Add refactor.mergeNodes procedure

cpp/refactor_module/algorithm/refactor.cpp

@@ -863,3 +863,129 @@ void Refactor::RenameType(mgp_list *args, mgp_graph *memgraph_graph, mgp_result
     return;
   }
 }
+
+void Refactor::MergeNodes(mgp_list *args, mgp_graph *memgraph_graph, mgp_result *result, mgp_memory *memory) {
+  mgp::MemoryDispatcherGuard guard{memory};
+  const auto arguments = mgp::List(args);
+  auto graph = mgp::Graph(memgraph_graph);
+  const auto record_factory = mgp::RecordFactory(result);
+
+  try {
+    const auto nodes = arguments[0].ValueList();
+    const auto config = arguments[1].ValueMap();
+
+    if (nodes.Empty()) {
+      throw mgp::ValueException(std::string(kMergeNodesEmptyListError).c_str());
+    }
+
+    // Verify all elements are nodes
+    for (const auto &node_value : nodes) {
+      if (!node_value.IsNode()) {
+        throw mgp::ValueException(std::string(kMergeNodesInvalidTypeError).c_str());
+      }
+    }
+
+    bool mergeRels = false;
+    if (config.KeyExists(kMergeNodesRelationshipsStrategy)) {
+      if (!config.At(kMergeNodesRelationshipsStrategy).IsBool()) {
+        throw mgp::ValueException(std::string(kMergeRelationshipsInvalidValueError).c_str());
+      }
+
+      mergeRels = config.At(kMergeNodesRelationshipsStrategy).ValueBool();
+    }
+
+    // Get the first node as the target node
+    auto target_node = nodes[0].ValueNode();
+
+    // Process remaining nodes
+    for (size_t i = 1; i < nodes.Size(); ++i) {
+      auto source_node = nodes[i].ValueNode();
+
+      // Get properties strategy from config
+      std::string prop_strategy = std::string(kMergeNodesPropertiesCombine);
+      if (config.KeyExists(kMergeNodesPropertiesStrategy)) {
+        prop_strategy = config.At(kMergeNodesPropertiesStrategy).ValueString();
+      }
+      if (config.KeyExists(kMergeNodesPropertiesStrategyAlternative)) {
+        prop_strategy = config.At(kMergeNodesPropertiesStrategyAlternative).ValueString();
+      }
+
+      // Convert to lowercase for case-insensitive comparison
+      std::transform(prop_strategy.begin(), prop_strategy.end(), prop_strategy.begin(), ::tolower);
+      // Validate property strategy
+      if (prop_strategy != std::string(kMergeNodesPropertiesCombine) &&
+          prop_strategy != std::string(kMergeNodesPropertiesDiscard) &&
+          prop_strategy != std::string(kMergeNodesPropertiesOverride) &&
+          prop_strategy != std::string(kMergeNodesPropertiesOverwrite)) {
+        throw mgp::ValueException(std::string(kMergeNodesInvalidPropertyStrategyError).c_str());
+      }
+
+      // Handle properties based on strategy
+      if (prop_strategy == std::string(kMergeNodesPropertiesCombine)) {
+        // Combine properties from both nodes
+        auto source_props = source_node.Properties();
+        for (const auto &[key, value] : source_props) {
+          if (!source_props.contains(key)) {
+            // nothing to combine
+            continue;
+          }
+
+          auto target_property = target_node.GetProperty(key);
+          if (target_property.IsList()) {
+            auto target_list = target_property.ValueList();
+            target_list.AppendExtend(source_props[key]);
+            target_node.SetProperty(key, mgp::Value(std::move(target_list)));
+          } else if (!target_property.IsNull()) {
+            auto combined_properties = mgp::List();
+            combined_properties.AppendExtend(target_property);
+            combined_properties.AppendExtend(source_props[key]);
+            target_node.SetProperty(key, mgp::Value(std::move(combined_properties)));
+          } else {
+            target_node.SetProperty(key, source_props[key]);
+          }
+        }
+      } else if (prop_strategy == std::string(kMergeNodesPropertiesDiscard)) {
+        // Keep only target node properties
+        // No action needed
+      } else if (prop_strategy == std::string(kMergeNodesPropertiesOverride) ||
+                 prop_strategy == std::string(kMergeNodesPropertiesOverwrite)) {
+        // Override/overwrite target properties with source properties
+        auto source_props = source_node.Properties();
+        for (const auto &[key, value] : source_props) {
+          target_node.SetProperty(key, value);
+        }
+      }
+
+      // Copy labels from source to target
+      auto source_labels = source_node.Labels();
+      for (size_t j = 0; j < source_labels.Size(); ++j) {
+        target_node.AddLabel(source_labels[j]);
+      }
+
+      // Handle relationships
+      // Copy all relationships from source to target
+      if (mergeRels) {
+        auto in_rels = source_node.InRelationships();
+        for (const auto &rel : in_rels) {
+          graph.CreateRelationship(rel.From(), target_node, rel.Type());
+        }
+
+        auto out_rels = source_node.OutRelationships();
+        for (const auto &rel : out_rels) {
+          graph.CreateRelationship(target_node, rel.To(), rel.Type());
+        }
+      }
+
+      // Delete the source node
+      graph.DetachDeleteNode(source_node);
+    }
+
+    // Return the merged node
+    auto record = record_factory.NewRecord();
+    record.Insert(std::string(kMergeNodesResult).c_str(), target_node);
+
+  } catch (const std::exception &e) {
+    record_factory.SetErrorMessage(e.what());
+    return;
+  }
+}

