Spring Batch简单入门（三）- ItemReader 与 ItemWriter

接上一章，Spring Batch本身提供了许多开箱即用的ItemReader与ItemWriter实现。Spring Batch 4针对这些组件又提供了Builder实现，用户可以很方便的通过Builder模式来使用这些组件。

本章，我们重点讨论以下几种ItemReader和ItemWriter的使用。

ItemReader

• FlatFileItemReader

对于FlatFileItemReader，官方示例已给出代码实现。我们这里不多赘述。

Reads from a flat file. Includes ItemStream and Skippable functionality

    @Bean
    public FlatFileItemReader flatFileItemReader() {
        FlatFileItemReader reader = new FlatFileItemReader<>();
        DefaultLineMapper lineMapper = new DefaultLineMapper() {
            {
                setLineTokenizer(new DelimitedLineTokenizer() {
                    {
                        setNames(new String[]{"firstName", "lastName"});
                    }
                });
                setFieldSetMapper(new BeanWrapperFieldSetMapper() {
                    {
                        setTargetType(User.class);
                    }
                });
            }
        };
        reader.setResource(new ClassPathResource("sample-data.csv"));
        reader.setLineMapper(lineMapper);
        return reader;
    }

• JpaPagingItemReader

Given a JPQL statement, pages through the rows, such that large datasets can be read without running out of memory.

以JPQL方式读取海量数据库记录，为避免内存泄漏，以分页方式读取。

    @Bean(destroyMethod = "")
    @Qualifier("jpaPagingItemReader")
    public ItemReader jpaPagingItemReader() {
        JpaPagingItemReader reader = new JpaPagingItemReader();
        String sqlQuery = "select * from user where id like :limit ";
        JpaNativeQueryProvider queryProvider = new JpaNativeQueryProvider();
        queryProvider.setSqlQuery(sqlQuery);
        queryProvider.setEntityClass(User.class);
        reader.setEntityManagerFactory(emf);
        reader.setPageSize(3);
        reader.setQueryProvider(queryProvider);
        reader.setParameterValues(Collections.singletonMap("limit", "%"));
        reader.setSaveState(true);
        return reader;
    }

• RepositoryItemReader

Given a Spring Data PagingAndSortingRepository object, a Sort, and the name of method to execute, returns items provided by the Spring Data repository implementation.

以Spring Data JPA(Spring Data Repository)方式读取数据，传入参数包含：PagingAndSortingRepository对象，结果集排序方式，调用目标方法等。返回Spring Data JPA方法实现执行结果。

比如，对于UserReaderRepository：

public interface UserReaderRepository extends PagingAndSortingRepository<User, Long> {
    @Override
    Page findAll(Pageable pageable);

    Page findAllByFirstNameLike(String firstName, Pageable pageable);
}

接口中定义findAllByFirstNameLike方法，查询FirstName以特定字符开始的数据库记录，因此，我们的RepositoryItemReader定义如下：

    @Bean
    @Qualifier("repositoryItemReaderWithParams")
    public RepositoryItemReader<User> repositoryItemReaderWithParams() {
        Map<String, Sort.Direction> map = new HashMap<>();
        map.put("id", Sort.Direction.DESC);
        List<String> params = new ArrayList();
        params.add("i%");
        RepositoryItemReader<User> repositoryItemReader = new RepositoryItemReader<>();
        repositoryItemReader.setRepository(userReaderRepository);
        repositoryItemReader.setPageSize(5);
        repositoryItemReader.setMethodName("findAllByFirstNameLike");
        repositoryItemReader.setArguments(params);
        repositoryItemReader.setSort(map);
        return repositoryItemReader;
    }

ItemWriter

• JdbcBatchItemWriter

Uses batching features from a PreparedStatement, if available, and can take rudimentary steps to locate a failure during a flush.

利用PreparedStatement批处理特性保存数据。

    @Bean
    @Qualifier("jdbcBatchItemWriter")
    public JdbcBatchItemWriter jdbcBatchItemWriter() {
        JdbcBatchItemWriter writer = new JdbcBatchItemWriter<>();
        writer.setItemSqlParameterSourceProvider(new BeanPropertyItemSqlParameterSourceProvider<>());
        writer.setSql("insert into people (person_id, first_name, last_name) VALUES (uuid(), :firstName, :lastName)");
        writer.setDataSource(w_dataSource);
        return writer;
    }

• RepositoryItemWriter

Given a Spring Data CrudRepository implementation, items are saved through the method specified in the configuration.

调用Spring Data JPA (Spring Data CrudRepository)指定方法实现，将数据保存到目标数据库中。比如，我们定义PeopleCrudRepository接口：

public interface PeopleCrudRepository extends CrudRepository<People, Long> {
}

我们调用CrudRepository中的默认save方法实现。

    @Bean
    @Qualifier("repositoryItemWriter")
    public RepositoryItemWriter repositoryItemWriter() {
        RepositoryItemWriter peopleRepositoryItemWriter = new RepositoryItemWriter<>();
        peopleRepositoryItemWriter.setRepository(peopleCrudRepository);
        peopleRepositoryItemWriter.setMethodName("save");
        return peopleRepositoryItemWriter;
    }

更多源码请参考:

https://github.com/ypmc/spring-cloud/tree/master/spring-batch