step 1 in making unittest platform independent.

lib/unittest/config.dart

Index: lib/unittest/config.dart
diff --git a/lib/unittest/config.dart b/lib/unittest/config.dart
new file mode 100644
index 0000000000000000000000000000000000000000..9507b508f92f8ad9dcb93bce14a37ceae4e88baa
--- /dev/null
+++ b/lib/unittest/config.dart
@@ -0,0 +1,65 @@
+// Copyright (c) 2011, 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.
+
+/**
+ * Hooks to configure the unittest library for different platforms. This class
+ * implements the API in a platform-independent way. Tests that want to take
+ * advantage of the platform can create a subclass and override methods from
+ * this class.
+ */
+class Configuration {
+  /**
+   * Called as soon as the unittest framework becomes initialized. This is done
+   * even before tests are added to the test framework. It might be used to
+   * determine/debug errors that occur before the test harness starts executing.
+   */
+  void onInit() {}
+
+  /**
+   * Called as soon as the unittest framework starts running. Used commonly to
+   * tell the vm or browser that tests are still running and the process should
+   * wait until they are done.
+   */
+  void onStart() {}
+
+  /**
+   * Called when each test is completed. Useful to show intermediate progress on
+   * a test suite.
+   */
+  void onTestResult(TestCase testCase) {}
+
+  /**
+   * Called with the result of all test cases. The default implementation prints
+   * the result summary using the built-in [print] command. Browser tests
+   * commonly override this to reformat the output.
+   */
+  void onDone(int passed, int failed, int errors, List<TestCase> results) {
+    // Print each test's result.
+    for (final test in _tests) {
+      print('${test.result.toUpperCase()}: ${test.description}');
+
+      if (test.message != '') {
+        print('  ${test.message}');
+      }
+    }
+
+    // Show the summary.
+    print('');
+
+    var success = false;
+    if (passed == 0 && failed == 0 && errors == 0) {
+      print('No tests found.');
+      // This is considered a failure too: if this happens you probably have a
+      // bug in your unit tests.
+    } else if (failed == 0 && errors == 0) {
+      print('All $passed tests passed.');
+      success = true;
+    } else {
+      print('$passed PASSED, $failed FAILED, $errors ERRORS');
+    }
+
+    // An exception is used by the test infrastructure to detect failure.
+    if (!success) throw new Exception("Some tests failed.");
+  }
+}