OLD | NEW |
1 // Copyright (c) 2012 The Chromium Authors. All rights reserved. | 1 // Copyright (c) 2012 The Chromium Authors. All rights reserved. |
2 // Use of this source code is governed by a BSD-style license that can be | 2 // Use of this source code is governed by a BSD-style license that can be |
3 // found in the LICENSE file. | 3 // found in the LICENSE file. |
4 | 4 |
5 #ifndef COMPONENTS_OMNIBOX_BROWSER_SCORED_HISTORY_MATCH_H_ | 5 #ifndef COMPONENTS_OMNIBOX_BROWSER_SCORED_HISTORY_MATCH_H_ |
6 #define COMPONENTS_OMNIBOX_BROWSER_SCORED_HISTORY_MATCH_H_ | 6 #define COMPONENTS_OMNIBOX_BROWSER_SCORED_HISTORY_MATCH_H_ |
7 | 7 |
8 #include <stddef.h> | 8 #include <stddef.h> |
9 | 9 |
10 #include <string> | 10 #include <string> |
11 #include <vector> | 11 #include <vector> |
12 | 12 |
13 #include "base/gtest_prod_util.h" | 13 #include "base/gtest_prod_util.h" |
14 #include "base/strings/string16.h" | 14 #include "base/strings/string16.h" |
15 #include "base/time/time.h" | 15 #include "base/time/time.h" |
16 #include "components/history/core/browser/history_match.h" | 16 #include "components/history/core/browser/history_match.h" |
17 #include "components/history/core/browser/history_types.h" | 17 #include "components/history/core/browser/history_types.h" |
18 #include "components/omnibox/browser/in_memory_url_index_types.h" | 18 #include "components/omnibox/browser/in_memory_url_index_types.h" |
| 19 #include "components/omnibox/browser/omnibox_field_trial.h" |
19 | 20 |
20 class ScoredHistoryMatchTest; | 21 class ScoredHistoryMatchTest; |
21 | 22 |
22 // An HistoryMatch that has a score as well as metrics defining where in the | 23 // An HistoryMatch that has a score as well as metrics defining where in the |
23 // history item's URL and/or page title matches have occurred. | 24 // history item's URL and/or page title matches have occurred. |
24 struct ScoredHistoryMatch : public history::HistoryMatch { | 25 struct ScoredHistoryMatch : public history::HistoryMatch { |
25 // ScoreMaxRelevance maps from an intermediate-score to the maximum | 26 // ScoreMaxRelevance maps from an intermediate-score to the maximum |
26 // final-relevance score given to a URL for this intermediate score. | 27 // final-relevance score given to a URL for this intermediate score. |
27 // This is used to store the score ranges of relevance buckets. | 28 // This is used to store the score ranges of relevance buckets. |
28 // Please see GetFinalRelevancyScore() for details. | 29 // Please see GetFinalRelevancyScore() for details. |
(...skipping 12 matching lines...) Expand all Loading... |
41 // history item given in |row| with recent visits as indicated in |visits|. It | 42 // history item given in |row| with recent visits as indicated in |visits|. It |
42 // first determines if the row qualifies by seeing if all of the terms in | 43 // first determines if the row qualifies by seeing if all of the terms in |
43 // |terms_vector| occur in |row|. If so, calculates a raw score. This raw | 44 // |terms_vector| occur in |row|. If so, calculates a raw score. This raw |
44 // score is in part determined by whether the matches occur at word | 45 // score is in part determined by whether the matches occur at word |
45 // boundaries, the locations of which are stored in |word_starts|. For some | 46 // boundaries, the locations of which are stored in |word_starts|. For some |
46 // terms, it's appropriate to look for the word boundary within the term. For | 47 // terms, it's appropriate to look for the word boundary within the term. For |
47 // instance, the term ".net" should look for a word boundary at the "n". | 48 // instance, the term ".net" should look for a word boundary at the "n". |
48 // These offsets (".net" should have an offset of 1) come from | 49 // These offsets (".net" should have an offset of 1) come from |
49 // |terms_to_word_starts_offsets|. |is_url_bookmarked| indicates whether the | 50 // |terms_to_word_starts_offsets|. |is_url_bookmarked| indicates whether the |
50 // match's URL is referenced by any bookmarks, which can also affect the raw | 51 // match's URL is referenced by any bookmarks, which can also affect the raw |
