使用分区步骤在spring批处理中访问@JobScope bean

Question

是否有方法访问分区步骤中定义为JobScope的bean？

我们将http客户端bean定义为JobScope，因为每个作业都是唯一的，但它是动态创建的，我们需要在从属步骤中使用它来发出post请求。当我们自动连线我们得到的一切

Error creating bean with name 'scopedTarget.captureErpStockTasklet':
Scope 'step' is not active for the current thread; consider defining a 
scoped proxy for this bean if you intend to refer to it from a singleton;
nested exception is java.lang.IllegalStateException: No context holder 
available for step scope

这里是作业配置类（我删除了所有不需要的内容，只剩下分区步骤和客户机类的配置，它们位于作业范围中，因为在每次作业运行时都需要特定于作业：

@Configuration
@EnableBatchProcessing
@ComponentScan(basePackages = {"com.example.importjob"})
public class FrozenFileImportJobConfig {
@Value("${thread.pool:10}")
private int threadPoolSize;
@Value("${partitioner.max.thread.pool:10}")
private int maxThreadPoolSize;
@Value("${grid.size:10}")
private int gridSize;
@Value("${client.url:some_url}")
private int url;
@Value("${client.id:client_id}")
private int clientId;
@Value("${client.secret:secret}")
private int clientSecret;
@Autowired
private JobBuilderFactory jobBuilderFactory;
@Autowired
private StepBuilderFactory stepBuilderFactory;
 @Autowired
private ErpStockIdPartitioner erpStockIdPartitioner;
@Autowired
private CaptureErpStockTasklet captureErpStockTasklet;
@Bean(destroyMethod = "destroy")
@JobScope
public Client client() {
  return new Client(url, clientId, clientSecret);
}

public ExecutionContextPromotionListener contextPromotionListener() {
  final ExecutionContextPromotionListener executionContextPromotionListener = new ExecutionContextPromotionListener();
  executionContextPromotionListener.setKeys(new String[] {"erp_stock_ids"});
  return executionContextPromotionListener;
}

public Step captureErpStockDBTaskletStep() {
  return stepBuilderFactory.get("captureErpStockDBTaskletStep").tasklet(captureErpStockTasklet).build();
}

@Bean
public ThreadPoolTaskExecutor erpStockIdTaskExecutor() {
  final ThreadPoolTaskExecutor threadPoolTaskExecutor = new ThreadPoolTaskExecutor();
  threadPoolTaskExecutor.setCorePoolSize(threadPoolSize);
  threadPoolTaskExecutor.setMaxPoolSize(maxThreadPoolSize);
  threadPoolTaskExecutor.setAllowCoreThreadTimeOut(true);
  return threadPoolTaskExecutor;
}
public Step partitioningCaptureStep() {
  return stepBuilderFactory.get("partitioningCaptureStep").partitioner(captureErpStockDBTaskletStep())
.partitioner("erpStockIdPartitioner", erpStockIdPartitioner).taskExecutor(erpStockIdTaskExecutor())
.gridSize(gridSize).allowStartIfComplete(true).build();
}
@Bean
public Job riverIslandFrozenFileImportJob() {
  return jobBuilderFactory.get("riverIslandFrozenFileImportJob").start(partitioningCaptureStep()).build();
}
}

下面是CaptureerStockDBTasklet，它作为主分区CaptureStep的分区步骤调用：

@Component
@StepScope
public class CaptureErpStockTasklet implements Tasklet {
  private static final Logger LOG = LoggerFactory.getLogger(CaptureErpStockTasklet.class);
  @Value("#{jobParameters[" + FtpJobParameters.ORGANIZATION_ID + "]}")
  private Long organizationId;
  @Autowired
  private ErpStockRepository erpStockRepository;
  @Autowired
  private Client client;
  @Override
  public RepeatStatus execute(final StepContribution contribution, final ChunkContext chunkContext) throws Exception {
    //importing of erp stock from DB, nothing special in that code so removed
     client.captureErpStock(stock);
     LOG.info("[{}] Finished importing ERP {}", organizationId, erpStock);
     return RepeatStatus.FINISHED;
  }
}

当我将Clinet的配置更改为StepScope时，它工作得很好，但我会为每个步骤创建客户端，这不是我想要的，我希望整个作业都有一个客户端。

Answer 1

我认为，当在分区步骤中使用作业范围bean时，这是spring批处理中的一个问题。

请参阅：https://jira.spring.io/browse/BATCH-2269和多线程访问Spring Batch 3.0中的作业范围bean