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

Handle repeating IDs for a numerical value during filtering.

kishorencf1d7d323b8fa
diff --git a/include/num_tree.h b/include/num_tree.h
index 7b0449c6..8f42c1cd 100644
--- a/include/num_tree.h
+++ b/include/num_tree.h
@@ -22,7 +22,9 @@ public:
int64map.emplace(value, new sorted_array);
}
- int64map[value]->append(id);
+ if(!int64map[value]->contains(id)) {
+ int64map[value]->append(id);
+ }
}
void range_inclusive_search(int64_t start, int64_t end, uint32_t** ids, size_t& ids_len) {
@@ -104,6 +106,7 @@ public:
}
std::sort(consolidated_ids.begin(), consolidated_ids.end());
+ consolidated_ids.erase(unique(consolidated_ids.begin(), consolidated_ids.end()), consolidated_ids.end());
uint32_t *out = nullptr;
ids_len = ArrayUtils::or_scalar(&consolidated_ids[0], consolidated_ids.size(),
@@ -134,6 +137,7 @@ public:
}
std::sort(consolidated_ids.begin(), consolidated_ids.end());
+ consolidated_ids.erase(unique(consolidated_ids.begin(), consolidated_ids.end()), consolidated_ids.end());
uint32_t *out = nullptr;
ids_len = ArrayUtils::or_scalar(&consolidated_ids[0], consolidated_ids.size(),
diff --git a/test/collection_filtering_test.cpp b/test/collection_filtering_test.cpp
index 82704872..2682c1aa 100644
--- a/test/collection_filtering_test.cpp
+++ b/test/collection_filtering_test.cpp
@@ -1001,5 +1001,80 @@ TEST_F(CollectionFilteringTest, NumericalFilteringWithAnd) {
ASSERT_EQ(0, results["found"].get<size_t>());
+ collectionManager.drop_collection("coll1");
+}
+
+TEST_F(CollectionFilteringTest, NumericalFilteringWithArray) {
+ Collection *coll1;
+
+ std::vector<field> fields = {field("title", field_types::STRING, false),
+ field("prices", field_types::INT32_ARRAY, false),};
+
+ coll1 = collectionManager.get_collection("coll1").get();
+ if (coll1 == nullptr) {
+ coll1 = collectionManager.create_collection("coll1", 1, fields, "").get();
+ }
+
+ std::vector<std::vector<std::string>> records = {
+ {"1", "T Shirt 1", "1", "2", "3"},
+ {"2", "T Shirt 2", "1", "2", "3"},
+ {"3", "T Shirt 3", "1", "2", "3"},
+ {"4", "T Shirt 4", "1", "1", "1"},
+ };
+
+ for (size_t i = 0; i < records.size(); i++) {
+ nlohmann::json doc;
+
+ doc["id"] = records[i][0];
+ doc["title"] = records[i][1];
+
+ std::vector<int32_t> prices;
+ for(size_t j = 2; j <= 4; j++) {
+ prices.push_back(std::stoi(records[i][j]));
+ }
+
+ doc["prices"] = prices;
+
+ ASSERT_TRUE(coll1->add(doc.dump()).ok());
+ }
+
+ // check equals on a repeating price
+ auto results = coll1->search("*",
+ {}, "prices:1",
+ {}, {}, 0, 10, 1, FREQUENCY, true).get();
+
+ ASSERT_EQ(4, results["found"].get<size_t>());
+ ASSERT_EQ(4, results["hits"].size());
+
+ // check ranges
+
+ results = coll1->search("*",
+ {}, "prices:>=1",
+ {}, {}, 0, 10, 1, FREQUENCY, true).get();
+
+ ASSERT_EQ(4, results["found"].get<size_t>());
+ ASSERT_EQ(4, results["hits"].size());
+
+ results = coll1->search("*",
+ {}, "prices:>=2",
+ {}, {}, 0, 10, 1, FREQUENCY, true).get();
+
+ ASSERT_EQ(3, results["found"].get<size_t>());
+ ASSERT_EQ(3, results["hits"].size());
+
+ results = coll1->search("*",
+ {}, "prices:<4",
+ {}, {}, 0, 10, 1, FREQUENCY, true).get();
+
+ ASSERT_EQ(4, results["found"].get<size_t>());
+ ASSERT_EQ(4, results["hits"].size());
+
+ results = coll1->search("*",
+ {}, "prices:<=2",
+ {}, {}, 0, 10, 1, FREQUENCY, true).get();
+
+ ASSERT_EQ(4, results["found"].get<size_t>());
+ ASSERT_EQ(4, results["hits"].size());
+
collectionManager.drop_collection("coll1");
}
\ No newline at end of file