Add MERGE ON CREATE SET / ON MATCH SET support (#1619)

.gitignore

@@ -15,3 +15,4 @@ __pycache__
 **/apache_age_python.egg-info
 
 drivers/python/build
+*.bc

regress/expected/cypher_merge.out

@@ -2001,9 +2001,258 @@ SELECT * FROM cypher('issue_1954', $$ MATCH (n) DETACH DELETE n $$) AS (a agtype
 ---
 (0 rows)
 
+--
+-- ON CREATE SET / ON MATCH SET tests (issue #1619)
+--
+SELECT create_graph('merge_actions');
+NOTICE:  graph "merge_actions" has been created
+ create_graph 
+--------------
+
+(1 row)
+
+-- Basic ON CREATE SET: first run creates the node
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Alice'})
+    ON CREATE SET n.created = true
+  RETURN n.name, n.created
+$$) AS (name agtype, created agtype);
+  name   | created 
+---------+---------
+ "Alice" | true
+(1 row)
+
+-- ON MATCH SET: second run matches the existing node
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Alice'})
+    ON MATCH SET n.found = true
+  RETURN n.name, n.created, n.found
+$$) AS (name agtype, created agtype, found agtype);
+  name   | created | found 
+---------+---------+-------
+ "Alice" | true    | true
+(1 row)
+
+-- Both ON CREATE SET and ON MATCH SET (first run = create)
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Bob'})
+    ON CREATE SET n.created = true
+    ON MATCH SET n.matched = true
+  RETURN n.name, n.created, n.matched
+$$) AS (name agtype, created agtype, matched agtype);
+ name  | created | matched 
+-------+---------+---------
+ "Bob" | true    | 
+(1 row)
+
+-- Both ON CREATE SET and ON MATCH SET (second run = match)
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Bob'})
+    ON CREATE SET n.created = true
+    ON MATCH SET n.matched = true
+  RETURN n.name, n.created, n.matched
+$$) AS (name agtype, created agtype, matched agtype);
+ name  | created | matched 
+-------+---------+---------
+ "Bob" | true    | true
+(1 row)
+
+-- ON CREATE SET with MERGE after MATCH (Case 1: has predecessor, first run = create)
+SELECT * FROM cypher('merge_actions', $$
+  MATCH (a:Person {name: 'Alice'})
+  MERGE (a)-[:KNOWS]->(b:Person {name: 'Charlie'})
+    ON CREATE SET b.source = 'merge_create'
+  RETURN a.name, b.name, b.source
+$$) AS (a agtype, b agtype, source agtype);
+    a    |     b     |     source     
+---------+-----------+----------------
+ "Alice" | "Charlie" | "merge_create"
+(1 row)
+
+-- ON MATCH SET with MERGE after MATCH (Case 1: has predecessor, second run = match)
+SELECT * FROM cypher('merge_actions', $$
+  MATCH (a:Person {name: 'Alice'})
+  MERGE (a)-[:KNOWS]->(b:Person {name: 'Charlie'})
+    ON MATCH SET b.visited = true
+  RETURN a.name, b.name, b.visited
+$$) AS (a agtype, b agtype, visited agtype);
+    a    |     b     | visited 
+---------+-----------+---------
+ "Alice" | "Charlie" | true
+(1 row)
+
+-- Multiple SET items in a single ON CREATE SET
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Dave'})
+    ON CREATE SET n.a = 1, n.b = 2
+  RETURN n.name, n.a, n.b
+$$) AS (name agtype, a agtype, b agtype);
+  name  | a | b 
+--------+---+---
+ "Dave" | 1 | 2
+(1 row)
+
+-- Reverse order: ON MATCH before ON CREATE should work
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Eve'})
+    ON MATCH SET n.seen = true
+    ON CREATE SET n.new = true
+  RETURN n.name, n.new
+$$) AS (name agtype, new agtype);
+ name  | new  
+-------+------
+ "Eve" | true
+(1 row)
+
+-- Error: ON CREATE SET specified more than once
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Bad'})
+    ON CREATE SET n.a = 1
+    ON CREATE SET n.b = 2
+  RETURN n
+$$) AS (n agtype);
+ERROR:  ON CREATE SET specified more than once
+LINE 1: SELECT * FROM cypher('merge_actions', $$
+                                               ^
+-- Error: ON MATCH SET specified more than once
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (n:Person {name: 'Bad'})
+    ON MATCH SET n.a = 1
+    ON MATCH SET n.b = 2
+  RETURN n
+$$) AS (n agtype);
+ERROR:  ON MATCH SET specified more than once
+LINE 1: SELECT * FROM cypher('merge_actions', $$
+                                               ^
+-- Chained (non-terminal) MERGE with ON CREATE SET (eager-buffering path)
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (a:Person {name: 'Frank'})
+    ON CREATE SET a.created = true
