Neo4j Cyber 使用

Neo4j是一个世界领先，具有完备事务的开源图形数据库。

图形数据库是以图形结构的形式存储数据的数据库。它以节点，关系和属性的形式存储应用程序的数据。正如RDBMS以表的“行，列”的形式存储数据，GDBMS以“图形”的形式存储数据。

图数据库中每个节点都直接包含一个关系列表，关系列表中存放此节点与其他节点的关系记录。这些关系记录按类型和方向组织起来，并且可以保存附加属性。图数据库可以使用此列表直接访问连接的节点，无需进行记录的搜索，匹配计算。

1. 单向关系

1 -> 2 -> 5 -> 9
       -> 6
  -> 3 -> 7 -> 10
  -> 4 -> 8 

创建节点和单向关系

create (zhang1:People{name:'zhang1'}) 
create (zhang2:People{name:'zhang2'}) 
create (zhang3:People{name:'zhang3'})
create (zhang4:People{name:'zhang4'})
create (zhang5:People{name:'zhang5'})
create (zhang6:People{name:'zhang6'})
create (zhang7:People{name:'zhang7'})
create (zhang8:People{name:'zhang8'})
create (zhang9:People{name:'zhang9'})
create (zhang10:People{name:'zhang10'})

create (zhang1)-[:FOLLOW]->(zhang2)
create (zhang1)-[:FOLLOW]->(zhang3)
create (zhang1)-[:FOLLOW]->(zhang4)
create (zhang2)-[:FOLLOW]->(zhang5)
create (zhang2)-[:FOLLOW]->(zhang6)
create (zhang3)-[:FOLLOW]->(zhang7)
create (zhang4)-[:FOLLOW]->(zhang8)
create (zhang5)-[:FOLLOW]->(zhang9)
create (zhang7)-[:FOLLOW]->(zhang10)

[r:FOLLOW*0..]：查询的关系深度从0到无止尽

in： 即 a -> b

match p=(a:People) - [r:FOLLOW*0..] -> (b:People) where b.name = 'zhang2' return p

out： 即 b -> c

match p=(b:People) - [rr:FOLLOW*0..] -> (c:People) where b.name = 'zhang2' return p

both： 即 a <-> b <-> c

match p=(a:People) <- [r:FOLLOW*0..] -> (b:People) <- [rr:FOLLOW*0..] -> (c:People)
where b.name = 'zhang9' return p

2. 单双向关系

1 <-> 2 <-> 5 -> 9
         -> 6
   -> 3 <-> 7 -> 10
   -> 4 <-> 8   

创建节点和单双向关系

create (zhang1:User{name:'zhang1'}) 
create (zhang2:User{name:'zhang2'}) 
create (zhang3:User{name:'zhang3'})
create (zhang4:User{name:'zhang4'})
create (zhang5:User{name:'zhang5'})
create (zhang6:User{name:'zhang6'})
create (zhang7:User{name:'zhang7'})
create (zhang8:User{name:'zhang8'})
create (zhang9:User{name:'zhang9'})
create (zhang10:User{name:'zhang10'})

create (zhang1)-[:FOLLOW]->(zhang2)
create (zhang1)-[:FOLLOW]->(zhang3)
create (zhang1)-[:FOLLOW]->(zhang4)
create (zhang2)-[:FOLLOW]->(zhang1)
create (zhang2)-[:FOLLOW]->(zhang5)
create (zhang2)-[:FOLLOW]->(zhang6)
create (zhang3)-[:FOLLOW]->(zhang7)
create (zhang4)-[:FOLLOW]->(zhang8)
create (zhang5)-[:FOLLOW]->(zhang2)
create (zhang5)-[:FOLLOW]->(zhang9)
create (zhang7)-[:FOLLOW]->(zhang3)
create (zhang7)-[:FOLLOW]->(zhang10)
create (zhang8)-[:FOLLOW]->(zhang4)

MATCH 语句：像SQL语句一样

match (u:User) return u
match (u:User) where u.name='zhang5' return u
match p=(u:User)-[r:FOLLOW]->(uu:User) return p

[r:FOLLOW*0..]：查询的关系深度从0到无止尽

in： 即 a -> b

match p=(a:User) - [r:FOLLOW*0..] -> (b:User) where b.name = 'zhang2' return p

out： 即 b -> c

match p=(b:User) - [rr:FOLLOW*0..] -> (c:User) where b.name = 'zhang2' return p

both： 即 a <-> b <-> c

match p=(a:User) <- [r:FOLLOW*0..] -> (b:User) <- [rr:FOLLOW*0..] -> (c:User) 
where b.name = 'zhang9' return p

参考：
Neo4j官方文档
Neo4j官方Cyber手册
neo4j 教程 w3cschool