English | 中文
NGBATIS is a database ORM framework base NebulaGraph + spring-boot, which takes advantage of the mybatis’ fashion development, including some de-factor operations in single table and vertex-edge, like mybatis-plus.
If you prefer JPA, graph-ocean is a good choice.
- Springboot
- Maven
- Java 8+
You could refer to ngbatis-demo in this repo.
-
Include in your
pom.xml-
Maven
<dependency> <groupId>org.nebula-contrib</groupId> <artifactId>ngbatis</artifactId> <version>1.1.5</version> </dependency>
-
Gradle
implementation 'org.nebula-contrib:ngbatis:1.1.5'
-
-
Referring to ngbatis-demo, which was smoothly integrated with spring-boot. The API examples could be found under the test of it for all features of ngbatis.
-
Configure the NebulaGraph Database
Configure
application.ymlwith the host and credential to enable access to the NebulaGraph Cluster.
nebula:
ngbatis:
session-life-length: 300000 # since v1.1.2
check-fixed-rate: 300000 # since v1.1.2
# space name needs to be informed through annotations(@Space) or xml(space="test")
# default false(false: Session pool map will not be initialized)
use-session-pool: false # since v1.1.2
hosts: 127.0.0.1:19669, 127.0.0.1:9669
username: root
password: nebula
space: test
pool-config:
min-conns-size: 0
max-conns-size: 10
timeout: 0
idle-time: 0
interval-idle: -1
wait-time: 0
min-cluster-health-rate: 1.0
enable-ssl: false- Dynamically register beans
@SpringBootApplication(scanBasePackages = { "org.nebula", "your.domain"})
public class YourApplication {
public static void main(String[] args) {
new SpringApplication(YourApplication.class).run(args);
}
}If SpringCloud is used in your project, please use
@ComponentScan( basePackages = {"org.nebula.contrib", "your.domain"} )instead.
package ye.weicheng.ngbatis.demo.repository;
import ye.weicheng.ngbatis.demo.pojo.Person;
import java.util.List;
import java.util.Map;
import java.util.Set;
public interface TestRepository {
Person selectPerson();
Person selectByPerson(Person person);
List<Person> selectAgeGt(Integer age);
List<String> selectListString();
List<Map> selectPersonsMap();
Map<String, Object> selectTriple();
}resource/mapper/TestRepository.xml
<mapper
namespace=
"ye.weicheng.ngbatis.demo.repository.TestRepository"
>
<select id="selectPerson" resultType="ye.weicheng.ngbatis.demo.pojo.Person">
match (v:person) return v.person.name as name, v.person.age as age limit 1
</select>
<select id="selectAgeGt" resultType="ye.weicheng.ngbatis.demo.pojo.Person">
MATCH (n: person)
WHERE n.person.age > $p0
RETURN n
LIMIT 100
</select>
<select id="selectByPerson" resultType="ye.weicheng.ngbatis.demo.pojo.Person">
MATCH (n: person)
WHERE n.person.name == $p0.name
RETURN n
LIMIT 100
</select>
<select id="selectListString" resultType="java.lang.String">
match (v:person) return v.person.name as name limit 100
</select>
<select id="selectPersonsMap" resultType="java.util.Map">
match (v:person) return v.person.name as name, v.person.age as age limit 100
</select>
<select id="selectTriple" resultType="java.util.Map">
MATCH (n: person)-[r: like]->(n2: person)
RETURN n, r, n2
LIMIT 100
</select>
<!--
More complex `nGQL` may need to be fully tested.
The two-layer object data structure of the project I am currently using is also satisfying.
`Path` is not yet supported because it can basically be handled by the `n, r, n2` structure in development.
-->
</mapper>package com.example.model.vertex.Person;
import lombok.Data;
import javax.persistence.Id;
import javax.persistence.Table;
@Data
@Table(name = "person")
public class Person {
@Id
private String name;
private Integer age;
}package com.example.model.edge.Like;
import lombok.AllArgsConstructor;
import lombok.NoArgsConstructor;
import lombok.Data;
import javax.persistence.Table;
@Data
@Table(name = "like")
@NoArgsConstructor
@AllArgsConstructor
public class Like {
private Double likeness;
}package com.example.dao;
import org.nebula.contrib.ngbatis.proxy.NebulaDaoBasic;
import com.example.model.vertex.Person;
public interface PersonDao extends NebulaDaoBasic<Person, String>{}Note, this is a mandatory requirement.
<mapper
namespace=
"com.example.dao.PersonDao"
>
</mapper>package com.example.service;
import org.nebula.contrib.ngbatis.utils.Page;
import com.example.dao.PersonDao;
import com.example.model.vertex.Person;
import com.example.model.edge.Like;
@Service
public class PersonServiceImpl {
@Autowired private PersonDao dao;
public void demos() {
// Implement two node insertions
Person tom = new Person();
tom.setName("Tom");
dao.insert( tom );
Person jerry = new Person();
jerry.setName( "Jerry" );
dao.insert( jerry );
// Establishing the relationship between two nodes
Like like = new Like( 0.99999 );
dao.insertEdge( tom, like, jerry );
// Find people who like jerry
String jerryId = jerry.getName();
List<Person> whoLikeJerry = dao.listStartNodes( Like.class, jerryId );
// Find the only people who like jerry, Non-Unique Times Error。（Limited to scenarios where there is only one upstream for a given relationship）
Person tom = dao.startNode( Like.class, jerryId );
// See the Like relationship between Tom and Jerry
String tomId = tom.getName();
Boolean tomLikeJerry = dao.existsEdge( tomId, Like.class, jerryId ); // true
Boolean jerryLikeTom = dao.existsEdge( jerryId, Like.class, tomId ); // false
// Poor Tom
// Find all information by Tom's name
Person tomDb = dao.selectById( "Tom" );
// Search by page
Page<Person> page = new Page<>();
List<Person> personPage = dao.selectPage( page );
page.getTotal(); // 2 rows， Tom and Jerry
Boolean theyAreFamily = page.getRows() == personPage; // true
// The story always wants to have a good ending
dao.insertEdge( jerry, like, tom );
// More base class operations are still under development；Expectations
}
}- beetl, BSD-3, we proudly use the beetl template language as our template engine, which is consumed in binary package(as is).
NGBATIS is under the Apache License, Version 2.0.