Tweaking WebPageSerializerImpl to emit a BOM for UTF16/32.

chrome/browser/download/save_page_browsertest.cc

 // Copyright (c) 2012 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "base/bind.h"
 #include "base/bind_helpers.h"
 #include "base/command_line.h"
 #include "base/files/file_path.h"
 #include "base/files/file_util.h"
 #include "base/files/scoped_temp_dir.h"
@@ skipping 955 matching lines @@
     EXPECT_EQ(expected_number_of_frames, actual_number_of_frames);
 
     for (const auto& expected_substring : expected_substrings) {
       int actual_number_of_matches = ui_test_utils::FindInPage(
           GetCurrentTab(browser()), base::UTF8ToUTF16(expected_substring),
           true,  // |forward|
           true,  // |case_sensitive|
           nullptr, nullptr);
 
       EXPECT_EQ(1, actual_number_of_matches)
-          << "Verifying if \"" << expected_substring << "\" appears "
-          << "exactly once in the web-contents text";
+          << "Verifying that \"" << expected_substring << "\" appears "
+          << "exactly once in the text of web contents";
     }
 
-    int actual_number_of_errors = ui_test_utils::FindInPage(
-        GetCurrentTab(browser()), base::UTF8ToUTF16("err"),
-        true,   // |forward|
-        false,  // |case_sensitive|
-        nullptr, nullptr);
-    EXPECT_EQ(0, actual_number_of_errors);
+    std::vector<std::string> forbidden_substrings{
+        "head"  // Html markup should not render.
+        "err",  // Used in error messages + included as text content
+                // of <object> elements (in case the object itself doesn't
+                // render).
+    };
+
+    for (const auto& forbidden_substring : forbidden_substrings) {
+      int actual_number_of_matches = ui_test_utils::FindInPage(
+          GetCurrentTab(browser()), base::UTF8ToUTF16(forbidden_substring),
+          true,  // |forward|
+          true,  // |case_sensitive|
+          nullptr, nullptr);
+
+      EXPECT_EQ(0, actual_number_of_matches)
+          << "Verifying that \"" << forbidden_substring << "\" doesn't "
+          << "appear in the text of web contents";
+    }
   }
 
   static void IncrementInteger(int* i, content::RenderFrameHost* /* unused */) {
     (*i)++;
   }
 };
 
 // Test coverage for OOPIFs for CompleteHtml (crbug.com/526786) and
 // MHTML (crbug.com/538766) as well as for redirected iframes saved
 // as MHTML (crbug.com/539936).
@@ skipping 73 matching lines @@
       "frames-nested2.htm: 6d23dc47-f283-4977-96ec-66bcf72301a4",
       "b.htm: 3a35f7fa-96a9-4487-9f18-4470263907fa",
   };
 
   GURL url(
       embedded_test_server()->GetURL("a.com", "/save_page/frames-nested.htm"));
 
   TestMultiFramePage(save_page_type, url, 3, expected_substrings);
 }
 
+// Test for saving frames with various encodings (crbug.com/541699).
+// - iso-8859-2: single-byte encoding, encoding declared via <meta> element
+// - utf32.htm - multi-byte encoding, detection via mocked http headers
+// - euc-kr.htm - encoding via html entities
+IN_PROC_BROWSER_TEST_P(SavePageMultiFrameBrowserTest, Encoding) {
+  content::SavePageType save_page_type = GetParam();
+
+  std::vector<std::string> expected_substrings{
+      "frames-encodings.htm: f53295dd-a95b-4b32-85f5-b6e15377fb20",
+      "euc-kr.htm: 樂樂",
+      "iso-8859-2.htm: Zażółć gęślą jaźń",
+      "utf32.htm: Zażółć gęślą jaźń",
+  };
+
+  GURL url(embedded_test_server()->GetURL("a.com",
+                                          "/save_page/frames-encodings.htm"));
+
+  // TODO(lukasza): crbug.com/541699: MHTML needs to handle multi-byte encodings
+  // by either:
+  // 1. Continuing to preserve the original encoding, but starting to round-trip
+  //    the encoding declaration (in Content-Type MIME/MHTML header?)
+  // 2. Saving html docs in UTF8.
+  if (save_page_type == content::SAVE_PAGE_TYPE_AS_MHTML)
+    return;
+
+  TestMultiFramePage(save_page_type, url, 4, expected_substrings);
+}
+
 INSTANTIATE_TEST_CASE_P(
     ,
     SavePageMultiFrameBrowserTest,
     ::testing::Values(content::SAVE_PAGE_TYPE_AS_COMPLETE_HTML,
                       content::SAVE_PAGE_TYPE_AS_MHTML));
 
 }  // namespace