Skip to content

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

adding found documents in results when group_by is enabled

krunal13130dffb36777d8
diff --git a/src/collection.cpp b/src/collection.cpp
index 15559c85..ead0c0bd 100644
--- a/src/collection.cpp
+++ b/src/collection.cpp
@@ -1384,7 +1384,7 @@ Option<nlohmann::json> Collection::search(std::string raw_query,
std::vector<std::vector<KV*>> raw_result_kvs;
std::vector<std::vector<KV*>> override_result_kvs;
- size_t total_found = 0;
+ size_t total_found = 0, total_groups_found = 0;
std::vector<uint32_t> excluded_ids;
std::vector<std::pair<uint32_t, uint32_t>> included_ids; // ID -> position
@@ -1557,10 +1557,10 @@ Option<nlohmann::json> Collection::search(std::string raw_query,
// for grouping we have to aggregate group set sizes to a count value
if(group_limit) {
- total_found = search_params->groups_processed.size() + override_result_kvs.size();
- } else {
- total_found = search_params->all_result_ids_len;
- }
+ total_groups_found = search_params->groups_processed.size() + override_result_kvs.size();
+ }
+
+ total_found = search_params->all_result_ids_len;
if(search_cutoff && total_found == 0) {
// this can happen if other requests stopped this request from being processed
@@ -1683,6 +1683,9 @@ Option<nlohmann::json> Collection::search(std::string raw_query,
}
nlohmann::json result = nlohmann::json::object();
+ if(group_limit != 0) {
+ result["found_groups"] = total_groups_found;
+ }
result["found"] = total_found;
if(exclude_fields.count("out_of") == 0) {
diff --git a/test/collection_grouping_test.cpp b/test/collection_grouping_test.cpp
index 9507a91c..8991bbbd 100644
--- a/test/collection_grouping_test.cpp
+++ b/test/collection_grouping_test.cpp
@@ -69,7 +69,8 @@ TEST_F(CollectionGroupingTest, GroupingBasics) {
"", 10,
{}, {}, {"size"}, 2).get();
- ASSERT_EQ(3, res["found"].get<size_t>());
+ ASSERT_EQ(12, res["found"].get<size_t>());
+ ASSERT_EQ(3, res["found_groups"].get<size_t>());
ASSERT_EQ(3, res["grouped_hits"].size());
ASSERT_EQ(11, res["grouped_hits"][0]["group_key"][0].get<size_t>());
@@ -116,7 +117,8 @@ TEST_F(CollectionGroupingTest, GroupingBasics) {
{}, {}, {"rating"}, 2).get();
// 7 unique ratings
- ASSERT_EQ(7, res["found"].get<size_t>());
+ ASSERT_EQ(12, res["found"].get<size_t>());
+ ASSERT_EQ(7, res["found_groups"].get<size_t>());
ASSERT_EQ(7, res["grouped_hits"].size());
ASSERT_FLOAT_EQ(4.4, res["grouped_hits"][0]["group_key"][0].get<float>());
@@ -167,8 +169,8 @@ TEST_F(CollectionGroupingTest, GroupingCompoundKey) {
spp::sparse_hash_set<std::string>(), 10, "", 30, 5,
"", 10,
{}, {}, {"size", "brand"}, 2).get();
-
- ASSERT_EQ(10, res["found"].get<size_t>());
+ ASSERT_EQ(12, res["found"].get<size_t>());
+ ASSERT_EQ(10, res["found_groups"].get<size_t>());
ASSERT_EQ(10, res["grouped_hits"].size());
ASSERT_EQ(1, res["grouped_hits"][0]["found"].get<int32_t>());
@@ -227,7 +229,8 @@ TEST_F(CollectionGroupingTest, GroupingCompoundKey) {
ASSERT_STREQ("0", res["grouped_hits"][0]["hits"][1]["document"]["id"].get<std::string>().c_str());
// total count and facet counts should be the same
- ASSERT_EQ(10, res["found"].get<size_t>());
+ ASSERT_EQ(12, res["found"].get<size_t>());
+ ASSERT_EQ(10, res["found_groups"].get<size_t>());
ASSERT_EQ(2, res["grouped_hits"].size());
ASSERT_EQ(10, res["grouped_hits"][0]["group_key"][0].get<size_t>());
