Chromium Code Reviews| 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 #include "components/omnibox/browser/url_index_private_data.h" | 5 #include "components/omnibox/browser/url_index_private_data.h" |
| 6 | 6 |
| 7 #include <stdint.h> | 7 #include <stdint.h> |
| 8 | 8 |
| 9 #include <functional> | 9 #include <functional> |
| 10 #include <iterator> | 10 #include <iterator> |
| (...skipping 191 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 202 if (lower_words.empty()) | 202 if (lower_words.empty()) |
| 203 continue; | 203 continue; |
| 204 HistoryIDSet history_id_set = HistoryIDSetFromWords(lower_words); | 204 HistoryIDSet history_id_set = HistoryIDSetFromWords(lower_words); |
| 205 pre_filter_item_count_ += history_id_set.size(); | 205 pre_filter_item_count_ += history_id_set.size(); |
| 206 // Trim the candidate pool if it is large. Note that we do not filter out | 206 // Trim the candidate pool if it is large. Note that we do not filter out |
| 207 // items that do not contain the search terms as proper substrings -- | 207 // items that do not contain the search terms as proper substrings -- |
| 208 // doing so is the performance-costly operation we are trying to avoid in | 208 // doing so is the performance-costly operation we are trying to avoid in |
| 209 // order to maintain omnibox responsiveness. | 209 // order to maintain omnibox responsiveness. |
| 210 const size_t kItemsToScoreLimit = 500; | 210 const size_t kItemsToScoreLimit = 500; |
| 211 if (history_id_set.size() > kItemsToScoreLimit) { | 211 if (history_id_set.size() > kItemsToScoreLimit) { |
| 212 HistoryIDVector history_ids; | 212 HistoryIDVector history_ids = {history_id_set.begin(), |
| 213 std::copy(history_id_set.begin(), history_id_set.end(), | 213 history_id_set.end()}; |
|
Peter Kasting
2017/02/17 01:33:46
Nit: Prefer the direct constructor:
History
dyaroshev
2017/02/17 20:17:22
Done.
| |
| 214 std::back_inserter(history_ids)); | 214 |
| 215 // Trim down the set by sorting by typed-count, visit-count, and last | 215 // Trim down the set by sorting by typed-count, visit-count, and last |
| 216 // visit. | 216 // visit. |
| 217 HistoryItemFactorGreater item_factor_functor(history_info_map_); | 217 HistoryItemFactorGreater item_factor_functor(history_info_map_); |
| 218 std::partial_sort(history_ids.begin(), | 218 std::nth_element(history_ids.begin(), |
|
Peter Kasting
2017/02/17 01:33:46
I've never seen this algorithm before. Good use o
dyaroshev
2017/02/17 20:17:22
thx) While we are here - I'm not 100% sure, but it
Peter Kasting
2017/02/17 21:52:26
Correct, we're going to score them and then sort t
| |
| 219 history_ids.begin() + kItemsToScoreLimit, | 219 history_ids.begin() + kItemsToScoreLimit, |
| 220 history_ids.end(), item_factor_functor); | 220 history_ids.end(), item_factor_functor); |
| 221 history_id_set.clear(); | 221 history_id_set = {history_ids.begin(), |
| 222 std::copy(history_ids.begin(), history_ids.begin() + kItemsToScoreLimit, | 222 history_ids.begin() + kItemsToScoreLimit}; |
| 223 std::inserter(history_id_set, history_id_set.end())); | |
| 224 post_filter_item_count_ += history_id_set.size(); | 223 post_filter_item_count_ += history_id_set.size(); |
| 225 } else { | 224 } else { |
| 226 post_filter_item_count_ += pre_filter_item_count_; | 225 post_filter_item_count_ += pre_filter_item_count_; |
| 227 } | 226 } |
| 228 ScoredHistoryMatches temp_scored_items; | 227 ScoredHistoryMatches temp_scored_items; |
| 229 HistoryIdSetToScoredMatches(history_id_set, lower_raw_string, | 228 HistoryIdSetToScoredMatches(history_id_set, lower_raw_string, |
| 230 template_url_service, bookmark_model, | 229 template_url_service, bookmark_model, |
| 231 &temp_scored_items); | 230 &temp_scored_items); |
| 232 scored_items.insert(scored_items.end(), temp_scored_items.begin(), | 231 scored_items.insert(scored_items.end(), temp_scored_items.begin(), |
| 233 temp_scored_items.end()); | 232 temp_scored_items.end()); |
| (...skipping 270 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 504 // a string like "http://www.somewebsite.com" which, from our perspective, | 503 // a string like "http://www.somewebsite.com" which, from our perspective, |
| 505 // is four words: 'http', 'www', 'somewebsite', and 'com'. | 504 // is four words: 'http', 'www', 'somewebsite', and 'com'. |
| 506 HistoryIDSet history_id_set; | 505 HistoryIDSet history_id_set; |
| 507 String16Vector words(unsorted_words); | 506 String16Vector words(unsorted_words); |
| 508 // Sort the words into the longest first as such are likely to narrow down | 507 // Sort the words into the longest first as such are likely to narrow down |
| 509 // the results quicker. Also, single character words are the most expensive | 508 // the results quicker. Also, single character words are the most expensive |
| 510 // to process so save them for last. | 509 // to process so save them for last. |
| 511 std::sort(words.begin(), words.end(), LengthGreater); | 510 std::sort(words.begin(), words.end(), LengthGreater); |
| 512 for (String16Vector::iterator iter = words.begin(); iter != words.end(); | 511 for (String16Vector::iterator iter = words.begin(); iter != words.end(); |
| 513 ++iter) { | 512 ++iter) { |
| 514 base::string16 uni_word = *iter; | 513 HistoryIDSet term_history_set = HistoryIDsForTerm(*iter); |
| 515 HistoryIDSet term_history_set = HistoryIDsForTerm(uni_word); | |
| 516 if (term_history_set.empty()) { | 514 if (term_history_set.empty()) { |
|
Peter Kasting
2017/02/17 01:33:46
Nit: No {}
dyaroshev
2017/02/17 20:17:22
Done.
| |
| 517 history_id_set.clear(); | 515 return {}; |
|
Peter Kasting
2017/02/17 01:33:46
Nit: I'd return HistoryIDSet(); (3 similar places)
dyaroshev
2017/02/17 20:17:22
Done.
