neo4j导入csv文件

1、我有两个节点，分别是Person和 Moive，有 Person- $:ACTED_IN$ ->Moive (ACTED_IN是参演关系，表示某个人参演了某个电影)，Person- $:DIRECTED$ ->Moive （DIRECTED表示是执导关系，表示某个人执导了某部电影），Person- $:REVIEWED$ ->Moive （REVIEWED是编剧关系，表示某个人是某部电影的编剧），Person- $:FOLLOWS$ ->Person（FOLLOWS是关注关系，表示某个人关注了另一个人）

2、首先导入Node节点的csv文件，首先 csv文件必须要放在 neo4j安装目录的import文件夹下

bash 复制代码

LOAD CSV WITH HEADERS FROM 'file:///person.csv' AS row
CREATE (n:Person {personId:row.personId ,name: row.name, born: toInteger(row.born)})


LOAD CSV WITH HEADERS FROM 'file:///movie.csv' AS row
CREATE (n:Movie {movieId:row.movieId,title: row.title, votes: toInteger(row.votes),tagline:row.tagline,released:toInteger(row.released)})

我本地的数据路径如下

3、导入关系文件

bash 复制代码

//参演
LOAD CSV WITH HEADERS FROM 'file:///r_ACTED_IN_person_2_moive.csv' AS row
MATCH (p:Person {name: row.`:START_ID`})
MATCH (m:Movie {tagline: row.`:END_ID`})
MERGE (p)-[r:ACTED_IN]->(m)
SET r.roles = row.roles

//指导
LOAD CSV WITH HEADERS FROM 'file:///r_DIRECTED_person_2_moive.csv' AS row
MATCH (p:Person {name: row.`:START_ID`})
MATCH (m:Movie {tagline: row.`:END_ID`})
MERGE (p)-[r:DIRECTED]->(m)

//REVIEWED 编剧
LOAD CSV WITH HEADERS FROM 'file:///r_REVIEWED_person_2_movie.csv' AS row
MATCH (p:Person {name: row.`:START_ID`})
MATCH (m:Movie {tagline: row.`:END_ID`})
MERGE (p)-[r:REVIEWED]->(m)
SET r.summary = row.summary ,r.rating = row.rating

//关注
LOAD CSV WITH HEADERS FROM 'file:///r_FOLLOWS_person_2_person.csv' AS row
MATCH (p:Person {name: row.`:START_ID`})
MATCH (m:Person {name: row.`:END_ID`})
MERGE (p)-[r:FOLLOWS]->(m)

3、导入完成后，可以看到如下

4、如果想导出可以直接执行的csv文件，首先要先写match语句，如下（适合数据量不大的情况）

bash 复制代码

MATCH (p:Person)-[r:ACTED_IN]->(m:Movie)
RETURN 
     p.name AS `:START_ID`, 
     'ACTED_IN' AS `:TYPE`, 
     m.tagline AS `:END_ID`, 
     r.roles AS roles

MATCH (p:Person)-[r:DIRECTED]->(m:Movie)
RETURN 
     p.name AS `:START_ID`, 
     'DIRECTED' AS `:TYPE`, 
     m.tagline AS `:END_ID` 

MATCH (p:Person)-[r:REVIEWED]->(m:Movie)
RETURN 
     p.name AS `:START_ID`, 
     'REVIEWED' AS `:TYPE`, 
     m.tagline AS `:END_ID`,
     r.summary as summary,
     r.rating as rating

MATCH (p:Person)-[r:FOLLOWS]->(m:Person)
RETURN 
     p.name AS `:START_ID`, 
     'FOLLOWS' AS `:TYPE`, 
     m.name AS `:END_ID`