如何在Neo4j中交换两个节点的指定属性及所有关系？

先修正你原查询的错误 #

你之前的CREATE (Temp:Student{properties(node1)})写法不符合Cypher语法——Cypher不支持在节点创建的属性字典中直接调用properties()函数。正确的临时节点属性复制方式是先创建空节点，再用SET Temp = node1全量复制属性。

核心需求实现：无需临时节点的方案 #

1. 属性交换（排除id） #

如果节点属性较少，可以直接用WITH子句暂存属性值，避免赋值时互相覆盖：

MATCH (node1 {id: 'id1'}), (node2 {id: 'ID1'})
WITH node1, node2, 
     node1.name AS n1_name, node1.city AS n1_city,
     node2.name AS n2_name, node2.city AS n2_city
SET node1.name = n2_name, node1.city = n2_city,
    node2.name = n1_name, node2.city = n1_city
RETURN node1, node2

如果节点有大量属性，推荐用APOC工具库简化操作（需提前安装APOC），自动排除id属性：

MATCH (node1 {id: 'id1'}), (node2 {id: 'ID1'})
WITH node1, node2, 
     apoc.map.removeKey(properties(node1), 'id') AS props1,
     apoc.map.removeKey(properties(node2), 'id') AS props2
SET node1 += props2, node2 += props1
RETURN node1, node2

+=运算符会合并属性，仅覆盖除id外的所有字段，保留原节点的id值。

2. 所有关系交换 #

要把两个节点的全部关系（包括入、出方向）互换，同样用APOC的重构函数批量处理：

MATCH (node1 {id: 'id1'}), (node2 {id: 'ID1'})
// 把node1的出关系转移给node2
OPTIONAL MATCH (node1)-[r1]->(other)
CALL apoc.refactor.from(r1, node2) YIELD input
// 把node1的入关系转移给node2
OPTIONAL MATCH (other)-[r2]->(node1)
CALL apoc.refactor.to(r2, node2) YIELD input
// 把node2的出关系转移给node1
OPTIONAL MATCH (node2)-[r3]->(other)
CALL apoc.refactor.from(r3, node1) YIELD input
// 把node2的入关系转移给node1
OPTIONAL MATCH (other)-[r4]->(node2)
CALL apoc.refactor.to(r4, node1) YIELD input
RETURN node1, node2

3. 属性+关系交换的完整整合查询 #

MATCH (node1 {id: 'id1'}), (node2 {id: 'ID1'})
WITH node1, node2, 
     apoc.map.removeKey(properties(node1), 'id') AS props1,
     apoc.map.removeKey(properties(node2), 'id') AS props2
// 交换属性
SET node1 += props2, node2 += props1
// 交换所有关系
OPTIONAL MATCH (node1)-[r1]->(other) CALL apoc.refactor.from(r1, node2) YIELD input
OPTIONAL MATCH (other)-[r2]->(node1) CALL apoc.refactor.to(r2, node2) YIELD input
OPTIONAL MATCH (node2)-[r3]->(other) CALL apoc.refactor.from(r3, node1) YIELD input
OPTIONAL MATCH (other)-[r4]->(node2) CALL apoc.refactor.to(r4, node1) YIELD input
RETURN node1, node2

备选：修正后的临时节点方案 #

如果仍偏好使用临时节点，正确写法如下：

MATCH (node1 {id: 'id1'}), (node2 {id: 'ID1'})
CREATE (Temp:Student)
SET Temp = node1  // 全量复制node1的属性到临时节点
SET node1.name = node2.name, node1.city = node2.city
SET node2.name = Temp.name, node2.city = Temp.city
DELETE Temp
RETURN node1, node2

内容的提问来源于stack exchange，提问作者nik