| |
| 518 break; | |
| 519 } | 516 } |
| 520 if (iter == words.begin()) { | 517 if (iter == words.begin()) { |
| 521 history_id_set.swap(term_history_set); | 518 history_id_set = std::move(term_history_set); |
| 522 } else { | 519 } else { |
| 523 HistoryIDSet new_history_id_set = base::STLSetIntersection<HistoryIDSet>( | 520 history_id_set = base::STLSetIntersection<HistoryIDSet>(history_id_set, |
| 524 history_id_set, term_history_set); | 521 term_history_set); |
| 525 history_id_set.swap(new_history_id_set); | |
| 526 } | 522 } |
| 527 } | 523 } |
| 528 return history_id_set; | 524 return history_id_set; |
| 529 } | 525 } |
| 530 | 526 |
| 531 HistoryIDSet URLIndexPrivateData::HistoryIDsForTerm( | 527 HistoryIDSet URLIndexPrivateData::HistoryIDsForTerm( |
| 532 const base::string16& term) { | 528 const base::string16& term) { |
| 533 if (term.empty()) | 529 if (term.empty()) |
| 534 return HistoryIDSet(); | 530 return HistoryIDSet(); |
| 535 | 531 |
| (...skipping 49 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 585 | 581 |
| 586 // Reduce the word set with any leftover, unprocessed characters. | 582 // Reduce the word set with any leftover, unprocessed characters. |
| 587 if (!unique_chars.empty()) { | 583 if (!unique_chars.empty()) { |
| 588 WordIDSet leftover_set(WordIDSetForTermChars(unique_chars)); | 584 WordIDSet leftover_set(WordIDSetForTermChars(unique_chars)); |
| 589 // We might come up empty on the leftovers. | 585 // We might come up empty on the leftovers. |
| 590 if (leftover_set.empty()) { | 586 if (leftover_set.empty()) { |
| 591 search_term_cache_[term] = SearchTermCacheItem(); | 587 search_term_cache_[term] = SearchTermCacheItem(); |
| 592 return HistoryIDSet(); | 588 return HistoryIDSet(); |
| 593 } | 589 } |
| 594 // Or there may not have been a prefix from which to start. | 590 // Or there may not have been a prefix from which to start. |
| 595 if (prefix_chars.empty()) { | 591 word_id_set = prefix_chars.empty() ? leftover_set |
| 596 word_id_set.swap(leftover_set); | 592 : base::STLSetIntersection<WordIDSet>( |
| 597 } else { | 593 word_id_set, leftover_set); |
| 598 WordIDSet new_word_id_set = base::STLSetIntersection<WordIDSet>( | |
| 599 word_id_set, leftover_set); | |
| 600 word_id_set.swap(new_word_id_set); | |
| 601 } | |
| 602 } | 594 } |
| 603 | 595 |
| 604 // We must filter the word list because the resulting word set surely | 596 // We must filter the word list because the resulting word set surely |
| 605 // contains words which do not have the search term as a proper subset. | 597 // contains words which do not have the search term as a proper subset. |
| 606 for (WordIDSet::iterator word_set_iter = word_id_set.begin(); | 598 word_id_set.erase(std::remove_if(word_id_set.begin(), word_id_set.end(), |
| 607 word_set_iter != word_id_set.end(); ) { | 599 [&](WordID word_id) { |
| 608 if (word_list_[*word_set_iter].find(term) == base::string16::npos) | 600 return word_list_[word_id].find(term) == |
| 609 word_set_iter = word_id_set.erase(word_set_iter); | 601 base::string16::npos; |
| 610 else | 602 }), |
| 611 ++word_set_iter; | 603 word_id_set.end()); |
|
Peter Kasting
2017/02/17 01:33:46
Nit: Slightly shorter, reads a bit less awkwardly
| |
| 612 } | 604 |
| 613 } else { | 605 } else { |
| 614 word_id_set = WordIDSetForTermChars(Char16SetFromString16(term)); | 606 word_id_set = WordIDSetForTermChars(Char16SetFromString16(term)); |
| 615 } | 607 } |
| 616 | 608 |
| 617 // If any words resulted then we can compose a set of history IDs by unioning | 609 // If any words resulted then we can compose a set of history IDs by unioning |
| 618 // the sets from each word. | 610 // the sets from each word. |
| 619 HistoryIDSet history_id_set; | 611 auto history_id_set = [&]() -> HistoryIDSet { |
|
Peter Kasting
2017/02/17 01:33:46
Why do this in a lambda? Why not just "inline" th
| |
| 620 if (!word_id_set.empty()) { | 612 HistoryIDVector buffer; |
|
Peter Kasting
2017/02/17 01:33:46
Nit: Deserves a comment about why it's important t
| |
| 621 for (WordIDSet::iterator word_id_iter = word_id_set.begin(); | 613 |
|
Peter Kasting
2017/02/17 01:33:46
Nit: No blank line
| |
| 622 word_id_iter != word_id_set.end(); ++word_id_iter) { | 614 for (WordID word_id : word_id_set) { |
| 623 WordID word_id = *word_id_iter; | |
| 624 WordIDHistoryMap::iterator word_iter = word_id_history_map_.find(word_id); | 615 WordIDHistoryMap::iterator word_iter = word_id_history_map_.find(word_id); |
| 625 if (word_iter != word_id_history_map_.end()) { | 616 if (word_iter == word_id_history_map_.end()) |
| 626 HistoryIDSet& word_history_id_set(word_iter->second); | 617 continue; |
|
Peter Kasting
2017/02/17 01:33:46
Nit: I probably wouldn't change the old form here
| |
| 627 history_id_set.insert(word_history_id_set.begin(), | 618 HistoryIDSet& word_history_id_set(word_iter->second); |
| 628 word_history_id_set.end()); | 619 buffer.insert(buffer.end(), word_history_id_set.begin(), |
| 629 } | 620 word_history_id_set.end()); |
| 630 } | 621 } |
| 631 } | 622 |
| 623 return {buffer.begin(), buffer.end()}; | |
| 624 }(); | |
| 632 | 625 |
| 633 // Record a new cache entry for this word if the term is longer than | 626 // Record a new cache entry for this word if the term is longer than |
| 634 // a single character. | 627 // a single character. |
| 635 if (term_length > 1) | 628 if (term_length > 1) |
| 636 search_term_cache_[term] = SearchTermCacheItem(word_id_set, history_id_set); | 629 search_term_cache_[term] = SearchTermCacheItem(word_id_set, history_id_set); |
| 637 | 630 |
| 638 return history_id_set; | 631 return history_id_set; |
| 639 } | 632 } |
| 640 | 633 |
| 641 WordIDSet URLIndexPrivateData::WordIDSetForTermChars( | 634 WordIDSet URLIndexPrivateData::WordIDSetForTermChars( |
| 642 const Char16Set& term_chars) { | 635 const Char16Set& term_chars) { |
| 643 WordIDSet word_id_set; | 636 WordIDSet word_id_set; |
| 644 for (Char16Set::const_iterator c_iter = term_chars.begin(); | |
| 645 c_iter != term_chars.end(); ++c_iter) { | |
| 646 CharWordIDMap::iterator char_iter = char_word_map_.find(*c_iter); | |
| 647 if (char_iter == char_word_map_.end()) { | |
| 648 // A character was not found so there are no matching results: bail. | |
| 649 word_id_set.clear(); | |
| 650 break; | |
| 651 } | |
| 652 WordIDSet& char_word_id_set(char_iter->second); | |
| 653 // It is possible for there to no longer be any words associated with | |
| 654 // a particular character. Give up in that case. | |
| 655 if (char_word_id_set.empty()) { | |
| 656 word_id_set.clear(); | |
| 657 break; | |
| 658 } | |
| 659 | 637 |
| 660 if (c_iter == term_chars.begin()) { | 638 for (base::char16 c : term_chars) { |
| 661 // First character results becomes base set of results. | 639 CharWordIDMap::iterator char_iter = char_word_map_.find(c); |
| 662 word_id_set = char_word_id_set; | 640 if (char_iter == char_word_map_.end()) |
|
Peter Kasting
2017/02/17 01:33:46
Nit: Preserve the old comments
dyaroshev
2017/02/17 20:17:23
Done.