cpp/refactor_module/algorithm/refactor.hpp

@@ -129,6 +129,23 @@ constexpr std::string_view kRenameTypePropertyArg2 = "new_property";
 constexpr std::string_view kRenameTypePropertyArg3 = "rels";
 constexpr std::string_view kRenameTypePropertyResult = "relationships_changed";
 
+/* merge_nodes constants */
+constexpr std::string_view kProcedureMergeNodes = "merge_nodes";
+constexpr std::string_view kMergeNodesArgNodes = "nodes";
+constexpr std::string_view kMergeNodesArgConfig = "config";
+constexpr std::string_view kMergeNodesResult = "node";
+constexpr std::string_view kMergeNodesPropertiesStrategy = "properties";
+constexpr std::string_view kMergeNodesPropertiesStrategyAlternative = ".*";
+constexpr std::string_view kMergeNodesRelationshipsStrategy = "mergeRels";
+constexpr std::string_view kMergeNodesPropertiesCombine = "combine";
+constexpr std::string_view kMergeNodesPropertiesDiscard = "discard";
+constexpr std::string_view kMergeNodesPropertiesOverride = "override";
+constexpr std::string_view kMergeNodesPropertiesOverwrite = "overwrite";
+constexpr std::string_view kMergeNodesEmptyListError = "The nodes list cannot be empty";
+constexpr std::string_view kMergeNodesInvalidTypeError = "All elements in the nodes list must be of type Node";
+constexpr std::string_view kMergeRelationshipsInvalidValueError = "mergeRels parameter must either be true or false";
+constexpr std::string_view kMergeNodesInvalidPropertyStrategyError = "Property strategy must be one of: combine, discard, override/overwrite";
+
 void From(mgp_list *args, mgp_graph *memgraph_graph, mgp_result *result, mgp_memory *memory);
 
 void To(mgp_list *args, mgp_graph *memgraph_graph, mgp_result *result, mgp_memory *memory);
@@ -171,4 +188,6 @@ void NormalizeAsBoolean(mgp_list *args, mgp_graph *memgraph_graph, mgp_result *r
 void ExtractNode(mgp_list *args, mgp_graph *memgraph_graph, mgp_result *result, mgp_memory *memory);
 
 void RenameType(mgp_list *args, mgp_graph *memgraph_graph, mgp_result *result, mgp_memory *memory);
+
+void MergeNodes(mgp_list *args, mgp_graph *memgraph_graph, mgp_result *result, mgp_memory *memory);
 }  // namespace Refactor

cpp/refactor_module/refactor_module.cpp

@@ -124,6 +124,11 @@ extern "C" int mgp_init_module(struct mgp_module *module, struct mgp_memory *mem
                  {mgp::Return(std::string(Refactor::kRenameTypePropertyResult).c_str(), mgp::Type::Int)}, module,
                  memory);
 
+    AddProcedure(Refactor::MergeNodes, Refactor::kProcedureMergeNodes, mgp::ProcedureType::Write,
+                 {mgp::Parameter(Refactor::kMergeNodesArgNodes, {mgp::Type::List, mgp::Type::Node}),
+                  mgp::Parameter(Refactor::kMergeNodesArgConfig, mgp::Type::Map, mgp::Value(mgp::Map{}))},
+                 {mgp::Return(Refactor::kMergeNodesResult, mgp::Type::Node)}, module, memory);
+
   } catch (const std::exception &e) {
     return 1;
   }