ASSERT_STREQ("Omega", res["grouped_hits"][0]["group_key"][1].get<std::string>().c_str());
@@ -313,7 +316,8 @@ TEST_F(CollectionGroupingTest, GroupingWithMultiFieldRelevance) {
"", 10,
{}, {}, {"genre"}, 2).get();
- ASSERT_EQ(3, results["found"].get<size_t>());
+ ASSERT_EQ(7, results["found"].get<size_t>());
+ ASSERT_EQ(3, results["found_groups"].get<size_t>());
ASSERT_EQ(3, results["grouped_hits"].size());
ASSERT_EQ(3, results["grouped_hits"][0]["found"].get<int32_t>());
@@ -345,7 +349,8 @@ TEST_F(CollectionGroupingTest, GroupingWithGropLimitOfOne) {
"", 10,
{}, {}, {"brand"}, 1).get();
- ASSERT_EQ(5, res["found"].get<size_t>());
+ ASSERT_EQ(12, res["found"].get<size_t>());
+ ASSERT_EQ(5, res["found_groups"].get<size_t>());
ASSERT_EQ(5, res["grouped_hits"].size());
// all hits array must be of size 1
@@ -430,7 +435,8 @@ TEST_F(CollectionGroupingTest, GroupingWithArrayFieldAndOverride) {
"", 10,
{}, {}, {"colors"}, 2).get();
- ASSERT_EQ(4, res["found"].get<size_t>());
+ ASSERT_EQ(9, res["found"].get<size_t>());
+ ASSERT_EQ(4, res["found_groups"].get<size_t>());
ASSERT_EQ(4, res["grouped_hits"].size());
ASSERT_EQ(1, res["grouped_hits"][0]["group_key"][0].size());
@@ -611,7 +617,8 @@ TEST_F(CollectionGroupingTest, SortingOnGroupCount) {
"", 10,
{}, {}, {"size"}, 2).get();
- ASSERT_EQ(3, res["found"].get<size_t>());
+ ASSERT_EQ(12, res["found"].get<size_t>());
+ ASSERT_EQ(3, res["found_groups"].get<size_t>());
ASSERT_EQ(3, res["grouped_hits"].size());
ASSERT_EQ(10, res["grouped_hits"][0]["group_key"][0].get<size_t>());
@@ -635,7 +642,8 @@ TEST_F(CollectionGroupingTest, SortingOnGroupCount) {
"", 10,
{}, {}, {"size"}, 2).get();
- ASSERT_EQ(3, res2["found"].get<size_t>());
+ ASSERT_EQ(12, res2["found"].get<size_t>());
+ ASSERT_EQ(3, res2["found_groups"].get<size_t>());
ASSERT_EQ(3, res2["grouped_hits"].size());
ASSERT_EQ(11, res2["grouped_hits"][0]["group_key"][0].get<size_t>());
@@ -715,7 +723,8 @@ TEST_F(CollectionGroupingTest, SortingMoreThanMaxTopsterSize) {
"", 10,
{}, {}, {"size"}, 2).get();
- ASSERT_EQ(300, res["found"].get<size_t>());
+ ASSERT_EQ(1000, res["found"].get<size_t>());
+ ASSERT_EQ(300, res["found_groups"].get<size_t>());
ASSERT_EQ(100, res["grouped_hits"].size());
ASSERT_EQ(4, res["grouped_hits"][4]["found"].get<int32_t>());
@@ -734,7 +743,8 @@ TEST_F(CollectionGroupingTest, SortingMoreThanMaxTopsterSize) {
"", 10,
{}, {}, {"size"}, 2).get();
- ASSERT_EQ(300, res["found"].get<size_t>());
+ ASSERT_EQ(1000, res["found"].get<size_t>());
+ ASSERT_EQ(300, res["found_groups"].get<size_t>());
ASSERT_EQ(100, res["grouped_hits"].size());
ASSERT_EQ(3, res["grouped_hits"][4]["found"].get<int32_t>());
@@ -757,7 +767,8 @@ TEST_F(CollectionGroupingTest, SortingMoreThanMaxTopsterSize) {
"", 10,
{}, {}, {"size"}, 2).get();
- ASSERT_EQ(300, res2["found"].get<size_t>());
+ ASSERT_EQ(1000, res2["found"].get<size_t>());
+ ASSERT_EQ(300, res2["found_groups"].get<size_t>());
ASSERT_EQ(100, res2["grouped_hits"].size());
ASSERT_EQ(2, res2["grouped_hits"][0]["found"].get<int32_t>());
@@ -775,7 +786,8 @@ TEST_F(CollectionGroupingTest, SortingMoreThanMaxTopsterSize) {
"", 10,
{}, {}, {"size"}, 2).get();
- ASSERT_EQ(300, res2["found"].get<size_t>());
+ ASSERT_EQ(1000, res2["found"].get<size_t>());
+ ASSERT_EQ(300, res2["found_groups"].get<size_t>());
ASSERT_EQ(100, res2["grouped_hits"].size());
ASSERT_EQ(3, res2["grouped_hits"][0]["found"].get<int32_t>());