LibDiff: Add new API to generate hunks from two pieces of text

For now this is just a standard implementation of the longest
common subsequence algorithm over the lines, except that it doesn't
do any coalescing of the lines. This isn't really ideal since
we get a single Hunk per changed line, and is definitely something
to improve in the future.

Userland/Libraries/LibDiff/Generator.h

@@ -0,0 +1,15 @@
+/*
+ * Copyright (c) 2021, Mustafa Quraish <mustafa@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#pragma once
+
+#include "Hunks.h"
+
+namespace Diff {
+
+Vector<Hunk> from_text(StringView const& old_text, StringView const& new_text);
+
+}