e2e/refactor_test/test_merge_nodes_asterisk/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}) CREATE (n3:Person {name: 'Charlie', age: 35, city: 'London'}) CREATE (n1)-[:KNOWS {since: 2020}]->(n2) CREATE (n2)-[:WORKS_WITH {project: 'Project X'}]->(n3) CREATE (n3)-[:FRIENDS_WITH {since: 2019}]->(n1); 

e2e/refactor_test/test_merge_nodes_asterisk/test.yml

@@ -0,0 +1,8 @@
+query: >
+  MATCH (n1:Person {age: 30}), (n2:Person {name: 'Bob'}), (n3:Person {name: 'Charlie'})
+  CALL refactor.merge_nodes([n1, n2, n3], {`.*`: 'combine', mergeRels: true}) YIELD node
+  RETURN node.name as name, outDegree(node) as out_degree, inDegree(node) as in_degree;
+output:
+  - name: [Bob, Charlie]
+    out_degree: 3
+    in_degree: 3

e2e/refactor_test/test_merge_nodes_case_insensitive/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {name: 'Alice', age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}); 

e2e/refactor_test/test_merge_nodes_case_insensitive/test.yml

@@ -0,0 +1,9 @@
+query: >
+  MATCH (n1:Person {name: 'Alice'}), (n2:Person {name: 'Bob'})
+  CALL refactor.merge_nodes([n1, n2], {properties: 'COMBINE'}) YIELD node
+  RETURN node.name as name, node.age as age, node.city as city, node.country as country;
+output:
+  - name: [Alice, Bob]
+    age: [30, 25]
+    city: New York
+    country: USA 

e2e/refactor_test/test_merge_nodes_combine_basic/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {name: 'Alice', age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}); 

e2e/refactor_test/test_merge_nodes_combine_basic/test.yml

@@ -0,0 +1,9 @@
+query: >
+  MATCH (n1:Person {name: 'Alice'}), (n2:Person {name: 'Bob'})
+  CALL refactor.merge_nodes([n1, n2], {properties: 'combine'}) YIELD node
+  RETURN node.name as name, node.age as age, node.city as city, node.country as country;
+output: 
+  - name: [Alice, Bob]
+    age: [30, 25]
+    city: New York
+    country: USA

e2e/refactor_test/test_merge_nodes_combine_basic_2/input.cyp

@@ -0,0 +1 @@
+CREATE (n4:Employee {name: 'David', salary: 50000}) CREATE (n5:Manager {name: 'Eve', department: 'IT'}); 

e2e/refactor_test/test_merge_nodes_combine_basic_2/test.yml

@@ -0,0 +1,9 @@
+query: >
+  MATCH (n4:Employee {name: 'David'}), (n5:Manager {name: 'Eve'})
+  CALL refactor.merge_nodes([n4, n5], {properties: 'combine'}) YIELD node
+  RETURN node.name as name, node.salary as salary, node.department as department, labels(node) as labels;
+output:
+- name: [David, Eve]
+  salary: 50000
+  department: IT
+  labels: [Employee, Manager] 

e2e/refactor_test/test_merge_nodes_combine_merge/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}) CREATE (n3:Person {name: 'Charlie', age: 35, city: 'London'}) CREATE (n1)-[:KNOWS {since: 2020}]->(n2) CREATE (n2)-[:WORKS_WITH {project: 'Project X'}]->(n3) CREATE (n3)-[:FRIENDS_WITH {since: 2019}]->(n1); 

e2e/refactor_test/test_merge_nodes_combine_merge/test.yml

@@ -0,0 +1,8 @@
+query: >
+  MATCH (n1:Person {age: 30}), (n2:Person {name: 'Bob'}), (n3:Person {name: 'Charlie'})
+  CALL refactor.merge_nodes([n1, n2, n3], {properties: 'combine', mergeRels: true}) YIELD node
+  RETURN node.name as name, outDegree(node) as out_degree, inDegree(node) as in_degree;
+output:
+  - name: [Bob, Charlie]
+    out_degree: 3
+    in_degree: 3

e2e/refactor_test/test_merge_nodes_discard_basic/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {name: 'Alice', age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}); 

e2e/refactor_test/test_merge_nodes_discard_basic/test.yml

@@ -0,0 +1,9 @@
+query: >
+  MATCH (n1:Person {name: 'Alice'}), (n2:Person {name: 'Bob'})
+  CALL refactor.merge_nodes([n1, n2], {properties: 'discard'}) YIELD node
+  RETURN node.name as name, node.age as age, node.city as city, node.country as country;
+output:
+  - name: Alice
+    age: 30
+    city: New York
+    country: null 

