webauthn: add language and direction tags to UI example.

https://github.com/w3c/webauthn/pull/1530 specifies a method for
encoding language and direction information in strings that will be
stored on authenticators. Since we may start to see this in credentials
in the wild, this change adds examples to the UI test so that we see
what they'll look like.

(On Linux, at least, they at least don't render as garbage or crash.
However the zh-Hant example doesn't display, possibly because I'm
missing suitable fonts.)

Change-Id: I9e778a08df85b8ae34032341c6bcd84291ff210c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2576795
Commit-Queue: Martin Kreichgauer <martinkr@google.com>
Auto-Submit: Adam Langley <agl@chromium.org>
Reviewed-by: Martin Kreichgauer <martinkr@google.com>
Cr-Commit-Position: refs/heads/master@{#834370}

chrome/browser/ui/webauthn/authenticator_dialog_browsertest.cc

@@ -144,11 +144,42 @@ class AuthenticatorDialogTest : public DialogBrowserTest {
       model->SetCurrentStep(
           AuthenticatorRequestDialogModel::Step::kStorageFull);
     } else if (name == "account_select") {
+      // These strings attempt to exercise the encoding of direction and
+      // language from https://github.com/w3c/webauthn/pull/1530.
+
+      // lang_and_dir_encoded contains a string with right-to-left and ar-SA
+      // tags. It's the UTF-8 encoding of the code points {0xE0001, 0xE0061,
+      // 0xE0072, 0xE002D, 0xE0053, 0xE0041, 0x200F, 0xFEA2, 0xFE92, 0xFBFF,
+      // 0xFE91, 0x20, 0xFE8E, 0xFEDF, 0xFEAE, 0xFEA4, 0xFEE3, 0xFE8E, 0xFEE7}.
+      const std::string lang_and_dir_encoded =
+          "\xf3\xa0\x80\x81\xf3\xa0\x81\xa1\xf3\xa0\x81\xb2\xf3\xa0\x80\xad\xf3"
+          "\xa0\x81\x93\xf3\xa0\x81\x81\xe2\x80\x8f\xef\xba\xa2\xef\xba\x92\xef"
+          "\xaf\xbf\xef\xba\x91\x20\xef\xba\x8e\xef\xbb\x9f\xef\xba\xae\xef\xba"
+          "\xa4\xef\xbb\xa3\xef\xba\x8e\xef\xbb\xa7";
+      // lang_jp_encoded specifies a kanji with language jp. This is the middle
+      // glyph from the example given in
+      // https://www.w3.org/TR/string-meta/#capturing-the-text-processing-language.
+      // It's the UTF-8 encoding of the code points {0xE0001, 0xE006a, 0xE0070,
+      // 0x76f4}.
+      const std::string lang_jp_encoded =
+          "\xf3\xa0\x80\x81\xf3\xa0\x81\xaa\xf3\xa0\x81\xb0\xe7\x9b\xb4";
+      // lang_zh_hant_encoded specifies the same code point as
+      // |lang_jp_encoded|, but with the language set to zh-Hant. According to
+      // the W3C document referenced above, this should display differently.
+      // It's the UTF-8 encoding of the code points {0xE0001, 0xe007a, 0xe0068,
+      // 0xe002d, 0xe0048, 0xe0061, 0xe006e, 0xe0074}.
+      const std::string lang_zh_hant_encoded =
+          "\xf3\xa0\x80\x81\xf3\xa0\x81\xba\xf3\xa0\x81\xa8\xf3\xa0\x80\xad\xf3"
+          "\xa0\x81\x88\xf3\xa0\x81\xa1\xf3\xa0\x81\xae\xf3\xa0\x81\xb4";
+
       const std::vector<std::pair<std::string, std::string>> infos = {
           {"foo@example.com", "Test User 1"},
           {"", "Test User 2"},
           {"", ""},
           {"bat@example.com", "Test User 4"},
+          {"encoded@example.com", lang_and_dir_encoded},
+          {"encoded2@example.com", lang_jp_encoded},
+          {"encoded3@example.com", lang_zh_hant_encoded},
           {"verylong@"
            "reallylongreallylongreallylongreallylongreallylongreallylong.com",
            "Very Long String Very Long String Very Long String Very Long "