| |
| 663 } else { | 641 return {}; |
| 664 // Subsequent character results get intersected in. | 642 |
| 665 WordIDSet new_word_id_set = base::STLSetIntersection<WordIDSet>( | 643 const WordIDSet& char_word_id_set(char_iter->second); |
| 666 word_id_set, char_word_id_set); | 644 if (char_word_id_set.empty()) |
| 667 word_id_set.swap(new_word_id_set); | 645 return {}; |
| 668 } | 646 |
| 647 word_id_set = | |
| 648 base::STLSetIntersection<WordIDSet>(word_id_set, char_word_id_set); | |
|
Peter Kasting
2017/02/17 01:33:46
Does dropping the assignment the old code used in
dyaroshev
2017/02/17 20:17:23
Yes, my mistake. Done. Reverted almost everything
| |
| 669 } | 649 } |
| 650 | |
| 670 return word_id_set; | 651 return word_id_set; |
| 671 } | 652 } |
| 672 | 653 |
| 673 void URLIndexPrivateData::HistoryIdSetToScoredMatches( | 654 void URLIndexPrivateData::HistoryIdSetToScoredMatches( |
| 674 HistoryIDSet history_id_set, | 655 HistoryIDSet history_id_set, |
| 675 const base::string16& lower_raw_string, | 656 const base::string16& lower_raw_string, |
| 676 const TemplateURLService* template_url_service, | 657 const TemplateURLService* template_url_service, |
| 677 bookmarks::BookmarkModel* bookmark_model, | 658 bookmarks::BookmarkModel* bookmark_model, |
| 678 ScoredHistoryMatches* scored_items) const { | 659 ScoredHistoryMatches* scored_items) const { |
| 679 if (history_id_set.empty()) | 660 if (history_id_set.empty()) |
| (...skipping 17 matching lines...) Expand all Loading... | |
| 697 // are some form of whitespace), but this is such a rare edge case that it's | 678 // are some form of whitespace), but this is such a rare edge case that it's |
| 698 // not worth the time. | 679 // not worth the time. |
| 699 if (lower_raw_terms.empty()) | 680 if (lower_raw_terms.empty()) |
| 700 return; | 681 return; |
| 701 | 682 |
| 702 WordStarts lower_terms_to_word_starts_offsets; | 683 WordStarts lower_terms_to_word_starts_offsets; |
| 703 CalculateWordStartsOffsets(lower_raw_terms, | 684 CalculateWordStartsOffsets(lower_raw_terms, |
| 704 &lower_terms_to_word_starts_offsets); | 685 &lower_terms_to_word_starts_offsets); |
| 705 | 686 |
| 706 // Filter bad matches and other matches we don't want to display. | 687 // Filter bad matches and other matches we don't want to display. |
| 707 for (auto it = history_id_set.begin();;) { | 688 history_id_set.erase( |
| 708 it = std::find_if(it, history_id_set.end(), | 689 std::remove_if(history_id_set.begin(), history_id_set.end(), |
| 709 [this, template_url_service](const HistoryID history_id) { | 690 [this, template_url_service](const HistoryID history_id) { |
| 710 return ShouldFilter(history_id, template_url_service); | 691 return ShouldFilter(history_id, template_url_service); |
| 711 }); | 692 }), |
| 712 if (it == history_id_set.end()) | 693 history_id_set.end()); |
| 713 break; | |
| 714 it = history_id_set.erase(it); | |
| 715 } | |
| 716 | 694 |
| 717 // Score the matches. | 695 // Score the matches. |
| 718 const size_t num_matches = history_id_set.size(); | 696 const size_t num_matches = history_id_set.size(); |
| 719 const base::Time now = base::Time::Now(); | 697 const base::Time now = base::Time::Now(); |
| 720 std::transform( | 698 std::transform( |
| 721 history_id_set.begin(), history_id_set.end(), | 699 history_id_set.begin(), history_id_set.end(), |
| 722 std::back_inserter(*scored_items), [&](const HistoryID history_id) { | 700 std::back_inserter(*scored_items), [&](const HistoryID history_id) { |
| 723 auto hist_pos = history_info_map_.find(history_id); | 701 auto hist_pos = history_info_map_.find(history_id); |
| 724 const history::URLRow& hist_item = hist_pos->second.url_row; | 702 const history::URLRow& hist_item = hist_pos->second.url_row; |
| 725 auto starts_pos = word_starts_map_.find(history_id); | 703 auto starts_pos = word_starts_map_.find(history_id); |
| (...skipping 97 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 823 HistoryID history_id = static_cast<HistoryID>(row.id()); | 801 HistoryID history_id = static_cast<HistoryID>(row.id()); |
| 824 // Split URL into individual, unique words then add in the title words. | 802 // Split URL into individual, unique words then add in the title words. |
| 825 const GURL& gurl(row.url()); | 803 const GURL& gurl(row.url()); |
| 826 const base::string16& url = | 804 const base::string16& url = |
| 827 bookmarks::CleanUpUrlForMatching(gurl, nullptr); | 805 bookmarks::CleanUpUrlForMatching(gurl, nullptr); |
| 828 String16Set url_words = String16SetFromString16(url, | 806 String16Set url_words = String16SetFromString16(url, |
| 829 word_starts ? &word_starts->url_word_starts_ : nullptr); | 807 word_starts ? &word_starts->url_word_starts_ : nullptr); |
| 830 const base::string16& title = bookmarks::CleanUpTitleForMatching(row.title()); | 808 const base::string16& title = bookmarks::CleanUpTitleForMatching(row.title()); |
| 831 String16Set title_words = String16SetFromString16(title, | 809 String16Set title_words = String16SetFromString16(title, |
| 832 word_starts ? &word_starts->title_word_starts_ : nullptr); | 810 word_starts ? &word_starts->title_word_starts_ : nullptr); |
| 833 String16Set words = base::STLSetUnion<String16Set>(url_words, title_words); | 811 for (const auto& word : |
| 834 for (String16Set::iterator word_iter = words.begin(); | 812 base::STLSetUnion<String16Set>(url_words, title_words)) |
| 835 word_iter != words.end(); ++word_iter) | 813 AddWordToIndex(word, history_id); |
| 836 AddWordToIndex(*word_iter, history_id); | |
| 837 | 814 |
| 838 search_term_cache_.clear(); // Invalidate the term cache. | 815 search_term_cache_.clear(); // Invalidate the term cache. |
| 839 } | 816 } |
| 840 | 817 |
| 841 void URLIndexPrivateData::AddWordToIndex(const base::string16& term, | 818 void URLIndexPrivateData::AddWordToIndex(const base::string16& term, |
| 842 HistoryID history_id) { | 819 HistoryID history_id) { |
| 843 WordMap::iterator word_pos = word_map_.find(term); | 820 WordMap::iterator word_pos = word_map_.find(term); |
| 844 if (word_pos != word_map_.end()) | 821 if (word_pos != word_map_.end()) |
| 845 UpdateWordHistory(word_pos->second, history_id); | 822 UpdateWordHistory(word_pos->second, history_id); |
| 846 else | 823 else |
| (...skipping 12 matching lines...) Expand all Loading... | |
| 859 } | 836 } |
| 860 word_map_[term] = word_id; | 837 word_map_[term] = word_id; |
| 861 | 838 |
| 862 HistoryIDSet history_id_set; | 839 HistoryIDSet history_id_set; |
| 863 history_id_set.insert(history_id); | 840 history_id_set.insert(history_id); |
| 864 word_id_history_map_[word_id] = history_id_set; | 841 word_id_history_map_[word_id] = history_id_set; |
| 865 AddToHistoryIDWordMap(history_id, word_id); | 842 AddToHistoryIDWordMap(history_id, word_id); |
| 866 | 843 |
| 867 // For each character in the newly added word (i.e. a word that is not | 844 // For each character in the newly added word (i.e. a word that is not |
| 868 // already in the word index), add the word to the character index. | 845 // already in the word index), add the word to the character index. |
| 869 Char16Set characters = Char16SetFromString16(term); | 846 for (base::char16 uni_char : Char16SetFromString16(term)) { |
|
Peter Kasting
2017/02/17 01:33:46
Nit: No {}
| |
| 870 for (Char16Set::iterator uni_char_iter = characters.begin(); | 847 char_word_map_[uni_char].insert(word_id); |
|
Peter Kasting
2017/02/17 01:33:46
Nice simplification!