51 // score. The raw score allows the matches to be ordered and can be used to | 52 // score. |num_matching_pages| indicates how many URLs in the eligible URL |
52 // influence the final score calculated by the client of this index. If the | 53 // database match the user's input; it can also affect the raw score. The raw |
53 // row does not qualify the raw score will be 0. | 54 // score allows the matches to be ordered and can be used to influence the |
| 55 // final score calculated by the client of this index. If the row does not |
| 56 // qualify the raw score will be 0. |
54 ScoredHistoryMatch(const history::URLRow& row, | 57 ScoredHistoryMatch(const history::URLRow& row, |
55 const VisitInfoVector& visits, | 58 const VisitInfoVector& visits, |
56 const base::string16& lower_string, | 59 const base::string16& lower_string, |
57 const String16Vector& terms_vector, | 60 const String16Vector& terms_vector, |
58 const WordStarts& terms_to_word_starts_offsets, | 61 const WordStarts& terms_to_word_starts_offsets, |
59 const RowWordStarts& word_starts, | 62 const RowWordStarts& word_starts, |
60 bool is_url_bookmarked, | 63 bool is_url_bookmarked, |
| 64 size_t num_matching_pages, |
61 base::Time now); | 65 base::Time now); |
62 | 66 |
63 ~ScoredHistoryMatch(); | 67 ~ScoredHistoryMatch(); |
64 | 68 |
65 // Compares two matches by score. Functor supporting URLIndexPrivateData's | 69 // Compares two matches by score. Functor supporting URLIndexPrivateData's |
66 // HistoryItemsForTerms function. Looks at particular fields within | 70 // HistoryItemsForTerms function. Looks at particular fields within |
67 // with url_info to make tie-breaking a bit smarter. | 71 // with url_info to make tie-breaking a bit smarter. |
68 static bool MatchScoreGreater(const ScoredHistoryMatch& m1, | 72 static bool MatchScoreGreater(const ScoredHistoryMatch& m1, |
69 const ScoredHistoryMatch& m2); | 73 const ScoredHistoryMatch& m2); |
70 | 74 |
(...skipping 20 matching lines...) Expand all Loading... |
91 // unimportant matches are eliminated by GetTopicalityScore(), called | 95 // unimportant matches are eliminated by GetTopicalityScore(), called |
92 // during construction.) | 96 // during construction.) |
93 | 97 |
94 // Term matches within the URL. | 98 // Term matches within the URL. |
95 TermMatches url_matches; | 99 TermMatches url_matches; |
96 // Term matches within the page title. | 100 // Term matches within the page title. |
97 TermMatches title_matches; | 101 TermMatches title_matches; |
98 | 102 |
99 private: | 103 private: |
100 friend class ScoredHistoryMatchTest; | 104 friend class ScoredHistoryMatchTest; |
| 105 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, GetDocumentSpecificityScore); |
101 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, GetFinalRelevancyScore); | 106 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, GetFinalRelevancyScore); |
102 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, GetFrequency); | 107 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, GetFrequency); |
103 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, GetHQPBucketsFromString); | 108 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, GetHQPBucketsFromString); |
104 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, ScoringBookmarks); | 109 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, ScoringBookmarks); |
105 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, ScoringScheme); | 110 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, ScoringScheme); |
106 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, ScoringTLD); | 111 FRIEND_TEST_ALL_PREFIXES(ScoredHistoryMatchTest, ScoringTLD); |
107 | 112 |
108 // Initialize ScoredHistoryMatch statics. Must be called before any other | 113 // Initialize ScoredHistoryMatch statics. Must be called before any other |
109 // method of ScoredHistoryMatch and before creating any instances. | 114 // method of ScoredHistoryMatch and before creating any instances. |
110 static void Init(); | 115 static void Init(); |
(...skipping 15 matching lines...) Expand all Loading... |
126 | 131 |
127 // Examines the first |max_visits_to_score_| and returns a score (higher is | 132 // Examines the first |max_visits_to_score_| and returns a score (higher is |
