Remove page id from "page capture".

ChromeRenderViewObserver::CapturePageInfo is a relic. It hasn't captured page content for indexing for about a year now (r212459), and the fancy footwork to avoid re-indexing isn't needed any more.

BUG=371151
TEST=everything still works

Review URL: https://codereview.chromium.org/363293005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@282479 0039d316-1c4b-4281-b951-d872f2087c98

chrome/renderer/chrome_render_view_observer.cc

@@ -100,12 +100,6 @@ static const char kTranslateCaptureText[] = "Translate.CaptureText";
 
 namespace {
 
-GURL StripRef(const GURL& url) {
-  GURL::Replacements replacements;
-  replacements.ClearRef();
-  return url.ReplaceComponents(replacements);
-}
-
 #if defined(OS_ANDROID)
 // Parses the DOM for a <meta> tag with a particular name.
 // |meta_tag_content| is set to the contents of the 'content' attribute.
@@ -165,7 +159,6 @@ ChromeRenderViewObserver::ChromeRenderViewObserver(
       chrome_render_process_observer_(chrome_render_process_observer),
       translate_helper_(new TranslateHelper(render_view)),
       phishing_classifier_(NULL),
-      last_indexed_page_id_(-1),
       capture_timer_(false, false) {
   const CommandLine& command_line = *CommandLine::ForCurrentProcess();
   if (!command_line.HasSwitch(switches::kDisableClientSidePhishingDetection))
@@ -357,7 +350,6 @@ void ChromeRenderViewObserver::DidStopLoading() {
     return;
 
   CapturePageInfoLater(
-      render_view()->GetPageId(),
       false,  // preliminary_capture
       base::TimeDelta::FromMilliseconds(
           render_view()->GetContentStateImmediately() ?
@@ -371,29 +363,21 @@ void ChromeRenderViewObserver::DidCommitProvisionalLoad(
     return;
 
   CapturePageInfoLater(
-      render_view()->GetPageId(),
       true,  // preliminary_capture
       base::TimeDelta::FromMilliseconds(kDelayForForcedCaptureMs));
 }
 
-void ChromeRenderViewObserver::CapturePageInfoLater(int page_id,
-                                                    bool preliminary_capture,
+void ChromeRenderViewObserver::CapturePageInfoLater(bool preliminary_capture,
                                                     base::TimeDelta delay) {
   capture_timer_.Start(
       FROM_HERE,
       delay,
       base::Bind(&ChromeRenderViewObserver::CapturePageInfo,
                  base::Unretained(this),
-                 page_id,
                  preliminary_capture));
 }
 
-void ChromeRenderViewObserver::CapturePageInfo(int page_id,
-                                               bool preliminary_capture) {
-  // If |page_id| is obsolete, we should stop indexing and capturing a page.
-  if (render_view()->GetPageId() != page_id)
-    return;
-
+void ChromeRenderViewObserver::CapturePageInfo(bool preliminary_capture) {
   if (!render_view()->GetWebView())
     return;
 
@@ -427,39 +411,6 @@ void ChromeRenderViewObserver::CapturePageInfo(int page_id,
   if (translate_helper_)
     translate_helper_->PageCaptured(contents);
 
-  // TODO(shess): Is indexing "Full text search" indexing?  In that
-  // case more of this can go.
-  // Skip indexing if this is not a new load.  Note that the case where
-  // page_id == last_indexed_page_id_ is more complicated, since we need to
-  // reindex if the toplevel URL has changed (such as from a redirect), even
-  // though this may not cause the page id to be incremented.
-  if (page_id < last_indexed_page_id_)
-    return;
-
-  bool same_page_id = last_indexed_page_id_ == page_id;
-  if (!preliminary_capture)
-    last_indexed_page_id_ = page_id;
-
-  // Get the URL for this page.
-  GURL url(main_frame->document().url());
-  if (url.is_empty()) {
-    if (!preliminary_capture)
-      last_indexed_url_ = GURL();
-    return;
-  }
-
-  // If the page id is unchanged, check whether the URL (ignoring fragments)
-  // has changed.  If so, we need to reindex.  Otherwise, assume this is a
-  // reload, in-page navigation, or some other load type where we don't want to
-  // reindex.  Note: subframe navigations after onload increment the page id,
-  // so these will trigger a reindex.
-  GURL stripped_url(StripRef(url));
-  if (same_page_id && stripped_url == last_indexed_url_)
-    return;
-
-  if (!preliminary_capture)
-    last_indexed_url_ = stripped_url;
-
   TRACE_EVENT0("renderer", "ChromeRenderViewObserver::CapturePageInfo");
 
 #if defined(FULL_SAFE_BROWSING)

chrome/renderer/chrome_render_view_observer.h

@@ -71,13 +71,12 @@ class ChromeRenderViewObserver : public content::RenderViewObserver {
   void OnSetClientSidePhishingDetection(bool enable_phishing_detection);
   void OnSetWindowFeatures(const blink::WebWindowFeatures& window_features);
 
-  void CapturePageInfoLater(int page_id,
-                            bool preliminary_capture,
+  void CapturePageInfoLater(bool preliminary_capture,
                             base::TimeDelta delay);
 
   // Captures the thumbnail and text contents for indexing for the given load
   // ID.  Kicks off analysis of the captured text.
-  void CapturePageInfo(int page_id, bool preliminary_capture);
+  void CapturePageInfo(bool preliminary_capture);
 
   // Retrieves the text from the given frame contents, the page text up to the
   // maximum amount kMaxIndexChars will be placed into the given buffer.
@@ -99,14 +98,6 @@ class ChromeRenderViewObserver : public content::RenderViewObserver {
   TranslateHelper* translate_helper_;
   safe_browsing::PhishingClassifierDelegate* phishing_classifier_;
 
-  // Page_id from the last page we indexed. This prevents us from indexing the
-  // same page twice in a row.
-  int32 last_indexed_page_id_;
-  // The toplevel URL that was last indexed.  This is used together with the
-  // page id to decide whether to reindex in certain cases like history
-  // replacement.
-  GURL last_indexed_url_;
-
   // A color page overlay when visually de-emaphasized.
   scoped_ptr<WebViewColorOverlay> dimmed_color_overlay_;