dyaroshev
2017/02/17 20:17:23
I rewrote AddWordToIndex, AddWordHistory and Updat
| |
| 871 uni_char_iter != characters.end(); ++uni_char_iter) { | |
| 872 base::char16 uni_char = *uni_char_iter; | |
| 873 CharWordIDMap::iterator char_iter = char_word_map_.find(uni_char); | |
| 874 if (char_iter != char_word_map_.end()) { | |
| 875 // Update existing entry in the char/word index. | |
| 876 WordIDSet& word_id_set(char_iter->second); | |
| 877 word_id_set.insert(word_id); | |
| 878 } else { | |
| 879 // Create a new entry in the char/word index. | |
| 880 WordIDSet word_id_set; | |
| 881 word_id_set.insert(word_id); | |
| 882 char_word_map_[uni_char] = word_id_set; | |
| 883 } | |
| 884 } | 848 } |
| 885 } | 849 } |
| 886 | 850 |
| 887 void URLIndexPrivateData::UpdateWordHistory(WordID word_id, | 851 void URLIndexPrivateData::UpdateWordHistory(WordID word_id, |
| 888 HistoryID history_id) { | 852 HistoryID history_id) { |
| 889 WordIDHistoryMap::iterator history_pos = word_id_history_map_.find(word_id); | 853 WordIDHistoryMap::iterator history_pos = word_id_history_map_.find(word_id); |
| 890 DCHECK(history_pos != word_id_history_map_.end()); | 854 DCHECK(history_pos != word_id_history_map_.end()); |
| 891 HistoryIDSet& history_id_set(history_pos->second); | 855 HistoryIDSet& history_id_set(history_pos->second); |
| 892 history_id_set.insert(history_id); | 856 history_id_set.insert(history_id); |
| 893 AddToHistoryIDWordMap(history_id, word_id); | 857 AddToHistoryIDWordMap(history_id, word_id); |
| 894 } | 858 } |
| 895 | 859 |
| 896 void URLIndexPrivateData::AddToHistoryIDWordMap(HistoryID history_id, | 860 void URLIndexPrivateData::AddToHistoryIDWordMap(HistoryID history_id, |
| 897 WordID word_id) { | 861 WordID word_id) { |
| 898 HistoryIDWordMap::iterator iter = history_id_word_map_.find(history_id); | 862 history_id_word_map_[history_id].insert(word_id); |
| 899 if (iter != history_id_word_map_.end()) { | |
| 900 WordIDSet& word_id_set(iter->second); | |
| 901 word_id_set.insert(word_id); | |
| 902 } else { | |
| 903 WordIDSet word_id_set; | |
| 904 word_id_set.insert(word_id); | |
| 905 history_id_word_map_[history_id] = word_id_set; | |
| 906 } | |
| 907 } | 863 } |
| 908 | 864 |
| 909 void URLIndexPrivateData::RemoveRowFromIndex(const history::URLRow& row) { | 865 void URLIndexPrivateData::RemoveRowFromIndex(const history::URLRow& row) { |
| 910 RemoveRowWordsFromIndex(row); | 866 RemoveRowWordsFromIndex(row); |
| 911 HistoryID history_id = static_cast<HistoryID>(row.id()); | 867 HistoryID history_id = static_cast<HistoryID>(row.id()); |
| 912 history_info_map_.erase(history_id); | 868 history_info_map_.erase(history_id); |
| 913 word_starts_map_.erase(history_id); | 869 word_starts_map_.erase(history_id); |
| 914 } | 870 } |
| 915 | 871 |
| 916 void URLIndexPrivateData::RemoveRowWordsFromIndex(const history::URLRow& row) { | 872 void URLIndexPrivateData::RemoveRowWordsFromIndex(const history::URLRow& row) { |
| 917 // Remove the entries in history_id_word_map_ and word_id_history_map_ for | 873 // Remove the entries in history_id_word_map_ and word_id_history_map_ for |
| 918 // this row. | 874 // this row. |
| 919 HistoryID history_id = static_cast<HistoryID>(row.id()); | 875 HistoryID history_id = static_cast<HistoryID>(row.id()); |
| 920 WordIDSet word_id_set = history_id_word_map_[history_id]; | 876 WordIDSet word_id_set = history_id_word_map_[history_id]; |
| 921 history_id_word_map_.erase(history_id); | 877 history_id_word_map_.erase(history_id); |
| 922 | 878 |
| 923 // Reconcile any changes to word usage. | 879 // Reconcile any changes to word usage. |
| 924 for (WordIDSet::iterator word_id_iter = word_id_set.begin(); | 880 for (WordID word_id : word_id_set) { |
| 925 word_id_iter != word_id_set.end(); ++word_id_iter) { | 881 auto word_id_history_map_iter = word_id_history_map_.find(word_id); |
| 926 WordID word_id = *word_id_iter; | 882 |
|
Peter Kasting
2017/02/17 01:33:46
Nit: No blank line
dyaroshev
2017/02/17 20:17:22
Done.
| |
| 927 word_id_history_map_[word_id].erase(history_id); | 883 if (word_id_history_map_iter == word_id_history_map_.end()) |
|
Peter Kasting
2017/02/17 01:33:46
Can this conditional succeed? The old code doesn'
dyaroshev
2017/02/17 20:17:22
Seems like no, we always add word to this map. And
| |
| 928 if (!word_id_history_map_[word_id].empty()) | 884 continue; |
| 929 continue; // The word is still in use. | 885 |
| 886 word_id_history_map_iter->second.erase(history_id); | |
| 887 if (!word_id_history_map_iter->second.empty()) | |
| 888 continue; | |
| 930 | 889 |
| 931 // The word is no longer in use. Reconcile any changes to character usage. | 890 // The word is no longer in use. Reconcile any changes to character usage. |
| 932 base::string16 word = word_list_[word_id]; | 891 base::string16 word = word_list_[word_id]; |
| 933 Char16Set characters = Char16SetFromString16(word); | 892 for (base::char16 uni_char : Char16SetFromString16(word)) { |
| 934 for (Char16Set::iterator uni_char_iter = characters.begin(); | 893 auto char_word_map_iter = char_word_map_.find(uni_char); |
| 935 uni_char_iter != characters.end(); ++uni_char_iter) { | 894 char_word_map_iter->second.erase(word_id); |
| 936 base::char16 uni_char = *uni_char_iter; | 895 if (char_word_map_iter->second.empty()) |
| 937 char_word_map_[uni_char].erase(word_id); | 896 char_word_map_.erase(char_word_map_iter); |
| 938 if (char_word_map_[uni_char].empty()) | |
| 939 char_word_map_.erase(uni_char); // No longer in use. | |
| 940 } | 897 } |
| 941 | 898 |
| 942 // Complete the removal of references to the word. | 899 // Complete the removal of references to the word. |
| 943 word_id_history_map_.erase(word_id); | 900 word_id_history_map_.erase(word_id_history_map_iter); |
| 944 word_map_.erase(word); | 901 word_map_.erase(word); |
| 945 word_list_[word_id] = base::string16(); | 902 word_list_[word_id] = base::string16(); |
| 946 available_words_.insert(word_id); | 903 available_words_.insert(word_id); |
| 947 } | 904 } |
| 948 } | 905 } |
| 949 | 906 |
| 950 void URLIndexPrivateData::ResetSearchTermCache() { | 907 void URLIndexPrivateData::ResetSearchTermCache() { |
| 951 for (SearchTermCacheMap::iterator iter = search_term_cache_.begin(); | 908 for (auto& item : search_term_cache_) |
| 952 iter != search_term_cache_.end(); ++iter) | 909 item.second.used_ = false; |
| 953 iter->second.used_ = false; | |
| 954 } | 910 } |
| 955 | 911 |
| 956 bool URLIndexPrivateData::SaveToFile(const base::FilePath& file_path) { | 912 bool URLIndexPrivateData::SaveToFile(const base::FilePath& file_path) { |
| 957 base::TimeTicks beginning_time = base::TimeTicks::Now(); | 913 base::TimeTicks beginning_time = base::TimeTicks::Now(); |
| 958 InMemoryURLIndexCacheItem index_cache; | 914 InMemoryURLIndexCacheItem index_cache; |
| 959 SavePrivateData(&index_cache); | 915 SavePrivateData(&index_cache); |
| 960 std::string data; | 916 std::string data; |
| 961 if (!index_cache.SerializeToString(&data)) { | 917 if (!index_cache.SerializeToString(&data)) { |
| 962 LOG(WARNING) << "Failed to serialize the InMemoryURLIndex cache."; | 918 LOG(WARNING) << "Failed to serialize the InMemoryURLIndex cache."; |
