Add a Pub validator for READMEs that are invalid utf-8.

utils/pub/validator/utf8_readme.dart

Index: utils/pub/validator/utf8_readme.dart
diff --git a/utils/pub/validator/utf8_readme.dart b/utils/pub/validator/utf8_readme.dart
new file mode 100644
index 0000000000000000000000000000000000000000..b660291b2b9c5dd9532f4083de994ec5ad4fa923
--- /dev/null
+++ b/utils/pub/validator/utf8_readme.dart
@@ -0,0 +1,41 @@
+// Copyright (c) 2013, the Dart project authors.  Please see the AUTHORS file
+// for details. All rights reserved. Use of this source code is governed by a
+// BSD-style license that can be found in the LICENSE file.
+
+library utf8_readme_validator;
+
+import 'dart:async';
+import 'dart:utf';
+
+import '../../../pkg/path/lib/path.dart' as path;
+
+import '../entrypoint.dart';
+import '../io.dart';
+import '../utils.dart';
+import '../validator.dart';
+
+/// A validator that validates that a package doesn't contain compiled Dartdoc
+/// output.

[Bob Nystrom, 2013/01/31 19:08:29]

Fix doc comment.

[nweiz, 2013/01/31 21:39:10]

Done.

+class Utf8ReadmeValidator extends Validator {
+  Utf8ReadmeValidator(Entrypoint entrypoint)
+    : super(entrypoint);
+
+  Future validate() {
+    return entrypoint.readmePath.then((readme) {
+      if (readme == null) return;
+      return readByteFile(readme).then((bytes) {
+        try {
+          // The second and third arguments here are the default values. The
+          // fourth tells [decodeUtf8] to throw an ArgumentError if `bytes`
+          // isn't valid utf-8.

[Bob Nystrom, 2013/01/31 19:08:29]

Good God, that's a bad API.

[nweiz, 2013/01/31 21:39:10]

It would be a little better if it used keyword args...

+          decodeUtf8(bytes, 0, null, null);
+        } on ArgumentError catch (_) {
+          warnings.add("$readme contains invalid UTF-8.\n"
+              "This will cause it to be displayed incorrectly on "
+                  "pub.dartlang.org.");
+        }
+      });
+    });
+  }
+}
+