For character predicates use String.fromCharCode instead of extract…

…ing the matched range from the source; it brings a small performance benefit.

lib/src/parser/predicate/single_character.dart

@@ -18,9 +18,11 @@ class SingleCharacterParser extends CharacterParser {
   Result<String> parseOn(Context context) {
     final buffer = context.buffer;
     final position = context.position;
-    if (position < buffer.length &&
-        predicate.test(buffer.codeUnitAt(position))) {
-      return context.success(buffer[position], position + 1);
+    if (position < buffer.length) {
+      final codeUnit = buffer.codeUnitAt(position);
+      if (predicate.test(codeUnit)) {
+        return context.success(String.fromCharCode(codeUnit), position + 1);
+      }
     }
     return context.failure(message);
   }

lib/src/parser/predicate/unicode_character.dart

@@ -28,8 +28,7 @@ class UnicodeCharacterParser extends CharacterParser {
         }
       }
       if (predicate.test(codeUnit)) {
-        return context.success(
-            buffer.substring(position, nextPosition), nextPosition);
+        return context.success(String.fromCharCode(codeUnit), nextPosition);
       }
     }
     return context.failure(message);