| 963 return false; | 919 return false; |
| (...skipping 24 matching lines...) Expand all Loading... | |
| 988 SaveWordIDHistoryMap(cache); | 944 SaveWordIDHistoryMap(cache); |
| 989 SaveHistoryInfoMap(cache); | 945 SaveHistoryInfoMap(cache); |
| 990 SaveWordStartsMap(cache); | 946 SaveWordStartsMap(cache); |
| 991 } | 947 } |
| 992 | 948 |
| 993 void URLIndexPrivateData::SaveWordList(InMemoryURLIndexCacheItem* cache) const { | 949 void URLIndexPrivateData::SaveWordList(InMemoryURLIndexCacheItem* cache) const { |
| 994 if (word_list_.empty()) | 950 if (word_list_.empty()) |
| 995 return; | 951 return; |
| 996 WordListItem* list_item = cache->mutable_word_list(); | 952 WordListItem* list_item = cache->mutable_word_list(); |
| 997 list_item->set_word_count(word_list_.size()); | 953 list_item->set_word_count(word_list_.size()); |
| 998 for (String16Vector::const_iterator iter = word_list_.begin(); | 954 for (const base::string16& word : word_list_) |
| 999 iter != word_list_.end(); ++iter) | 955 list_item->add_word(base::UTF16ToUTF8(word)); |
| 1000 list_item->add_word(base::UTF16ToUTF8(*iter)); | |
| 1001 } | 956 } |
| 1002 | 957 |
| 1003 void URLIndexPrivateData::SaveWordMap(InMemoryURLIndexCacheItem* cache) const { | 958 void URLIndexPrivateData::SaveWordMap(InMemoryURLIndexCacheItem* cache) const { |
| 1004 if (word_map_.empty()) | 959 if (word_map_.empty()) |
| 1005 return; | 960 return; |
| 1006 WordMapItem* map_item = cache->mutable_word_map(); | 961 WordMapItem* map_item = cache->mutable_word_map(); |
| 1007 map_item->set_item_count(word_map_.size()); | 962 map_item->set_item_count(word_map_.size()); |
| 1008 for (WordMap::const_iterator iter = word_map_.begin(); | 963 for (const auto& elem : word_map_) { |
| 1009 iter != word_map_.end(); ++iter) { | |
| 1010 WordMapEntry* map_entry = map_item->add_word_map_entry(); | 964 WordMapEntry* map_entry = map_item->add_word_map_entry(); |
| 1011 map_entry->set_word(base::UTF16ToUTF8(iter->first)); | 965 map_entry->set_word(base::UTF16ToUTF8(elem.first)); |
| 1012 map_entry->set_word_id(iter->second); | 966 map_entry->set_word_id(elem.second); |
| 1013 } | 967 } |
| 1014 } | 968 } |
| 1015 | 969 |
| 1016 void URLIndexPrivateData::SaveCharWordMap( | 970 void URLIndexPrivateData::SaveCharWordMap( |
| 1017 InMemoryURLIndexCacheItem* cache) const { | 971 InMemoryURLIndexCacheItem* cache) const { |
| 1018 if (char_word_map_.empty()) | 972 if (char_word_map_.empty()) |
| 1019 return; | 973 return; |
| 1020 CharWordMapItem* map_item = cache->mutable_char_word_map(); | 974 CharWordMapItem* map_item = cache->mutable_char_word_map(); |
| 1021 map_item->set_item_count(char_word_map_.size()); | 975 map_item->set_item_count(char_word_map_.size()); |
| 1022 for (CharWordIDMap::const_iterator iter = char_word_map_.begin(); | 976 for (const auto& elem : char_word_map_) { |
| 1023 iter != char_word_map_.end(); ++iter) { | |
| 1024 CharWordMapEntry* map_entry = map_item->add_char_word_map_entry(); | 977 CharWordMapEntry* map_entry = map_item->add_char_word_map_entry(); |
| 1025 map_entry->set_char_16(iter->first); | 978 map_entry->set_char_16(elem.first); |
| 1026 const WordIDSet& word_id_set(iter->second); | 979 const WordIDSet& word_id_set(elem.second); |
| 1027 map_entry->set_item_count(word_id_set.size()); | 980 map_entry->set_item_count(word_id_set.size()); |
| 1028 for (WordIDSet::const_iterator set_iter = word_id_set.begin(); | 981 for (WordID word_id : word_id_set) |
| 1029 set_iter != word_id_set.end(); ++set_iter) | 982 map_entry->add_word_id(word_id); |
| 1030 map_entry->add_word_id(*set_iter); | |
| 1031 } | 983 } |
| 1032 } | 984 } |
| 1033 | 985 |
| 1034 void URLIndexPrivateData::SaveWordIDHistoryMap( | 986 void URLIndexPrivateData::SaveWordIDHistoryMap( |
| 1035 InMemoryURLIndexCacheItem* cache) const { | 987 InMemoryURLIndexCacheItem* cache) const { |
| 1036 if (word_id_history_map_.empty()) | 988 if (word_id_history_map_.empty()) |
| 1037 return; | 989 return; |
| 1038 WordIDHistoryMapItem* map_item = cache->mutable_word_id_history_map(); | 990 WordIDHistoryMapItem* map_item = cache->mutable_word_id_history_map(); |
| 1039 map_item->set_item_count(word_id_history_map_.size()); | 991 map_item->set_item_count(word_id_history_map_.size()); |
| 1040 for (WordIDHistoryMap::const_iterator iter = word_id_history_map_.begin(); | 992 for (const auto& elem : word_id_history_map_) { |
| 1041 iter != word_id_history_map_.end(); ++iter) { | |
| 1042 WordIDHistoryMapEntry* map_entry = | 993 WordIDHistoryMapEntry* map_entry = |
| 1043 map_item->add_word_id_history_map_entry(); | 994 map_item->add_word_id_history_map_entry(); |
| 1044 map_entry->set_word_id(iter->first); | 995 map_entry->set_word_id(elem.first); |
| 1045 const HistoryIDSet& history_id_set(iter->second); | 996 const HistoryIDSet& history_id_set(elem.second); |
| 1046 map_entry->set_item_count(history_id_set.size()); | 997 map_entry->set_item_count(history_id_set.size()); |
| 1047 for (HistoryIDSet::const_iterator set_iter = history_id_set.begin(); | 998 for (HistoryID history_id : history_id_set) |
| 1048 set_iter != history_id_set.end(); ++set_iter) | 999 map_entry->add_history_id(history_id); |
| 1049 map_entry->add_history_id(*set_iter); | |
| 1050 } | 1000 } |
| 1051 } | 1001 } |
| 1052 | 1002 |
| 1053 void URLIndexPrivateData::SaveHistoryInfoMap( | 1003 void URLIndexPrivateData::SaveHistoryInfoMap( |
| 1054 InMemoryURLIndexCacheItem* cache) const { | 1004 InMemoryURLIndexCacheItem* cache) const { |
| 1055 if (history_info_map_.empty()) | 1005 if (history_info_map_.empty()) |
| 1056 return; | 1006 return; |
| 1057 HistoryInfoMapItem* map_item = cache->mutable_history_info_map(); | 1007 HistoryInfoMapItem* map_item = cache->mutable_history_info_map(); |
| 1058 map_item->set_item_count(history_info_map_.size()); | 1008 map_item->set_item_count(history_info_map_.size()); |
| 1059 for (HistoryInfoMap::const_iterator iter = history_info_map_.begin(); | 1009 for (const auto& elem : history_info_map_) { |
| 1060 iter != history_info_map_.end(); ++iter) { | |
| 1061 HistoryInfoMapEntry* map_entry = map_item->add_history_info_map_entry(); | 1010 HistoryInfoMapEntry* map_entry = map_item->add_history_info_map_entry(); |
| 1062 map_entry->set_history_id(iter->first); | 1011 map_entry->set_history_id(elem.first); |
| 1063 const history::URLRow& url_row(iter->second.url_row); | 1012 const history::URLRow& url_row(elem.second.url_row); |
| 1064 // Note: We only save information that contributes to the index so there | 1013 // Note: We only save information that contributes to the index so there |
| 1065 // is no need to save search_term_cache_ (not persistent). | 1014 // is no need to save search_term_cache_ (not persistent). |
| 1066 map_entry->set_visit_count(url_row.visit_count()); | 1015 map_entry->set_visit_count(url_row.visit_count()); |
| 1067 map_entry->set_typed_count(url_row.typed_count()); | 1016 map_entry->set_typed_count(url_row.typed_count()); |
| 1068 map_entry->set_last_visit(url_row.last_visit().ToInternalValue()); | 1017 map_entry->set_last_visit(url_row.last_visit().ToInternalValue()); |
