Clarify ElmTreeWalker pattern matching logic

Added comments to explain the behavior and intent of the
pattern matching logic in the `ElmTreeWalker` creation
within the `MissingResultTypeSpecifierCorrector` class.
The comments highlight the importance of processing order
and why a switch expression is not suitable. These changes
improve code readability and help prevent unintended
modifications that could disrupt functionality.

Author: baseTwo

Cql/Cql.Compiler/Preprocessing/MissingResultTypeSpecifierCorrector.cs

@@ -19,6 +19,11 @@ internal class MissingResultTypeSpecifierCorrector(ILogger<MissingResultTypeSpec
 {
     private readonly ElmTreeWalker _walker = ElmTreeWalker.Create((self, node) =>
     {
+        // DO NOT turn this into a switch expression.
+        // It is possible to match multiple patterns in one node, e.g.
+        // Element with resultTypeName then Query with ListTypeSpecifier
+        // The order of these blocks matters too, starting with the most general ones first.
+
         {
             // If an Element has a resultTypeName but no resultTypeSpecifier, set resultTypeSpecifier to a NamedTypeSpecifier with the name from resultTypeName
             if (node is Element { resultTypeSpecifier: null, resultTypeName: { } resultTypeName } element)