path appended to the originally specified path. This is
// the behavior of the various Directory APIs that return a list of strings.
//
// RootDirectory has the final separator trimmed, OriginalRootDirectory does not. Our legacy behavior would
// effectively account for this by appending subdirectory names as it recursed. As such we need to trim one
// separator when combining with the relative path (Directory.Slice(RootDirectory.Length)).
//
// Original => Root => Directory => FileName => relativePath => Specified
// C:\foo C:\foo C:\foo bar "" C:\foo\bar
// C:\foo\ C:\foo C:\foo bar "" C:\foo\bar
// C:\foo/ C:\foo C:\foo bar "" C:\foo/bar
// C:\foo\\ C:\foo C:\foo bar "" C:\foo\\bar
// C:\foo C:\foo C:\foo\bar jar "bar" C:\foo\bar\jar
// C:\foo\ C:\foo C:\foo\bar jar "bar" C:\foo\bar\jar
// C:\foo/ C:\foo C:\foo\bar jar "bar" C:\foo/bar\jar
// If we're at the top level directory the Directory and RootDirectory will be identical. As there are no
// trailing slashes in play, once we're in a subdirectory, slicing off the root will leave us with an
// initial separator. We need to trim that off if it exists, but it isn't needed if the original root
// didn't have a separator. Join() would handle it if we did trim it, not doing so is an optimization.
ReadOnlySpan <char> relativePath = Directory.Slice(RootDirectory.Length);
if (Path.EndsInDirectorySeparator(OriginalRootDirectory) && PathInternal.StartsWithDirectorySeparator(relativePath))
{
relativePath = relativePath.Slice(1);
}
return(Path.Join(OriginalRootDirectory, relativePath, FileName));
}