| 1069 map_entry->set_url(url_row.url().spec()); | 1018 map_entry->set_url(url_row.url().spec()); |
| 1070 map_entry->set_title(base::UTF16ToUTF8(url_row.title())); | 1019 map_entry->set_title(base::UTF16ToUTF8(url_row.title())); |
| 1071 const VisitInfoVector& visits(iter->second.visits); | 1020 for (const auto& visit : elem.second.visits) { |
| 1072 for (VisitInfoVector::const_iterator visit_iter = visits.begin(); | |
| 1073 visit_iter != visits.end(); ++visit_iter) { | |
| 1074 HistoryInfoMapEntry_VisitInfo* visit_info = map_entry->add_visits(); | 1021 HistoryInfoMapEntry_VisitInfo* visit_info = map_entry->add_visits(); |
| 1075 visit_info->set_visit_time(visit_iter->first.ToInternalValue()); | 1022 visit_info->set_visit_time(visit.first.ToInternalValue()); |
| 1076 visit_info->set_transition_type(visit_iter->second); | 1023 visit_info->set_transition_type(visit.second); |
| 1077 } | 1024 } |
| 1078 } | 1025 } |
| 1079 } | 1026 } |
| 1080 | 1027 |
| 1081 void URLIndexPrivateData::SaveWordStartsMap( | 1028 void URLIndexPrivateData::SaveWordStartsMap( |
| 1082 InMemoryURLIndexCacheItem* cache) const { | 1029 InMemoryURLIndexCacheItem* cache) const { |
| 1083 if (word_starts_map_.empty()) | 1030 if (word_starts_map_.empty()) |
| 1084 return; | 1031 return; |
| 1085 // For unit testing: Enable saving of the cache as an earlier version to | 1032 // For unit testing: Enable saving of the cache as an earlier version to |
| 1086 // allow testing of cache file upgrading in ReadFromFile(). | 1033 // allow testing of cache file upgrading in ReadFromFile(). |
| 1087 // TODO(mrossetti): Instead of intruding on production code with this kind of | 1034 // TODO(mrossetti): Instead of intruding on production code with this kind of |
| 1088 // test harness, save a copy of an older version cache with known results. | 1035 // test harness, save a copy of an older version cache with known results. |
| 1089 // Implement this when switching the caching over to SQLite. | 1036 // Implement this when switching the caching over to SQLite. |
| 1090 if (saved_cache_version_ < 1) | 1037 if (saved_cache_version_ < 1) |
| 1091 return; | 1038 return; |
| 1092 | 1039 |
| 1093 WordStartsMapItem* map_item = cache->mutable_word_starts_map(); | 1040 WordStartsMapItem* map_item = cache->mutable_word_starts_map(); |
| 1094 map_item->set_item_count(word_starts_map_.size()); | 1041 map_item->set_item_count(word_starts_map_.size()); |
| 1095 for (WordStartsMap::const_iterator iter = word_starts_map_.begin(); | 1042 for (const auto& entrie : word_starts_map_) { |
|
Peter Kasting
2017/02/17 01:33:46
Nit: Did you mean "entry"? (many places)
dyaroshev
2017/02/17 20:17:22
Oops, spelling( Done.
| |
| 1096 iter != word_starts_map_.end(); ++iter) { | |
| 1097 WordStartsMapEntry* map_entry = map_item->add_word_starts_map_entry(); | 1043 WordStartsMapEntry* map_entry = map_item->add_word_starts_map_entry(); |
| 1098 map_entry->set_history_id(iter->first); | 1044 map_entry->set_history_id(entrie.first); |
| 1099 const RowWordStarts& word_starts(iter->second); | 1045 const RowWordStarts& word_starts(entrie.second); |
| 1100 for (WordStarts::const_iterator i = word_starts.url_word_starts_.begin(); | 1046 for (auto url_word_start : word_starts.url_word_starts_) |
| 1101 i != word_starts.url_word_starts_.end(); ++i) | 1047 map_entry->add_url_word_starts(url_word_start); |
| 1102 map_entry->add_url_word_starts(*i); | 1048 for (auto title_word_start : word_starts.title_word_starts_) |
| 1103 for (WordStarts::const_iterator i = word_starts.title_word_starts_.begin(); | 1049 map_entry->add_title_word_starts(title_word_start); |
| 1104 i != word_starts.title_word_starts_.end(); ++i) | |
| 1105 map_entry->add_title_word_starts(*i); | |
| 1106 } | 1050 } |
| 1107 } | 1051 } |
| 1108 | 1052 |
| 1109 bool URLIndexPrivateData::RestorePrivateData( | 1053 bool URLIndexPrivateData::RestorePrivateData( |
| 1110 const InMemoryURLIndexCacheItem& cache) { | 1054 const InMemoryURLIndexCacheItem& cache) { |
| 1111 last_time_rebuilt_from_history_ = | 1055 last_time_rebuilt_from_history_ = |
| 1112 base::Time::FromInternalValue(cache.last_rebuild_timestamp()); | 1056 base::Time::FromInternalValue(cache.last_rebuild_timestamp()); |
| 1113 const base::TimeDelta rebuilt_ago = | 1057 const base::TimeDelta rebuilt_ago = |
| 1114 base::Time::Now() - last_time_rebuilt_from_history_; | 1058 base::Time::Now() - last_time_rebuilt_from_history_; |
| 1115 if ((rebuilt_ago > base::TimeDelta::FromDays(7)) || | 1059 if ((rebuilt_ago > base::TimeDelta::FromDays(7)) || |
| (...skipping 22 matching lines...) Expand all Loading... | |
| 1138 bool URLIndexPrivateData::RestoreWordList( | 1082 bool URLIndexPrivateData::RestoreWordList( |
| 1139 const InMemoryURLIndexCacheItem& cache) { | 1083 const InMemoryURLIndexCacheItem& cache) { |
| 1140 if (!cache.has_word_list()) | 1084 if (!cache.has_word_list()) |
| 1141 return false; | 1085 return false; |
| 1142 const WordListItem& list_item(cache.word_list()); | 1086 const WordListItem& list_item(cache.word_list()); |
| 1143 uint32_t expected_item_count = list_item.word_count(); | 1087 uint32_t expected_item_count = list_item.word_count(); |
| 1144 uint32_t actual_item_count = list_item.word_size(); | 1088 uint32_t actual_item_count = list_item.word_size(); |
| 1145 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1089 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1146 return false; | 1090 return false; |
| 1147 const RepeatedPtrField<std::string>& words(list_item.word()); | 1091 const RepeatedPtrField<std::string>& words(list_item.word()); |
| 1148 for (RepeatedPtrField<std::string>::const_iterator iter = words.begin(); | 1092 word_list_.reserve(words.size()); |
| 1149 iter != words.end(); ++iter) | 1093 std::transform( |
| 1150 word_list_.push_back(base::UTF8ToUTF16(*iter)); | 1094 words.begin(), words.end(), std::back_inserter(word_list_), |
| 1095 [](const std::string& word) { return base::UTF8ToUTF16(word); }); | |
| 1151 return true; | 1096 return true; |
| 1152 } | 1097 } |
| 1153 | 1098 |
| 1154 bool URLIndexPrivateData::RestoreWordMap( | 1099 bool URLIndexPrivateData::RestoreWordMap( |
| 1155 const InMemoryURLIndexCacheItem& cache) { | 1100 const InMemoryURLIndexCacheItem& cache) { |
| 1156 if (!cache.has_word_map()) | 1101 if (!cache.has_word_map()) |
| 1157 return false; | 1102 return false; |
| 1158 const WordMapItem& list_item(cache.word_map()); | 1103 const WordMapItem& list_item(cache.word_map()); |
| 1159 uint32_t expected_item_count = list_item.item_count(); | 1104 uint32_t expected_item_count = list_item.item_count(); |
| 1160 uint32_t actual_item_count = list_item.word_map_entry_size(); | 1105 uint32_t actual_item_count = list_item.word_map_entry_size(); |
| 1161 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1106 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1162 return false; | 1107 return false; |
| 1163 const RepeatedPtrField<WordMapEntry>& entries(list_item.word_map_entry()); | 1108 for (const auto& entrie : list_item.word_map_entry()) |
| 1164 for (RepeatedPtrField<WordMapEntry>::const_iterator iter = entries.begin(); | 1109 word_map_[base::UTF8ToUTF16(entrie.word())] = entrie.word_id(); |