+  MERGE (a)-[:KNOWS]->(b:Person {name: 'Grace'})
+    ON CREATE SET b.created = true
+  RETURN a.name, a.created, b.name, b.created
+$$) AS (a_name agtype, a_created agtype, b_name agtype, b_created agtype);
+ a_name  | a_created | b_name  | b_created 
+---------+-----------+---------+-----------
+ "Frank" | true      | "Grace" | true
+(1 row)
+
+-- Chained (non-terminal) MERGE with ON MATCH SET (second run = match)
+SELECT * FROM cypher('merge_actions', $$
+  MERGE (a:Person {name: 'Frank'})
+    ON MATCH SET a.matched = true
+  MERGE (a)-[:KNOWS]->(b:Person {name: 'Grace'})
+    ON MATCH SET b.matched = true
+  RETURN a.name, a.matched, b.name, b.matched
+$$) AS (a_name agtype, a_matched agtype, b_name agtype, b_matched agtype);
+ a_name  | a_matched | b_name  | b_matched 
+---------+-----------+---------+-----------
+ "Frank" | true      | "Grace" | true
+(1 row)
+
+-- ON keyword as label name (backward compat via safe_keywords)
+SELECT * FROM cypher('merge_actions', $$
+  CREATE (n:on {name: 'test'})
+  RETURN n.name
+$$) AS (name agtype);
+  name  
+--------
+ "test"
+(1 row)
+
+-- Issue #2347: RHS of ON CREATE / ON MATCH SET referencing a bound
+-- variable crashed the backend when MERGE had a previous clause, because
+-- the lateral-join's ParseNamespaceItem had p_nscolumns=NULL.
+-- ON CREATE SET with RHS referencing the outer MATCH's variable
+SELECT * FROM cypher('merge_actions', $$ CREATE (:Person {name:'Anchor'}) $$) AS (a agtype);
+ a 
+---
+(0 rows)
+
+SELECT * FROM cypher('merge_actions', $$
+  MATCH (a:Person {name: 'Anchor'})
+  MERGE (b:Person {name: 'FromOuter'})
+    ON CREATE SET b.source_name = a.name
+  RETURN a.name, b.name, b.source_name
+$$) AS (a_name agtype, b_name agtype, b_source agtype);
+  a_name  |   b_name    | b_source 
+----------+-------------+----------
+ "Anchor" | "FromOuter" | "Anchor"
+(1 row)
+
+-- ON CREATE SET with RHS referencing the MERGE-bound variable itself
+SELECT * FROM cypher('merge_actions', $$
+  MATCH (a:Person {name: 'Anchor'})
+  MERGE (b:Person {name: 'SelfRef'})
+    ON CREATE SET b.echo_name = b.name
+  RETURN b.name, b.echo_name
+$$) AS (b_name agtype, b_echo agtype);
+  b_name   |  b_echo   
+-----------+-----------
+ "SelfRef" | "SelfRef"
+(1 row)
+
+-- ON CREATE SET driven by UNWIND with self-reference on the RHS
+-- (Muhammad's second reproducer)
+SELECT * FROM cypher('merge_actions', $$
+  UNWIND ['U1', 'U2'] AS nm
+  MERGE (n:Person {name: nm})
+    ON CREATE SET n.copy_name = n.name
+  RETURN n.name, n.copy_name
+$$) AS (n_name agtype, n_copy agtype);
+ n_name | n_copy 
+--------+--------
+ "U1"   | "U1"
+ "U2"   | "U2"
+(2 rows)
+
+-- Multiple SET items mixing outer-ref, self-ref, and literal RHS
+SELECT * FROM cypher('merge_actions', $$
+  MATCH (a:Person {name: 'Anchor'})
+  MERGE (b:Person {name: 'MultiItem'})
+    ON CREATE SET b.from_a = a.name, b.self = b.name, b.lit = 'literal'
+  RETURN b.from_a, b.self, b.lit
+$$) AS (fa agtype, sf agtype, lit agtype);
+    fa    |     sf      |    lit    
+----------+-------------+-----------
+ "Anchor" | "MultiItem" | "literal"
+(1 row)
+
+-- ON MATCH SET with variable RHS (second run on existing node)
+SELECT * FROM cypher('merge_actions', $$
+  MATCH (a:Person {name: 'Anchor'})
+  MERGE (b:Person {name: 'FromOuter'})
+    ON CREATE SET b.source_name = a.name
+    ON MATCH SET b.last_seen_by = a.name
+  RETURN b.source_name, b.last_seen_by
+$$) AS (src agtype, last agtype);
+   src    |   last   
+----------+----------
+ "Anchor" | "Anchor"
+(1 row)
+
+-- cleanup
+SELECT * FROM cypher('merge_actions', $$ MATCH (n) DETACH DELETE n $$) AS (a agtype);
+ a 
+---
+(0 rows)
+
 --
 -- delete graphs
 --
+SELECT drop_graph('merge_actions', true);
+NOTICE:  drop cascades to 5 other objects
+DETAIL:  drop cascades to table merge_actions._ag_label_vertex
+drop cascades to table merge_actions._ag_label_edge
+drop cascades to table merge_actions."Person"
+drop cascades to table merge_actions."KNOWS"
+drop cascades to table merge_actions."on"
+NOTICE:  graph "merge_actions" has been dropped
+ drop_graph 
+------------
+
+(1 row)
+
 SELECT drop_graph('issue_1907', true);
 NOTICE:  drop cascades to 4 other objects
 DETAIL:  drop cascades to table issue_1907._ag_label_vertex