Port back CLD to Linux and Mac and fix Trad Chinese detection

chrome/renderer/render_view.cc

 // Copyright (c) 2010 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 "chrome/renderer/render_view.h"
 
 #include <algorithm>
 #include <string>
 #include <vector>
 
@@ skipping 49 matching lines @@
 #include "chrome/renderer/webplugin_delegate_proxy.h"
 #include "chrome/renderer/websharedworker_proxy.h"
 #include "chrome/renderer/webworker_proxy.h"
 #include "grit/generated_resources.h"
 #include "grit/renderer_resources.h"
 #include "net/base/data_url.h"
 #include "net/base/escape.h"
 #include "net/base/net_errors.h"
 #include "skia/ext/bitmap_platform_device.h"
 #include "skia/ext/image_operations.h"
-#if defined(OS_WIN)
-// TODO(port): The compact language detection library works only for Windows.
 #include "third_party/cld/bar/toolbar/cld/i18n/encodings/compact_lang_det/win/cld_unicodetext.h"
-#endif
 #include "third_party/WebKit/WebKit/chromium/public/WebAccessibilityCache.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebAccessibilityObject.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebCString.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebDataSource.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebDevToolsAgent.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebDocument.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebDragData.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebFileChooserParams.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebFormElement.h"
 #include "third_party/WebKit/WebKit/chromium/public/WebFrame.h"
@@ skipping 132 matching lines @@
 
 // The maximum number of popups that can be spawned from one page.
 static const int kMaximumNumberOfUnacknowledgedPopups = 25;
 
 static const char* const kUnreachableWebDataURL =
     "chrome://chromewebdata/";
 
 static const char* const kBackForwardNavigationScheme = "history";
 
 // The string returned in DetectLanguage if we failed to detect the language.
-static const char* const kUnknownLanguageCode = "unknown";
+static const char* const kUnknownLanguageCode = "und";
 
 static void GetRedirectChain(WebDataSource* ds, std::vector<GURL>* result) {
   WebVector<WebURL> urls;
   ds->redirectChain(urls);
   result->reserve(urls.size());
   for (size_t i = 0; i < urls.size(); ++i)
     result->push_back(urls[i]);
 }
 
 static bool UrlMatchesPermissions(
@@ skipping 2838 matching lines @@
     return;
   }
 
   Send(new ViewHostMsg_PageLanguageDetermined(routing_id_, DetectLanguage()));
 }
 
 std::string RenderView::DetectLanguage() {
   if (!webview() || is_loading_)
     return kUnknownLanguageCode;
 
-  std::string language = kUnknownLanguageCode;
-#if defined(OS_WIN)  // CLD is only available on Windows at this time.
   WebFrame* main_frame = webview()->mainFrame();
   std::wstring contents;
   CaptureText(main_frame, &contents);
-  language = DetermineTextLanguage(contents);
-#endif
-
-  return language;
+  return DetermineTextLanguage(contents);
 }
 
 // static
 std::string RenderView::DetermineTextLanguage(const std::wstring& text) {
   std::string language = kUnknownLanguageCode;
-#if defined(OS_WIN)  // CLD is only available on Windows at this time.
   int num_languages = 0;
   bool is_reliable = false;
+  string16 input = WideToUTF16(text);
   Language cld_language =
-      DetectLanguageOfUnicodeText(NULL, text.c_str(), true, &is_reliable,
+      DetectLanguageOfUnicodeText(NULL, input.c_str(), true, &is_reliable,
                                   &num_languages, NULL);
   if (cld_language != NUM_LANGUAGES && cld_language != UNKNOWN_LANGUAGE &&
       cld_language != TG_UNKNOWN_LANGUAGE) {
-    language = LanguageCodeISO639_1(cld_language);
+    // We should not LanguageCode_ISO_639_1 because it does not cover all the
+    // languages CLD can detect. As a result, it'll return the invalid language
+    // code for tradtional Chinese among others. |LanguageCode| will go through
+    // ISO 639-1, ISO-639-2 and 'other' tables to do the 'right' thing.
+    language = LanguageCode(cld_language);
   }
-#endif
   return language;
 }
 
 void RenderView::DnsPrefetch(const std::vector<std::string>& host_names) {
   Send(new ViewHostMsg_DnsPrefetch(host_names));
 }
 
 void RenderView::OnZoom(PageZoom::Function function) {
   if (!webview())  // Not sure if this can happen, but no harm in being safe.
     return;
@@ skipping 929 matching lines @@
       new PluginMsg_SignalModalDialogEvent(host_window_));
 
   message->EnableMessagePumping();  // Runs a nested message loop.
   bool rv = Send(message);
 
   PluginChannelHost::Broadcast(
       new PluginMsg_ResetModalDialogEvent(host_window_));
 
   return rv;
 }