Open Source alternative to Algolia + Pinecone and an Easier-to-Use alternative to ElasticSearch ⚡ 🔍 ✨ Fast, typo tolerant, in-memory fuzzy Search Engine for building delightful search experiences
C++ algoliadatastoreelasticsearchenterprise-searchfacetingfull-text-searchfuzzy-searchgeosearchin-memoryinstantsearchmerchandisingpineconesearchsearch-enginesemantic-searchsimilarity-searchsite-searchsynonymstypo-tolerancevector-search
Handle match score not being part of sort fields.
Kishore Nallanc1aace925c75
diff --git a/include/topster.h b/include/topster.hindex 5d43920d..e1af2112 100644--- a/include/topster.h+++ b/include/topster.h@@ -8,7 +8,7 @@struct KV {uint8_t field_id{};- uint8_t match_score_index{};+ int8_t match_score_index{};uint16_t query_index{};uint16_t array_index{};uint32_t token_bits{};diff --git a/src/collection.cpp b/src/collection.cppindex 337f47de..78832b9d 100644--- a/src/collection.cpp+++ b/src/collection.cpp@@ -1382,7 +1382,7 @@ Option<nlohmann::json> Collection::search(const std::string & raw_query,if(field_order_kv->match_score_index == CURATED_RECORD_IDENTIFIER) {wrapper_doc["curated"] = true;- } else {+ } else if(field_order_kv->match_score_index >= 0) {wrapper_doc["text_match"] = field_order_kv->scores[field_order_kv->match_score_index];wrapper_doc["text_match_info"] = nlohmann::json::object();diff --git a/src/index.cpp b/src/index.cppindex 5f9d8eaf..009ce89a 100644--- a/src/index.cpp+++ b/src/index.cpp@@ -3446,7 +3446,7 @@ void Index::compute_sort_scores(const std::vector<sort_by>& sort_fields, const ifilter_index = found_index;}- scores[0] = int64_t(found);+ scores[2] = int64_t(found);} else {auto it = field_values[2]->find(seq_id);scores[2] = (it == field_values[2]->end()) ? default_score : it->second;@@ -3664,7 +3664,7 @@ void Index::do_infix_search(const size_t num_search_fields, const std::vector<se0, match_score, seq_id, sort_order, false, false, false, 1, -1, {});int64_t scores[3] = {0};- int64_t match_score_index = 0;+ int64_t match_score_index = -1;compute_sort_scores(sort_fields, sort_order, field_values, geopoint_indices, seq_id, filter_index,100, scores, match_score_index);@@ -3995,7 +3995,7 @@ void Index::search_wildcard(const std::vector<filter>& filters,match_score, seq_id, sort_order, false, false, false, 1, -1, plists);int64_t scores[3] = {0};- int64_t match_score_index = 0;+ int64_t match_score_index = -1;compute_sort_scores(sort_fields, sort_order, field_values, geopoint_indices, seq_id, filter_index,100, scores, match_score_index);diff --git a/test/collection_sorting_test.cpp b/test/collection_sorting_test.cppindex 80755c93..43c3e393 100644--- a/test/collection_sorting_test.cpp+++ b/test/collection_sorting_test.cpp@@ -1737,6 +1737,37 @@ TEST_F(CollectionSortingTest, RepeatingTokenRanking) {collectionManager.drop_collection("coll1");}+TEST_F(CollectionSortingTest, SortingDoesNotHaveTextMatchComponent) {+ // text_match_score field should not be present in response+ std::vector<field> fields = {field("title", field_types::STRING, false),+ field("points", field_types::INT32, false),};++ Collection* coll1 = collectionManager.create_collection("coll1", 1, fields, "points").get();++ nlohmann::json doc1;+ doc1["id"] = "0";+ doc1["title"] = "Test Title";+ doc1["points"] = 100;++ ASSERT_TRUE(coll1->add(doc1.dump()).ok());++ sort_fields = {+ sort_by("points", "DESC"),+ sort_by("points", "DESC"),+ sort_by("points", "DESC"),+ };++ auto results = coll1->search("test", {"title"}, "", {}, sort_fields, {2}, 10, 1, FREQUENCY, {true}).get();+ ASSERT_EQ(1, results["hits"].size());+ ASSERT_EQ(0, results["hits"][0].count("text_match"));++ results = coll1->search("*", {}, "", {}, sort_fields, {2}, 10, 1, FREQUENCY, {true}).get();+ ASSERT_EQ(1, results["hits"].size());+ ASSERT_EQ(0, results["hits"][0].count("text_match"));++ collectionManager.drop_collection("coll1");+}+TEST_F(CollectionSortingTest, IntegerFloatAndBoolShouldDefaultSortTrue) {std::string coll_schema = R"({