e2e/refactor_test/test_merge_nodes_invalid_empty_nodes/test.yml

@@ -0,0 +1,5 @@
+query: >
+  CALL refactor.merge_nodes([], {properties: 'combine'}) YIELD node
+  RETURN node;
+exception: >-
+  "The nodes list cannot be empty"

e2e/refactor_test/test_merge_nodes_invalid_merge_rels_value/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}) CREATE (n3:Person {name: 'Charlie', age: 35, city: 'London'}) CREATE (n1)-[:KNOWS {since: 2020}]->(n2) CREATE (n2)-[:WORKS_WITH {project: 'Project X'}]->(n3) CREATE (n3)-[:FRIENDS_WITH {since: 2019}]->(n1); 

e2e/refactor_test/test_merge_nodes_invalid_merge_rels_value/test.yml

@@ -0,0 +1,7 @@
+query: >
+  MATCH (n1:Person {age: 30}), (n2:Person {name: 'Bob'}), (n3:Person {name: 'Charlie'})
+  CALL refactor.merge_nodes([n1, n2, n3], {`.*`: 'combine', mergeRels: "nothing"}) YIELD node
+  RETURN node.name as name, outDegree(node) as out_degree, inDegree(node) as in_degree;
+exception: >
+  mergeRels parameter must either be true or false
+

e2e/refactor_test/test_merge_nodes_merge_false/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}) CREATE (n3:Person {name: 'Charlie', age: 35, city: 'London'}) CREATE (n1)-[:KNOWS {since: 2020}]->(n2) CREATE (n2)-[:WORKS_WITH {project: 'Project X'}]->(n3) CREATE (n3)-[:FRIENDS_WITH {since: 2019}]->(n1); 

e2e/refactor_test/test_merge_nodes_merge_false/test.yml

@@ -0,0 +1,8 @@
+query: >
+  MATCH (n1:Person {age: 30}), (n2:Person {name: 'Bob'}), (n3:Person {name: 'Charlie'})
+  CALL refactor.merge_nodes([n1, n2, n3], {`.*`: 'combine', mergeRels: false}) YIELD node
+  RETURN node.name as name, outDegree(node) as out_degree, inDegree(node) as in_degree;
+output:
+  - name: [Bob, Charlie]
+    out_degree: 0
+    in_degree: 0

e2e/refactor_test/test_merge_nodes_merge_rels/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {name: 'Alice', age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}) CREATE (n3:Person {name: 'Charlie', age: 35, city: 'London'}) CREATE (n1)-[:KNOWS {since: 2020}]->(n2) CREATE (n2)-[:WORKS_WITH {project: 'Project X'}]->(n3) CREATE (n3)-[:FRIENDS_WITH {since: 2019}]->(n1); 

e2e/refactor_test/test_merge_nodes_merge_rels/test.yml

@@ -0,0 +1,8 @@
+query: >
+  MATCH (n1:Person {name: 'Alice'}), (n2:Person {name: 'Bob'}), (n3:Person {name: 'Charlie'})
+  CALL refactor.merge_nodes([n1, n2, n3], {properties: 'combine', mergeRels: true}) YIELD node
+  RETURN node.name as name, outDegree(node) as out_degree, inDegree(node) as in_degree;
+output:
+  - name: [Alice, Bob, Charlie]
+    out_degree: 3
+    in_degree: 3

e2e/refactor_test/test_merge_nodes_override_basic/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {name: 'Alice', age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'});

e2e/refactor_test/test_merge_nodes_override_basic/test.yml

@@ -0,0 +1,9 @@
+query: >
+  MATCH (n1:Person {name: 'Alice'}), (n2:Person {name: 'Bob'})
+  CALL refactor.merge_nodes([n1, n2], {properties: 'override'}) YIELD node
+  RETURN node.name as name, node.age as age, node.city as city, node.country as country;
+output:
+  - name: Bob
+    age: 25
+    city: New York
+    country: USA

e2e/refactor_test/test_merge_nodes_overwrite_basic/input.cyp

@@ -0,0 +1 @@
+CREATE (n1:Person {name: 'Alice', age: 30, city: 'New York'}) CREATE (n2:Person {name: 'Bob', age: 25, country: 'USA'}); 

e2e/refactor_test/test_merge_nodes_overwrite_basic/test.yml

@@ -0,0 +1,9 @@
+query: >
+  MATCH (n1:Person {name: 'Alice'}), (n2:Person {name: 'Bob'})
+  CALL refactor.merge_nodes([n1, n2], {properties: 'overwrite'}) YIELD node
+  RETURN node.name as name, node.age as age, node.city as city, node.country as country;
+output:
+  - name: Bob
+    age: 25
+    city: New York
+    country: USA