Introduction:

Database data generator based on Spring Data Repository. This tool can be used in your spring data project for generating some data.

Requirements

• spring data module
• spring data repos
• jpa entities

Work stages:

• Find all JpaRepository beans through @EnableJpaRepositories, @EntityScan.Get entity classes.
• Store metadata from @Table, @ManyToMany,@OneToMany,@OneToOne,@ManyToOne and etc .
• Traversal all entities and construct relations(dependents and neighbours) each other.
• Generate plain instances for entities without dependencies.
• Generate parent instances having @OneToOne(optional=true),@OneToMany and etc
• Generate child instances having parent object as dependency(usual haves a column with parent id)
  • Put generated objects to cache if the cache is empty.
• Generate neighbour relations(ManyToMany link).

Properties:

• set default jpa properties - @EnableJpaRepositories, @EntityScan and etc
• should set @EnableDatabaseDataGenerator
• Tables with @OneToOne relations should field optional=true to be exist for one of that tables
• Tables with @oneToOne relations and @PrimaryKeyJoinColumn should have this annotation only one for this relation
  • should have cascade for secondary entity with merge or all
• add gradle dependency from your local repo :)

compile group: 'ru.generator.db.data', name: 'db-data-generator', version: '0.2'

• (optional) add property generator.cache-entity-size to your application file.
  • It manages generated size between many2many relations. By default it is 20. For example, we have 2 tables A and B with m2m rel. Generator takes 20 entities from each table and generate 400 relations each other.
• if you have id fields without @GeneratedValue generator will use sequence for numeric and random for uuid. For set start sequence value , using

    generator.startId() // 0 by default

Subjects:

• DatabaseDataGeneratorFactory - generator factory.
• - common class generates common logic.
• InnerLog - log entity contains inner log stats.
• InnerCache - inner cache with generated objects.It can be used for getting generated objects without an db impact.
• MetaData - inner data class has all parsed information from JpaRepository.
• PlainTypeGenerator - interface describes a method for generation plain types(integer, long, string and etc).
  • It takes Metadata instance for custom generation.
• Action - SPA for customizing generated value
• ColumnPredicate - Spa for filtering rules.

Notes:

• if a field hasn't an annotation @Column with column name the generator takes column name from field name converting camel case to snake case.
• if objects has a cycle(obj1 has obj2 field then obj2 has obj3 field then obj3 has obj1 field) one of that columns must be optional.

Usage examples:

Get the factory:

     @Autowired
     private DatabaseDataGeneratorFactory factory;
     

Generate all values

    String report =
            factory
                .generator()
                .generateAll()
                .report();

Generate by Class

    String report =
            factory
                .generator()
                .generateBy(Limit.class)
                .generateBy(Customer.class)
                .generateBy(Currency.class)
                .generateBy(OrganizationUnit.class)
                .report();

Generate by Table

       String report =
            factory
                .generator()
                .generateBy("schema","table")
                .report();

Generate by Table with exception

    try {
      InnerLog log =
          factory
              .generator()
              .generateBy("schema", "table")
              .withException()
              .log;
    } catch (DataGenerationException e) {
      e.printStackTrace();
    }

Generate by Class and repeate it 10 times

 InnerLog log =
          factory
              .generator().repeate(10)
              .generateBy(Limit.class)
              .log;

Generate by Class and do it till exprected log size

    InnerLog s =
        factory
            .generator().metronome(1, TimeUnit.SECONDS)
            .predicate(countPredicate(10))
            .generateBy(Customer.class)
            .log();

    InnerLog s =
        factory
            .generator()
            .metronome(1, TimeUnit.SECONDS,ctx -> ctx.log.markerValue() < 10)
            .generateBy(Customer.class)
            .log();

Generate by Class and get cache

    InnerCache cache = factory
        .generator()
        .metronome(10,TimeUnit.MILLISECONDS,COUNT(10))
        .generateBy(SimplePlainObject.class)
        .cache();

Generate by Class and do it async

    factory
        .generator().async()
        .repeate(100)
        .generateBy(SimplePlainObject1.class)
        .generateBy(SimplePlainObject2.class)
        .generateBy(SimplePlainObject3.class)
        .finish();

Add specific logic for generated value - rule(ColumnPredicate, Action, Class<?>)

    import static ru.gpb.utils.db.data.generator.worker.Action.*;
    import static ru.gpb.utils.db.data.generator.worker.ColumnPredicate.*;
    ... 
    
    NakedObject spo = factory
    	  .generator()
    	  .rule(FIELD("fieldWithCamel"), CONST("newValue"), String.class)
    	  .generateBy(NakedObject.class)
    	  .cache()
    	  .getValueList(NakedObject.class)
    	  .get(0);
    
    	assertEquals("newValue", spo.getFieldWithCamel());
    	
    NakedObject spo = factory
          .generator()
          .rule(
            COMPOSE(CLASS(NakedObject.class), FIELD("fieldWithCamel")), 
            CONST("newValue"), String.class)
          .generateBy(NakedObject.class)
          .generateBy(NakedObject2.class)
          .generateBy(NakedObject3.class)
          
    List<SeqIncObject> list =
    	  factory.generator()
    		.repeate(10)
    		.rule(FIELD("lRight"), INCREMENT_L(0), long.class)
    		.rule(FIELD("random"), RANDOM(10), int.class)
    		.rule(FIELD("random"), PEEK(System.out::println), int.class)
    		.generateBy(SeqIncObject.class)
    		.cache()
    		.getValueList(SeqIncObject.class);      
    	

License

This project is licensed under the terms of the MIT license.