Implementing URL prefix match for history thumbnail cache.

chrome/browser/history/top_sites_cache.h

Index: chrome/browser/history/top_sites_cache.h
diff --git a/chrome/browser/history/top_sites_cache.h b/chrome/browser/history/top_sites_cache.h
index 822d18903dc00e05ceb538e3975f71b2a6a8e3cf..d8874879108efce328216c0ff897473e94ea0e2e 100644
--- a/chrome/browser/history/top_sites_cache.h
+++ b/chrome/browser/history/top_sites_cache.h
@@ -10,9 +10,31 @@
 
 #include "base/memory/ref_counted.h"
 #include "chrome/browser/history/history_types.h"
+#include "chrome/browser/history/url_utils.h"
 
 namespace history {
 
+// TopSiteCache caches thumbnails for visited pages. Retrieving thumbnails from
+// a given [input URL] is a two-stage process:
+//
+//   [input URL] --(map 1)--> [canonical URL] --(map 2)--> image.
+//
+// (map 1) searches [input URL] in |canonical_urls_|. [canonical URL] is
+// assigned to the resulting value if found; else [input URL].
+//
+// (map 2) simply looks up [canonical URL] in |images_|.
+//
+// TopSiteCache also provides GetCanonicalURLsIteratorForPrefix(), which is an
+// alternative implementation of (map 1) that does the following:
+// - if [canonical URL] is a key in |canonical_urls_|, return the value.
+// - else if [canonical URL] is a "URL prefix" (see comment in url_utils.h) of
+//   some key in |canonical_urls_|, return the value corresponding to the key.
+// - else return [input URL].
+
+// The entries in CanonicalURLs, see CanonicalURLs for details. The second
+// argument gives the index of the URL into MostVisitedURLs redirects.
+typedef std::pair<MostVisitedURL*, size_t> CanonicalURLEntry;

[sky, 2013/09/12 21:22:18]

Why is this public?

[huangs, 2013/09/12 23:33:47]

A relic from the previous change. Moved back into TopSitesCache as private.

+
 // TopSitesCache caches the top sites and thumbnails for TopSites.
 class TopSitesCache {
  public:
@@ -41,9 +63,14 @@ class TopSitesCache {
   // there is a thumbnail score for the specified url.
   bool GetPageThumbnailScore(const GURL& url, ThumbnailScore* score);
 
-  // Returns the canonical URL for |url|.
+  // Returns the [canonical URL] for |url|.
   const GURL& GetCanonicalURL(const GURL& url);
 
+  // Returns the [canonical URL] for |url_prefix| that matches by prefix.

[sky, 2013/09/12 21:22:18]

What's with [all the brackets] ?

[huangs, 2013/09/12 23:33:47]

The purpose is to highlight proper terms -- but I suppose that's not the style I
should use. Removed.

+  // Multiple matches exst, returns the [canonical URL] for the first
+  // matching entry under lexicographical order.
+  const GURL& GetCanonicalURLForPrefix(const GURL& url_prefix);
+
   // Returns true if |url| is known.
   bool IsKnownURL(const GURL& url);
 
@@ -51,17 +78,11 @@ class TopSitesCache {
   size_t GetURLIndex(const GURL& url);
 
  private:
-  // The entries in CanonicalURLs, see CanonicalURLs for details. The second
-  // argument gives the index of the URL into MostVisitedURLs redirects.
-  typedef std::pair<MostVisitedURL*, size_t> CanonicalURLEntry;
-
-  // Comparator used for CanonicalURLs.
+  // Comparator used for CanonicalURLs. Its main purpose
   class CanonicalURLComparator {
    public:
     bool operator()(const CanonicalURLEntry& e1,
-                    const CanonicalURLEntry& e2) const {
-      return e1.first->redirects[e1.second] < e2.first->redirects[e2.second];

[sky, 2013/09/12 21:22:18]

Why nuke the inlining here?

[huangs, 2013/09/12 23:33:47]

Need to call CanonicalURLStringCompare(), and I wanted to avoid including
url_utils.h (even though I did, which was a mistake). Made inline again, and
keeping #include ".../url_utils.h".

-    }
+                    const CanonicalURLEntry& e2) const;
   };
 
   // This is used to map from redirect url to the MostVisitedURL the redirect is
@@ -80,9 +101,14 @@ class TopSitesCache {
   // Stores a set of redirects. This is used by GenerateCanonicalURLs.
   void StoreRedirectChain(const RedirectList& redirects, size_t destination);
 
-  // Returns the iterator into canconical_urls_ for the specified url.
+  // Returns the iterator into canonical_urls_ for the |url|.

[sky, 2013/09/12 21:22:18]

Style here is || around references to parameters and members.

[huangs, 2013/09/12 23:33:47]

Done.

   CanonicalURLs::iterator GetCanonicalURLsIterator(const GURL& url);
 
+  // Returns the first iterator into canonical_urls_ for whcih |previx_url|

[sky, 2013/09/12 21:22:18]

|canonical_urls_|.
previx->prefix

[huangs, 2013/09/12 23:33:47]

Done, also using the term "URL prefix" for consistency.

+  // is a path-prefix.
+  CanonicalURLs::iterator GetCanonicalURLsIteratorForPrefix(
+      const GURL& prefix_url);
+
   // The top sites.
   MostVisitedURLList top_sites_;