128 // better) based the rate of visits, whether the page is bookmarked, and | 133 // better) based the rate of visits, whether the page is bookmarked, and |
129 // how often those visits are typed navigations (i.e., explicitly | 134 // how often those visits are typed navigations (i.e., explicitly |
130 // invoked by the user). |now| is passed in to avoid unnecessarily | 135 // invoked by the user). |now| is passed in to avoid unnecessarily |
131 // recomputing it frequently. | 136 // recomputing it frequently. |
132 float GetFrequency(const base::Time& now, | 137 float GetFrequency(const base::Time& now, |
133 const bool bookmarked, | 138 const bool bookmarked, |
134 const VisitInfoVector& visits) const; | 139 const VisitInfoVector& visits) const; |
135 | 140 |
136 // Combines the two component scores into a final score that's an appropriate | 141 // Returns a document specificity score based on how many pages matched the |
137 // value to use as a relevancy score. | 142 // user's input. |
| 143 float GetDocumentSpecificityScore(size_t num_matching_pages) const; |
| 144 |
| 145 // Combines the three component scores into a final score that's |
| 146 // an appropriate value to use as a relevancy score. |
138 static float GetFinalRelevancyScore(float topicality_score, | 147 static float GetFinalRelevancyScore(float topicality_score, |
139 float frequency_score); | 148 float frequency_score, |
| 149 float specificity_score); |
140 | 150 |
141 // Helper function that returns the string containing the scoring buckets | 151 // Helper function that returns the string containing the scoring buckets |
142 // (either the default ones or ones specified in an experiment). | 152 // (either the default ones or ones specified in an experiment). |
143 static ScoreMaxRelevances GetHQPBuckets(); | 153 static ScoreMaxRelevances GetHQPBuckets(); |
144 | 154 |
145 // Helper function to parse the string containing the scoring buckets and | 155 // Helper function to parse the string containing the scoring buckets and |
146 // return the results. For example, with |buckets_str| as | 156 // return the results. For example, with |buckets_str| as |
147 // "0.0:400,1.5:600,12.0:1300,20.0:1399", it returns [(0.0, 400), (1.5, 600), | 157 // "0.0:400,1.5:600,12.0:1300,20.0:1399", it returns [(0.0, 400), (1.5, 600), |
148 // (12.0, 1300), (20.0, 1399)]. It returns an empty vector in the case of a | 158 // (12.0, 1300), (20.0, 1399)]. It returns an empty vector in the case of a |
149 // malformed |buckets_str|. | 159 // malformed |buckets_str|. |
(...skipping 34 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
184 | 194 |
185 // |topicality_threshold_| is used to control the topicality scoring. | 195 // |topicality_threshold_| is used to control the topicality scoring. |
186 // If |topicality_threshold_| > 0, then URLs with topicality-score less than | 196 // If |topicality_threshold_| > 0, then URLs with topicality-score less than |
187 // the threshold are given topicality score of 0. | 197 // the threshold are given topicality score of 0. |
188 static float topicality_threshold_; | 198 static float topicality_threshold_; |
189 | 199 |
190 // Used for testing. A possibly null pointer to a vector. If set, | 200 // Used for testing. A possibly null pointer to a vector. If set, |
191 // overrides the static local variable |relevance_buckets| declared in | 201 // overrides the static local variable |relevance_buckets| declared in |
192 // GetFinalRelevancyScore(). | 202 // GetFinalRelevancyScore(). |
193 static ScoreMaxRelevances* relevance_buckets_override_; | 203 static ScoreMaxRelevances* relevance_buckets_override_; |
| 204 |
| 205 // Used for testing. If this pointer is not null, it overrides the static |
| 206 // local variable |default_matches_to_specificity| declared in |
| 207 // GetDocumentSpecificityScore(). |
| 208 static OmniboxFieldTrial::NumMatchesScores* matches_to_specificity_override_; |
194 }; | 209 }; |
195 typedef std::vector<ScoredHistoryMatch> ScoredHistoryMatches; | 210 typedef std::vector<ScoredHistoryMatch> ScoredHistoryMatches; |
196 | 211 |
197 #endif // COMPONENTS_OMNIBOX_BROWSER_SCORED_HISTORY_MATCH_H_ | 212 #endif // COMPONENTS_OMNIBOX_BROWSER_SCORED_HISTORY_MATCH_H_ |
OLD | NEW |