Neo4j 图数据库入门（三）：从会写到写好——MERGE 防重、聚合分析、索引约束与完整实战

1
// ============ 第一步：清空数据库（从零开始） ============
2
MATCH (n) DETACH DELETE n;
3

4
// ============ 第二步：约束先行 ============
5
CREATE CONSTRAINT unique_person_name IF NOT EXISTS
6
FOR (p:Person) REQUIRE p.name IS UNIQUE;
7

8
CREATE CONSTRAINT unique_movie_title IF NOT EXISTS
9
FOR (m:Movie) REQUIRE m.title IS UNIQUE;
10

11
// ============ 第三步：索引加速 ============
12
CREATE INDEX person_name IF NOT EXISTS
13
FOR (p:Person) ON (p.name);
14

15
// ============ 第四步：MERGE 节点 ============
16
MERGE (p1:Person {name: '张三'}) ON CREATE SET p1.age = 30, p1.gender = '男';
17
MERGE (p2:Person {name: '李四'}) ON CREATE SET p2.age = 28, p2.gender = '男';
18
MERGE (p3:Person {name: '王五'}) ON CREATE SET p3.age = 35, p3.gender = '男';
19
MERGE (p4:Person {name: '赵六'}) ON CREATE SET p4.age = 26, p4.gender = '女';
20

21
MERGE (m1:Movie {title: '流浪地球'}) ON CREATE SET m1.year = 2019, m1.rating = 7.9;
22
MERGE (m2:Movie {title: '三体'})     ON CREATE SET m2.year = 2023, m2.rating = 8.5;
23
MERGE (m3:Movie {title: '满江红'})   ON CREATE SET m3.year = 2023, m3.rating = 7.0;
24

25
MERGE (c1:Company {name: '中影股份'}) ON CREATE SET c1.founded = 1999;
26
MERGE (c2:Company {name: '光线传媒'}) ON CREATE SET c2.founded = 1998;
27

28
// ============ 第五步：MERGE 关系 ============
29
MATCH (a:Person {name: '张三'}), (b:Person {name: '李四'})
30
MERGE (a)-[:KNOWS {since: 2018}]->(b);
31

32
MATCH (a:Person {name: '李四'}), (b:Person {name: '王五'})
33
MERGE (a)-[:KNOWS {since: 2020}]->(b);
34

35
MATCH (a:Person {name: '王五'}), (b:Person {name: '赵六'})
36
MERGE (a)-[:KNOWS {since: 2021}]->(b);
37

38
MATCH (a:Person {name: '张三'}), (b:Person {name: '王五'})
39
MERGE (a)-[:KNOWS {since: 2019}]->(b);
40

41
MATCH (p:Person {name: '张三'}), (m:Movie {title: '流浪地球'})
42
MERGE (p)-[:ACTED_IN {role: '工程师'}]->(m);
43

44
MATCH (p:Person {name: '李四'}), (m:Movie {title: '流浪地球'})
45
MERGE (p)-[:DIRECTED]->(m);
46

47
MATCH (p:Person {name: '赵六'}), (m:Movie {title: '三体'})
48
MERGE (p)-[:ACTED_IN {role: '科学家'}]->(m);
49

50
MATCH (p:Person {name: '王五'}), (c:Company {name: '中影股份'})
51
MERGE (p)-[:WORKS_AT {position: '制片人'}]->(c);
52

53
MATCH (p:Person {name: '张三'}), (m:Movie {title: '满江红'})
54
MERGE (p)-[:REVIEWED {score: 8}]->(m);
55

56
// ============ 第六步：验证与分析 ============
57

58
// 6.1 全图概览（Browser 画布上会看到第一篇的"世界观"全景图）
59
MATCH (n) RETURN n;
60

61
// 6.2 张三的二度人脉（第一篇开头的"五层 JOIN 问题"的正面回答）
62
MATCH (:Person {name: '张三'})-[:KNOWS]->()-[:KNOWS]->(fof:Person)
63
RETURN DISTINCT fof.name AS 张三的二度人脉;
64

65
// 6.3 张三到赵六的最短路径
66
MATCH path = shortestPath(
67
  (:Person {name: '张三'})-[*]-(:Person {name: '赵六'})
68
)
69
RETURN path;
70

71
// 6.4 每部电影参演人数排行
72
MATCH (m:Movie)<-[:ACTED_IN]-(p:Person)
73
RETURN m.title AS 电影, count(p) AS 演员数
74
ORDER BY 演员数 DESC;
75

76
// 6.5 社交达人榜（WITH + 聚合 + 过滤）
77
MATCH (p:Person)-[:KNOWS]->(friend:Person)
78
WITH p, count(friend) AS friendCount
79
WHERE friendCount > 1
80
RETURN p.name AS 达人, friendCount AS 朋友数;