| 1165 iter != entries.end(); ++iter) | 1110 |
| 1166 word_map_[base::UTF8ToUTF16(iter->word())] = iter->word_id(); | |
| 1167 return true; | 1111 return true; |
| 1168 } | 1112 } |
| 1169 | 1113 |
| 1170 bool URLIndexPrivateData::RestoreCharWordMap( | 1114 bool URLIndexPrivateData::RestoreCharWordMap( |
| 1171 const InMemoryURLIndexCacheItem& cache) { | 1115 const InMemoryURLIndexCacheItem& cache) { |
| 1172 if (!cache.has_char_word_map()) | 1116 if (!cache.has_char_word_map()) |
| 1173 return false; | 1117 return false; |
| 1174 const CharWordMapItem& list_item(cache.char_word_map()); | 1118 const CharWordMapItem& list_item(cache.char_word_map()); |
| 1175 uint32_t expected_item_count = list_item.item_count(); | 1119 uint32_t expected_item_count = list_item.item_count(); |
| 1176 uint32_t actual_item_count = list_item.char_word_map_entry_size(); | 1120 uint32_t actual_item_count = list_item.char_word_map_entry_size(); |
| 1177 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1121 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1178 return false; | 1122 return false; |
| 1179 const RepeatedPtrField<CharWordMapEntry>& | 1123 |
| 1180 entries(list_item.char_word_map_entry()); | 1124 for (const auto& entrie : list_item.char_word_map_entry()) { |
| 1181 for (RepeatedPtrField<CharWordMapEntry>::const_iterator iter = | 1125 expected_item_count = entrie.item_count(); |
| 1182 entries.begin(); iter != entries.end(); ++iter) { | 1126 actual_item_count = entrie.word_id_size(); |
| 1183 expected_item_count = iter->item_count(); | |
| 1184 actual_item_count = iter->word_id_size(); | |
| 1185 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1127 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1186 return false; | 1128 return false; |
| 1187 base::char16 uni_char = static_cast<base::char16>(iter->char_16()); | 1129 base::char16 uni_char = static_cast<base::char16>(entrie.char_16()); |
| 1188 WordIDSet word_id_set; | 1130 const RepeatedField<int32_t>& word_ids(entrie.word_id()); |
| 1189 const RepeatedField<int32_t>& word_ids(iter->word_id()); | 1131 char_word_map_[uni_char] = {word_ids.begin(), word_ids.end()}; |
| 1190 for (RepeatedField<int32_t>::const_iterator jiter = word_ids.begin(); | |
| 1191 jiter != word_ids.end(); ++jiter) | |
| 1192 word_id_set.insert(*jiter); | |
| 1193 char_word_map_[uni_char] = word_id_set; | |
| 1194 } | 1132 } |
| 1195 return true; | 1133 return true; |
| 1196 } | 1134 } |
| 1197 | 1135 |
| 1198 bool URLIndexPrivateData::RestoreWordIDHistoryMap( | 1136 bool URLIndexPrivateData::RestoreWordIDHistoryMap( |
| 1199 const InMemoryURLIndexCacheItem& cache) { | 1137 const InMemoryURLIndexCacheItem& cache) { |
| 1200 if (!cache.has_word_id_history_map()) | 1138 if (!cache.has_word_id_history_map()) |
| 1201 return false; | 1139 return false; |
| 1202 const WordIDHistoryMapItem& list_item(cache.word_id_history_map()); | 1140 const WordIDHistoryMapItem& list_item(cache.word_id_history_map()); |
| 1203 uint32_t expected_item_count = list_item.item_count(); | 1141 uint32_t expected_item_count = list_item.item_count(); |
| 1204 uint32_t actual_item_count = list_item.word_id_history_map_entry_size(); | 1142 uint32_t actual_item_count = list_item.word_id_history_map_entry_size(); |
| 1205 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1143 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1206 return false; | 1144 return false; |
| 1207 const RepeatedPtrField<WordIDHistoryMapEntry>& | 1145 for (const auto& entrie : list_item.word_id_history_map_entry()) { |
| 1208 entries(list_item.word_id_history_map_entry()); | 1146 expected_item_count = entrie.item_count(); |
| 1209 for (RepeatedPtrField<WordIDHistoryMapEntry>::const_iterator iter = | 1147 actual_item_count = entrie.history_id_size(); |
| 1210 entries.begin(); iter != entries.end(); ++iter) { | |
| 1211 expected_item_count = iter->item_count(); | |
| 1212 actual_item_count = iter->history_id_size(); | |
| 1213 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1148 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1214 return false; | 1149 return false; |
| 1215 WordID word_id = iter->word_id(); | 1150 WordID word_id = entrie.word_id(); |
| 1216 HistoryIDSet history_id_set; | 1151 const RepeatedField<int64_t>& history_ids(entrie.history_id()); |
| 1217 const RepeatedField<int64_t>& history_ids(iter->history_id()); | 1152 word_id_history_map_[word_id] = {history_ids.begin(), history_ids.end()}; |
| 1218 for (RepeatedField<int64_t>::const_iterator jiter = history_ids.begin(); | 1153 for (HistoryID history_id : history_ids) |
| 1219 jiter != history_ids.end(); ++jiter) { | 1154 AddToHistoryIDWordMap(history_id, word_id); |
| 1220 history_id_set.insert(*jiter); | |
| 1221 AddToHistoryIDWordMap(*jiter, word_id); | |
| 1222 } | |
| 1223 word_id_history_map_[word_id] = history_id_set; | |
| 1224 } | 1155 } |
| 1225 return true; | 1156 return true; |
| 1226 } | 1157 } |
| 1227 | 1158 |
| 1228 bool URLIndexPrivateData::RestoreHistoryInfoMap( | 1159 bool URLIndexPrivateData::RestoreHistoryInfoMap( |
| 1229 const InMemoryURLIndexCacheItem& cache) { | 1160 const InMemoryURLIndexCacheItem& cache) { |
| 1230 if (!cache.has_history_info_map()) | 1161 if (!cache.has_history_info_map()) |
| 1231 return false; | 1162 return false; |
| 1232 const HistoryInfoMapItem& list_item(cache.history_info_map()); | 1163 const HistoryInfoMapItem& list_item(cache.history_info_map()); |
| 1233 uint32_t expected_item_count = list_item.item_count(); | 1164 uint32_t expected_item_count = list_item.item_count(); |
| 1234 uint32_t actual_item_count = list_item.history_info_map_entry_size(); | 1165 uint32_t actual_item_count = list_item.history_info_map_entry_size(); |
| 1235 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1166 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1236 return false; | 1167 return false; |
| 1237 const RepeatedPtrField<HistoryInfoMapEntry>& | 1168 |
| 1238 entries(list_item.history_info_map_entry()); | 1169 for (const auto& entrie : list_item.history_info_map_entry()) { |
| 1239 for (RepeatedPtrField<HistoryInfoMapEntry>::const_iterator iter = | 1170 HistoryID history_id = entrie.history_id(); |
| 1240 entries.begin(); iter != entries.end(); ++iter) { | 1171 GURL url(entrie.url()); |
|
Peter Kasting
2017/02/17 01:33:46
Nit: Prefer = to () here
dyaroshev
2017/02/17 20:17:22
Removed the temporary altogether.
| |
| 1241 HistoryID history_id = iter->history_id(); | |
| 1242 GURL url(iter->url()); | |
| 1243 history::URLRow url_row(url, history_id); | 1172 history::URLRow url_row(url, history_id); |
| 1244 url_row.set_visit_count(iter->visit_count()); | 1173 url_row.set_visit_count(entrie.visit_count()); |
| 1245 url_row.set_typed_count(iter->typed_count()); | 1174 url_row.set_typed_count(entrie.typed_count()); |
| 1246 url_row.set_last_visit(base::Time::FromInternalValue(iter->last_visit())); | 1175 url_row.set_last_visit(base::Time::FromInternalValue(entrie.last_visit())); |
| 1247 if (iter->has_title()) { | 1176 if (entrie.has_title()) { |
|
Peter Kasting
2017/02/17 01:33:46
Nit: No {}
dyaroshev
2017/02/17 20:17:22
Done.
