Neo4j开发总结

最新推荐文章于 2025-11-10 15:56:11 发布

原创最新推荐文章于 2025-11-10 15:56:11 发布 · 1.1k 阅读

1 ·

CC 4.0 BY-SA版权

文章标签：

#Neo4j #cql

数据库专栏收录该内容

26 篇文章

订阅专栏

1、查看两个节点之间的关系使用：type()函数

2、DISTINCT()去重函数注意：该函数要放在RETURN()后面，并且后面加逗号（,）.放在最后会报错

3、查询几度关系数据

4、查询两个节点间的所有关系（不确定关系）

5、java代码中模糊检索Neo4j

6、case、when

7、处理字符串转数字

8、分页limit：对Neo4j中通过union后结果进行分页（教程：需要安装apoc插件；jar包地址：GitHub）

9、Neo4j内存配置

1、查看两个节点之间的关系使用：type()函数

MATCH (:paper)-[r]->(:field_sub) 
RETURN type(r)

2、DISTINCT()去重函数注意：该函数要放在RETURN()后面，并且后面加逗号（,）.放在最后会报错

//正确代码
MATCH (pe:person)-[:write_paper]->(pa:paper)
RETURN
	DISTINCT(pe.nationality),
	pe.nationality AS country,
    COUNT(pa)

//报错代码
MATCH (pe:person)-[:write_paper]->(pa:paper)
RETURN
	pe.nationality AS country,
    COUNT(pa),
    DISTINCT(pe.nationality)

//报错信息
Neo.ClientError.Statement.SyntaxError
Neo.ClientError.Statement.SyntaxError: Unknown function 'DISTINCT' (line 5, column 5 (offset: 99))
"    DISTINCT(pe.nationality)"
     ^

3、查询几度关系数据

MATCH p=(pe1:Person)-[*1..4]-(pe2:Person) RETURN p LIMIT 25

4、查询两个节点间的所有关系（不确定关系）

MATCH (o1:organization)-- (p:project) -- (o2:organization) 
WHERE o1.orgId = {orgId1}
AND p.startYear >= {startYear} AND p.startYear <= {endYear}
RETURN o2.orgId AS orgId, o2.org_name AS orgName
LIMIT 100

5、java代码中模糊检索Neo4j

#Neo4j图库中查询语句
MATCH (n:DataSet{code:'uci'}) 
WHERE n.name =~ '(?i).*blog.*'
RETURN n.datasetId AS datasetId, n.name as name 
LIMIT 10

#java中查询
MATCH (n:DataSet{code:'uci'}) 
WHERE n.name =~ {name}
RETURN n.datasetId AS datasetId, n.name as name 
LIMIT 10

将（(?i).*blog.*）整体作为一个参数传入
代码中提前拼接好：name = "(?i).*" + name + ".*";

6、case、when

                    "MATCH  (org:organization) -[ogor:org_own_reward]->(rd:reward) " +
                        "  	WHERE 	toInteger(subString(ogor.date,0,4)) >= #startYear and toInteger(subString(ogor.date,0,4))<= #endYear  " +
                        "   		and rd.rewardType = '国际科学技术合作奖'  " +
                        "  	RETURN " +
                        "   	case  " +
                        "    		when ogor.type is not null  " +
                        "     			then ogor.type " +
                        "    	else " +
                        "      		'组织'  " +
                        "   	end  " +
                        "   	AS rewardArea,count(ogor) as statis  " +
                        "  	ORDER BY rewardArea ";

7、处理字符串转数字

toInteger(substring(pj.startDate,0,4))

8、分页limit：对Neo4j中通过union后结果进行分页（教程：需要安装apoc插件；jar包地址：GitHub）

CALL apoc.cypher.run("
     MATCH (n:Disease) RETURN n.name AS field1, n.desc  AS field2
     ORDER BY n.field1 ASC LIMIT 100
     UNION
     MATCH (n:Symptom) RETURN n.name AS field1, n.name AS field2
     ORDER BY n.field1 ASC LIMIT 100
",{})  yield value
    return value.field1,value.field2
    order by value.field1 limit 100

    @Query("CALL apoc.cypher.run( {cql} ,{}) " +
            "yield value " +
            "RETURN value.achName AS achName, value.psName AS psName, value.psId AS psId, value.influence AS influence, value.achId AS achId " +
            "SKIP {skip} LIMIT {limit}")
    List<Map<String, Object>> queryAwardByPersonId(@Param("cql") String cql,
                                                   @Param("skip") int skip,
                                                   @Param("limit") int limit);

9、Neo4j内存配置

# java heap 初始值 
dbms.memory.heap.initial_size=1g 
# java heap 最大值，一般不要超过可用物理内存的80％ 
dbms.memory.heap.max_size=16g 
# pagecache大小，官方建议设为：(总内存-dbms.memory.heap.max_size)/2
dbms.memory.pagecache.size=2g