Question

我正在使用带有OpenCSV的Spring Batch Tasklet来读取我的CSV文件。在问这个问题之前，我知道很多块，但是在以后的步骤中文件之间会进行交叉验证，因此我必须继续使用Tasklet。

我想做的是向我的报告步骤报告丢失的文件或解析错误。我不确定应该向下一步报告失败的正确方法。我有以下代码。

读取文件的初始步骤。

public class CsvBatchReader<T> implements Tasklet, StepExecutionListener {

    private final Logger logger = LoggerFactory.getLogger(CsvBatchReader.class);

    private List batch;

    private final Class<T> clazz;

    private Path path;

    public CsvBatchReader(Class<T> clazz, Path path) {
        this.clazz = clazz;
        this.path = path;
    }

    @Override
    public void beforeStep(StepExecution stepExecution) {
        logger.info("Reader initialized - " + clazz.getSimpleName());

        batch = new ArrayList();
    }

    @Override
    public ExitStatus afterStep(StepExecution stepExecution) {
        logger.info("Reader ended - " + clazz.getSimpleName());
        return ExitStatus.COMPLETED;
    }

    @Override
    public RepeatStatus execute(StepContribution contribution, ChunkContext chunkContext) throws UnexpectedJobExecutionException {
        logger.info("Reader execute - " + clazz.getSimpleName());


        ICsvToBean csvToBean = new CsvToBean(clazz, path);

        try {
            batch = csvToBean.readCsv();
        } catch(IOException ex) {
            // error message being caught from my csvToBean class. 
            throw new UnexpectedJobExecutionException("Invalid file " + ex.getMessage());
        }

        return RepeatStatus.FINISHED;
    }

}

报告步骤

我不确定如何传递异常消息，或者是否有定义的方法来传递故障消息而不使用步骤执行上下文。

public class CsvBatchReporting implements Tasklet, StepExecutionListener {

    private final Logger logger = LoggerFactory.getLogger(CsvBatchCrossValidation.class);

    private List errorMessages;
    private List skippedInserts;

    @Override
    public void beforeStep(StepExecution stepExecution) {
        logger.info("Reporting initialized");

        ExecutionContext executionContext = stepExecution
                .getJobExecution()
                .getExecutionContext();

        System.out.println("description " + stepExecution.getStatus());


    }

    @Override
    public ExitStatus afterStep(StepExecution stepExecution) {
        logger.info("Reporting ended");
        return ExitStatus.COMPLETED;
    }

    @Override
    public RepeatStatus execute(StepContribution contribution, ChunkContext chunkContext) throws Exception {
        logger.info("Reporting execute");

        //Email Error 

        return RepeatStatus.FINISHED;
    }

}

作业配置

@Bean
public Job primaryCareJob(@Qualifier("reportingStep") Step reportingStep, @Qualifier("crossValidationStep") Step crossValidationStep) {
    logger.info("Start PrimaryCare Job");

    return jobs.get("primaryCareJob")
            .start(readPrimaryCareStep()).on("FAILED").to(reportingStep)
            .from(readPrimaryCareStep()).on("*").to(readPrimaryCareDetailStep())

            .from(readPrimaryCareDetailStep()).on("FAILED").to(reportingStep)
            .from(readPrimaryCareDetailStep()).on("*").to(processPrimaryCareStep())

            .from(processPrimaryCareStep()).on("INVALID").to(reportingStep)
            .from(processPrimaryCareStep()).on("*").to(processPrimaryCareDetailStep())

            .from(processPrimaryCareDetailStep()).on("INVALID").to(reportingStep)
            //Other steps

            .from(reportingStep).on("*").end()
            .from(reportingStep).on("*").fail()

            .build()
        .build();
}

我开始将工作模式更改为失败，而不是将其定义为无效，以获取异常以自动调用失败的步骤。在我的afterStep中，使用以下代码定义了我使用无效的其他步骤。

if(!errorMessages.isEmpty()) {
            chunkContext.getStepContext().getStepExecution().setExitStatus(new ExitStatus("INVALID"));
}

如何从阅读器中获取CSV异常消息以传递到我的报告步骤中，以便我可以将其作为电子邮件发送？

Answer 1

我不确定如何在不使用“步骤执行上下文”的情况下传递异常消息或是否定义了传递失败消息的方法。

您可以访问作业执行中上一步中引发的异常。这是一个示例：

import java.util.List;

import org.springframework.batch.core.Job;
import org.springframework.batch.core.JobExecution;
import org.springframework.batch.core.JobParameters;
import org.springframework.batch.core.Step;
import org.springframework.batch.core.StepExecution;
import org.springframework.batch.core.configuration.annotation.EnableBatchProcessing;
import org.springframework.batch.core.configuration.annotation.JobBuilderFactory;
import org.springframework.batch.core.configuration.annotation.StepBuilderFactory;
import org.springframework.batch.core.launch.JobLauncher;
import org.springframework.batch.repeat.RepeatStatus;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.context.ApplicationContext;
import org.springframework.context.annotation.AnnotationConfigApplicationContext;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

@Configuration
@EnableBatchProcessing
public class MyJob {

    @Autowired
    private JobBuilderFactory jobs;

    @Autowired
    private StepBuilderFactory steps;

    @Bean
    public Step step1() {
        return steps.get("step1")
                .tasklet((contribution, chunkContext) -> {
                    System.out.println("hello");
                    throw new Exception("Boom!");
                })
                .build();
    }

