NyBatis is Persistent Data Mapper Framework for Relation DataBase.

The name of NyBatis is abbreviation of Nextgeneration of Your Batis.

Nowaday general trend is JPA or Hibernate and many people think that Data Mapper Framework is old and not easy.

I agree JPA is very brilliant solution but believe that there is no silver bullet. DMF(Data Mapper Framework) is also appropriate in some business application.

I admire MyBatis but it's feature is gradually outdated and nothing to happy for novice developer. So I develop this an alternative for IBatis or MyBatis.

Focus of NyBatis is simple and easy. Many feature to be determined by developer was automated so it helps to keep code clear from SQL.

I confess my lazy. I developed it on circumstance with Java 8 so NyBatis is only supports Java 8. NyBatis also can be used in spring, struts and so on But it still doesn't have simple configruation to apply with above popular framework yet.

This project has JARs deployed to the Unofficial Github private repository.

Add repository to pom.xml.

<repositories>
  <repository>
    <id>UnofficialNybatisMavenRepository</id>
    <url>https://github.com/nayasis/UnofficialMavenRepository/raw/master/</url>
  </repository>
</repositories>

And add the following dependency to pom.xml.

<dependency>
  <groupId>org.nybatis</groupId>
  <artifactId>NyBatisCore</artifactId>
  <version>0.0.1-SNAPSHOT</version>
</dependency>

Default configuration file path is below.

/resources/config/db/config.xml

And load like below.

DatabaseConfigurator.build();

Also configuration path is changable.

DatabaseConfigurator.build( "/home/user/config/yourConfiguration.xml" );

And it is possible to load from your application running directory.

DatabaseConfigurator.build( Const.path.getRoot() + "/yourConfiguration.xml" );

SQL is written as XML in file like MyBatis. but SQL id defined as FileName.id only.

User.xml

<mapper>
  <sql id="select">
    SELECT  *
    FROM    USER
    WHERE   1=1
    AND     name = #{name}
      <if test="#{age} != empty">
      AND     age  = #{age}
      </if>
      <else>
      AND     age <= 20
      </else>
  </sql>
</mapper>

Java code is like below. (sql id is User + select )

SqlSession session = SessionManager.openSession();

Map param = new HashMap();
param.put( "name", "nayasis" );
param.put( "age",  18 );

List<Map> list = session.sqlId( "User.select", param ).list().select();

Also you can define specific return type in result set.

public class User {
   public String name;
   public int    age;
}
...

SqlSession session = SessionManager.openSession();

User param = new User();
param.setName( "nayasis" );
param.setAge( 18 );

List<User> list = session.sqlId( "User.mapper", param ).list().select( User.class );

XML SQL only contains id and query. it does not contains code dependent attribute like parameter type or return class.

SQL also can be injected in NyBatis like Spring JDBC template.

session.sql( "UPDATE USER SET age = #{age} WHERE name = #{name}", param ).execute();

When method execute() is called, transaction will be activated automatically. NyBatis manages transaction mechanism by itself and developer does not annoy to controll it. It will be explained in detail later.

Another tip is that you can write '<' charater in XML :)

NyBatis supports simple ORM mechanism.

@Table( "TB_DEV_USER" )
public class User {
   public String name;
   public int    age;
}
...

OrmSession session = SessionManager.openOrmSession( User.class );

User param = new User();
param.setName( "nayasis" );

User user = session.select( user );
user.setAge( 20 );

session.update( user ); // if user data was cached, cache will be refreshed.
session.delete( user ); // if user data was cached, cache will be deleted.
session.insert( user );

List data also can be retrived.

OrmSession session = SessionManager.openOrmSession( User.class );

List<User> list = session.list().setPage( 10, 50 ).select();

int count = session.list().count();

Dynamic condition can be written like below.

OrmSession session = SessionManager.openOrmSession( User.class );

List<User> list01 = session.list()
   .where( "age >= #{age}",  19 ).select();

List<User> list02 = session.list()
   .where( "name like #{name}", "%y%" )
   .where( "age >= #{age}",  19 )
   .orderBy( "name" )
   .select();

Map param = new HashMap();
param.put( "name", "%y%" );
param.put( "age", 19 );

List<User> list03 = session.list()
   .where( "name like #{name} OR age >= #{age}", param )
   .orderBy( "name, age DESC" )
   .select();

NyBatis supports mass data manifulation mode like addBatch in JDBC.

List<User> users = ...

// In SqlSession
SqlSession sqlSession = SessionManager.openSession();
sqlSession.batchSqlId( "User.insert", users ).execute();

// In OrmSession
OrmSession ormSession = SessionManager.openOrmSession( User.class );
ormSession.batch().insert( users );
ormSession.batch().update( users );
ormSession.batch().delete( users );

It does not needs any addtional configuration.

Dificullty of testing DAO (DataAccessObject) is that testing data is flexible so it is necessary to create each DAO's own testing data.

But creating test data is a little annoying to implement in code.

Using Spring Jdbc template or Pure jdbc statement or Interface query of MyBatis... Using many DMF in one source is not delight work.

In NyBatis, you can run sql script to make preliminary data easily in test code.

SqlSession sqlSession = SessionManager.openSession();
sqlSession.sql( "INSERT INTO USER ( name, age ) VALUES ( 'test', '20' )" ).execute();

... \\ Real Testing Code

sqlSession.sql( "DELETE USER WHERE name = 'test'" ).execute();