| |
| 1248 base::string16 title(base::UTF8ToUTF16(iter->title())); | 1177 url_row.set_title(base::UTF8ToUTF16(entrie.title())); |
| 1249 url_row.set_title(title); | |
| 1250 } | 1178 } |
| 1251 history_info_map_[history_id].url_row = url_row; | 1179 history_info_map_[history_id].url_row = std::move(url_row); |
| 1252 | 1180 |
| 1253 // Restore visits list. | 1181 // Restore visits list. |
| 1254 VisitInfoVector visits; | 1182 VisitInfoVector visits; |
| 1255 visits.reserve(iter->visits_size()); | 1183 visits.reserve(entrie.visits_size()); |
| 1256 for (int i = 0; i < iter->visits_size(); ++i) { | 1184 for (const auto& entrie_visit : entrie.visits()) { |
| 1257 visits.push_back(std::make_pair( | 1185 visits.emplace_back( |
| 1258 base::Time::FromInternalValue(iter->visits(i).visit_time()), | 1186 base::Time::FromInternalValue(entrie_visit.visit_time()), |
| 1259 ui::PageTransitionFromInt(iter->visits(i).transition_type()))); | 1187 ui::PageTransitionFromInt(entrie_visit.transition_type())); |
| 1260 } | 1188 } |
| 1261 history_info_map_[history_id].visits = visits; | 1189 history_info_map_[history_id].visits = std::move(visits); |
| 1262 } | 1190 } |
| 1263 return true; | 1191 return true; |
| 1264 } | 1192 } |
| 1265 | 1193 |
| 1266 bool URLIndexPrivateData::RestoreWordStartsMap( | 1194 bool URLIndexPrivateData::RestoreWordStartsMap( |
| 1267 const InMemoryURLIndexCacheItem& cache) { | 1195 const InMemoryURLIndexCacheItem& cache) { |
| 1268 // Note that this function must be called after RestoreHistoryInfoMap() has | 1196 // Note that this function must be called after RestoreHistoryInfoMap() has |
| 1269 // been run as the word starts may have to be recalculated from the urls and | 1197 // been run as the word starts may have to be recalculated from the urls and |
| 1270 // page titles. | 1198 // page titles. |
| 1271 if (cache.has_word_starts_map()) { | 1199 if (cache.has_word_starts_map()) { |
| 1272 const WordStartsMapItem& list_item(cache.word_starts_map()); | 1200 const WordStartsMapItem& list_item(cache.word_starts_map()); |
| 1273 uint32_t expected_item_count = list_item.item_count(); | 1201 uint32_t expected_item_count = list_item.item_count(); |
| 1274 uint32_t actual_item_count = list_item.word_starts_map_entry_size(); | 1202 uint32_t actual_item_count = list_item.word_starts_map_entry_size(); |
| 1275 if (actual_item_count == 0 || actual_item_count != expected_item_count) | 1203 if (actual_item_count == 0 || actual_item_count != expected_item_count) |
| 1276 return false; | 1204 return false; |
| 1277 const RepeatedPtrField<WordStartsMapEntry>& | 1205 const RepeatedPtrField<WordStartsMapEntry>& |
| 1278 entries(list_item.word_starts_map_entry()); | 1206 entries(list_item.word_starts_map_entry()); |
| 1279 for (RepeatedPtrField<WordStartsMapEntry>::const_iterator iter = | 1207 for (const auto& entrie : entries) { |
| 1280 entries.begin(); iter != entries.end(); ++iter) { | 1208 HistoryID history_id = entrie.history_id(); |
| 1281 HistoryID history_id = iter->history_id(); | |
| 1282 RowWordStarts word_starts; | 1209 RowWordStarts word_starts; |
| 1283 // Restore the URL word starts. | 1210 // Restore the URL word starts. |
| 1284 const RepeatedField<int32_t>& url_starts(iter->url_word_starts()); | 1211 const RepeatedField<int32_t>& url_starts(entrie.url_word_starts()); |
| 1285 for (RepeatedField<int32_t>::const_iterator jiter = url_starts.begin(); | 1212 word_starts.url_word_starts_ = {url_starts.begin(), url_starts.end()}; |
| 1286 jiter != url_starts.end(); ++jiter) | 1213 |
| 1287 word_starts.url_word_starts_.push_back(*jiter); | |
| 1288 // Restore the page title word starts. | 1214 // Restore the page title word starts. |
| 1289 const RepeatedField<int32_t>& title_starts(iter->title_word_starts()); | 1215 const RepeatedField<int32_t>& title_starts(entrie.title_word_starts()); |
| 1290 for (RepeatedField<int32_t>::const_iterator jiter = title_starts.begin(); | 1216 word_starts.title_word_starts_ = {title_starts.begin(), |
| 1291 jiter != title_starts.end(); ++jiter) | 1217 title_starts.end()}; |
| 1292 word_starts.title_word_starts_.push_back(*jiter); | 1218 |
| 1293 word_starts_map_[history_id] = word_starts; | 1219 word_starts_map_[history_id] = std::move(word_starts); |
| 1294 } | 1220 } |
| 1295 } else { | 1221 } else { |
| 1296 // Since the cache did not contain any word starts we must rebuild then from | 1222 // Since the cache did not contain any word starts we must rebuild then from |
| 1297 // the URL and page titles. | 1223 // the URL and page titles. |
| 1298 for (HistoryInfoMap::const_iterator iter = history_info_map_.begin(); | 1224 for (const auto& entrie : history_info_map_) { |
| 1299 iter != history_info_map_.end(); ++iter) { | |
| 1300 RowWordStarts word_starts; | 1225 RowWordStarts word_starts; |
| 1301 const history::URLRow& row(iter->second.url_row); | 1226 const history::URLRow& row(entrie.second.url_row); |
| 1302 const base::string16& url = | 1227 const base::string16& url = |
| 1303 bookmarks::CleanUpUrlForMatching(row.url(), nullptr); | 1228 bookmarks::CleanUpUrlForMatching(row.url(), nullptr); |
| 1304 String16VectorFromString16(url, false, &word_starts.url_word_starts_); | 1229 String16VectorFromString16(url, false, &word_starts.url_word_starts_); |
| 1305 const base::string16& title = | 1230 const base::string16& title = |
| 1306 bookmarks::CleanUpTitleForMatching(row.title()); | 1231 bookmarks::CleanUpTitleForMatching(row.title()); |
| 1307 String16VectorFromString16(title, false, &word_starts.title_word_starts_); | 1232 String16VectorFromString16(title, false, &word_starts.title_word_starts_); |
| 1308 word_starts_map_[iter->first] = word_starts; | 1233 word_starts_map_[entrie.first] = std::move(word_starts); |
| 1309 } | 1234 } |
| 1310 } | 1235 } |
| 1311 return true; | 1236 return true; |
| 1312 } | 1237 } |
| 1313 | 1238 |
| 1314 // static | 1239 // static |
| 1315 bool URLIndexPrivateData::URLSchemeIsWhitelisted( | 1240 bool URLIndexPrivateData::URLSchemeIsWhitelisted( |
| 1316 const GURL& gurl, | 1241 const GURL& gurl, |
| 1317 const std::set<std::string>& whitelist) { | 1242 const std::set<std::string>& whitelist) { |
| 1318 return whitelist.find(gurl.scheme()) != whitelist.end(); | 1243 return whitelist.find(gurl.scheme()) != whitelist.end(); |
| (...skipping 62 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 1381 // First cut: typed count, visit count, recency. | 1306 // First cut: typed count, visit count, recency. |
| 1382 // TODO(mrossetti): This is too simplistic. Consider an approach which ranks | 1307 // TODO(mrossetti): This is too simplistic. Consider an approach which ranks |
| 1383 // recently visited (within the last 12/24 hours) as highly important. Get | 1308 // recently visited (within the last 12/24 hours) as highly important. Get |
| 1384 // input from mpearson. | 1309 // input from mpearson. |
| 1385 if (r1.typed_count() != r2.typed_count()) | 1310 if (r1.typed_count() != r2.typed_count()) |
| 1386 return (r1.typed_count() > r2.typed_count()); | 1311 return (r1.typed_count() > r2.typed_count()); |
| 1387 if (r1.visit_count() != r2.visit_count()) | 1312 if (r1.visit_count() != r2.visit_count()) |
| 1388 return (r1.visit_count() > r2.visit_count()); | 1313 return (r1.visit_count() > r2.visit_count()); |
| 1389 return (r1.last_visit() > r2.last_visit()); | 1314 return (r1.last_visit() > r2.last_visit()); |
| 1390 } | 1315 } |
| OLD | NEW |