    @Bean
    public Step step2() {
        return steps.get("step2")
                .tasklet((contribution, chunkContext) -> {
                    JobExecution jobExecution = chunkContext.getStepContext().getStepExecution().getJobExecution();
                    StepExecution stepExecution = jobExecution.getStepExecutions().iterator().next(); // TODO properly get the stepExecution of the previous step
                    List<Throwable> failureExceptions = stepExecution.getFailureExceptions();
                    if (!failureExceptions.isEmpty()) {
                        Throwable throwable = failureExceptions.get(0);
                        System.out.println("Looks like step1 has thrown an exception: " + throwable.getMessage());
                    }
                    System.out.println("world");
                    return RepeatStatus.FINISHED;
                })
                .build();
    }

    @Bean
    public Job job() {
        return jobs.get("job")
                    .flow(step1())
                    .on("*").to(step2())
                    .build()
                .build();
    }

    public static void main(String[] args) throws Exception {
        ApplicationContext context = new AnnotationConfigApplicationContext(MyJob.class);
        JobLauncher jobLauncher = context.getBean(JobLauncher.class);
        Job job = context.getBean(Job.class);
        jobLauncher.run(job, new JobParameters());
    }

}

此样本打印：

hello
Looks like step1 has thrown an exception: Boom!
world

很显然，您需要确保在所有情况下步骤1都流到步骤2（因此定义流程）。

希望这会有所帮助。

Answer 2

考虑将errorMessages制成bean：

// somewhere convenient...
@Bean
public List<String> errorMessages() {
    return new ArrayList<>();
}

并将errorMessages注入两个小任务中。

在CsvBatchReader任务集中，检查execute(...)引发的任何异常，并根据需要更新errorMessages：

public class CsvBatchReader<T> implements Tasklet, StepExecutionListener {

    @Override
    public ExitStatus afterStep(StepExecution stepExecution) {
        List<Throwable> failures = stepExecution.getFailureExceptions();
        if (!failures.isEmpty())  {
            errorMessages.add(...);
        }

        logger.info("Reader ended - " + clazz.getSimpleName());
        return ExitStatus.COMPLETED;  // return INVALID if !failures.isEmpty() ?
    }
}

由于errorMessages被注入CsvBatchReporting中，因此：

public class CsvBatchReporting implements Tasklet, StepExecutionListener {

@Override
public RepeatStatus execute(StepContribution contribution, ChunkContext chunkContext) throws Exception {
    logger.info("Reporting execute");

    //Email Error 
    String body = null;
    String subject = "job finished: ";
    if (!errorMessages.isEmpty())  {
       subject += "ERROR";
       body = ... // from errorMessages
    } else {
       subject += "SUCCESS";
    }

    return RepeatStatus.FINISHED;
}

}

Answer 3

尽管我最初的问题是将异常从一个步骤传递到下一步骤，但我想指出一种使用exitStatus

的替代方法

在阅读器步骤中，我可以捕获我的csvToBean异常并像这样创建一个存在状态。

@Override
public RepeatStatus execute(StepContribution contribution, ChunkContext chunkContext) throws UnexpectedJobExecutionException {
    logger.info("Reader execute - " + clazz.getSimpleName());

    ICsvToBean csvToBean = new CsvToBean(clazz, path);

    try {
        batch = csvToBean.readCsv();
    } catch(IOException ex) {
        chunkContext.getStepContext().getStepExecution().setExitStatus(new ExitStatus("FAILED", ex.getMessage()));
    }

    return RepeatStatus.FINISHED;
}

然后我将像这样在我的工作中创建条件流。

@Bean
public Job primaryCareJob(@Qualifier("reportingStep") Step reportingStep, @Qualifier("crossValidationStep") Step crossValidationStep) {
    logger.info("Start PrimaryCare Job");

    return jobs.get("primaryCareJob")
            .start(readPrimaryCareStep()).on("FAILED").to(reportingStep)
            .from(readPrimaryCareStep()).on("*").to(readPrimaryCareDetailStep())

            .from(readPrimaryCareDetailStep()).on("FAILED").to(reportingStep)
            .from(readPrimaryCareDetailStep()).on("*").to(processPrimaryCareStep())

            .from(processPrimaryCareStep()).on("INVALID").to(reportingStep)
            .from(processPrimaryCareStep()).on("*").to(processPrimaryCareDetailStep())

            .from(processPrimaryCareDetailStep()).on("INVALID").to(reportingStep)
            //Other steps

            .from(reportingStep).on("*").end()
            .from(reportingStep).on("*").fail()

            .build()
        .build();
}

最后，在我的阅读器小任务步骤中，我将像这样检索存在状态

@Override
public RepeatStatus execute(StepContribution contribution, ChunkContext chunkContext) throws Exception {
    JobExecution jobExecution = chunkContext.getStepContext().getStepExecution().getJobExecution();

    Collection<StepExecution> stepExecutions = jobExecution.getStepExecutions();

    Map<String, String> result = stepExecutions.stream()
        .filter(x -> x.getExitStatus().getExitCode().equals("FAILED"))
        .collect(Collectors.toMap(StepExecution::getStepName, x -> x.getExitStatus().getExitDescription()));

    result.forEach((k, v) -> {
        System.out.println("Step " + k + " failure " + v);
    });
}

将Spring Batch Tasklet失败消息传递给报告步骤。

3 个答案: