如何处理fixedLengthformat数据?

时间:2019-05-15 09:48:30

标签: apache-camel bindy

我处理一些固定长度的数据。 因此,我使用绑定组件来处理这些数据。

此文件只有一个记录。 该记录具有页眉,多正文和页脚。

  header record (total length : 20)
  1 : record_type (length : 1)
  VOLTE : service_type (length : 5)
  20190515 : creation date (length : 8)
  3 : custom_flag (length : 6)

  3 body records (total length : 20)
  2 : record_type (length : 1)
  01012345678 : mobile number (length : 11)
  20190515 : call start date (length : 8)

  footer records (total length : 20)
  3 : record_type (length : 1)
  AAAA.DAT : FILE NAME (length : 19)

真实数据

  1VOLTE20190515     32010123456782019051520101234567820190516201012345678201905173AAAA.DAT           

我定义了如下数据格式。

标题

  @FixedLengthRecord(length=20, paddingChar=' ')
  public class VoLTEHeader {

  @DataField(pos=1, length=1, trim=true)
  String record_type;

  @DataField(pos=2, length=5, trim=true)
  String service_type;

  @DataField(pos=7, length=8, trim=true)
  String creation_date;

  @DataField(pos=15, length=6, trim=true, align="R")
  String custom_flag;

  @FixedLengthRecord(length=20, paddingChar=' ')
  public class VoLTEFooter {

  @DataField(pos=1, length=1, trim=true)
  String record_type;

  @DataField(pos=2, length=19, trim=true)
  String file_name;

身体

  @FixedLengthRecord(length=20, header=VoLTEHeader.class, footer=VoLTEFooter.class)
  public class VoLTEBody implements Serializable {

  @DataField(pos=1, length=1,trim=true)
  String record_type;

  @DataField(pos=2, length=11,trim=true)
  String mobile_number;

  @DataField(pos=13, length=8,trim=true)
  String call_start_date;

我执行了骆驼路线,但发生了异常。

  java.lang.IllegalArgumentException: Size of the record: 100 is not equal to the value provided in the model: 20
  at org.apache.camel.dataformat.bindy.fixed.BindyFixedLengthDataFormat.createModel(BindyFixedLengthDataFormat.java:295) ~[camel-bindy-2.23.2.jar:2.23.2]
at org.apache.camel.dataformat.bindy.fixed.BindyFixedLengthDataFormat.unmarshal(BindyFixedLengthDataFormat.java:209) ~[camel-bindy-2.23.2.jar:2.23.2]
at org.apache.camel.processor.UnmarshalProcessor.process(UnmarshalProcessor.java:69) ~[camel-core-2.23.2.jar:2.23.2]
at org.apache.camel.processor.RedeliveryErrorHandler.process(RedeliveryErrorHandler.java:548) [camel-core-2.23.2.jar:2.23.2]
at org.apache.camel.processor.CamelInternalProcessor.process(CamelInternalProcessor.java:201) [camel-core-2.23.2.jar:2.23.2]
at org.apache.camel.processor.Pipeline.process(Pipeline.java:138) [camel-core-2.23.2.jar:2.23.2]
at org.apache.camel.processor.Pipeline.process(Pipeline.java:101) [camel-core-2.23.2.jar:2.23.2]

我认为不一定需要在多行中创建fixedLengthDataFormat。

如何解决此问题?

2 个答案:

答案 0 :(得分:0)

Camel bindy将标头定义为“ 在文件/流的开头的单个标头记录”,将页脚定义为“ 在文件/流的末尾的单个页脚记录” ”,请参阅[camel-bindy doc] [1]。

您的测试数据在同一行中的多有效载荷零件之前和之后都具有元数据,因此您不能使用Bindy页眉/页脚模型来对此进行解析。

相反,为页眉,页脚和一个主体创建独立的Bindy模型(基本上从VoLTEBody中删除“ header = VoLTEHeader.class,footer = VoLTEFooter.class”),并分别处理页眉,页脚和主体:

public class MyRouteBuilderTest extends CamelTestSupport {
    @Produce(uri="direct:start")
    private ProducerTemplate producer;

    @Override
    protected RouteBuilder createRouteBuilder() {
        return new RouteBuilder() {
            public void configure() {
                DataFormat bindyHeader = new BindyFixedLengthDataFormat(VoLTEHeader.class);
                DataFormat bindyFooter = new BindyFixedLengthDataFormat(VoLTEFooter.class);
                DataFormat bindyOneBody = new BindyFixedLengthDataFormat(VoLTEBody.class);

                from("direct:start")
                    // assuming the raw record is in the body, we keep a copy in a Camel-header
                    .setHeader("record", body())

                    // get the VoLTE header string into the Camel-body and unmarshal, then put VoLTE header object into Camel-header for later use
                    .setBody().groovy("request.body.substring(0,20)")
                    .unmarshal(bindyHeader)
                    .setHeader("header", body())

                    // restore VoLTE record string to Camel-body, get the VoLTE footer string into the Camel-body and unmarshal, then put footer VoLTE object into Camel-header for later use
                    .setBody().header("record")
                    .setBody().groovy("request.body.substring(request.body.length()-20,request.body.length())")
                    .unmarshal(bindyFooter)
                    .setHeader("footer", body())

                    // restore VoLTE record string to Camel-body, get the multi-bodies string into the Camel-body
                    .setBody().header("record")
                    .setBody().groovy("request.body.substring(20,request.body.length()-20)")

                    // Split VoLTE bodies string to each 20 char length, unmarshal each and finally put the list of VoLTE body objects into a Camel-header
                    .setBody().groovy("request.body.split('(?<=\\\\G.{20})')")
                    .split(body(), AggregationStrategies.flexible().storeInHeader("bodyList").accumulateInCollection(ArrayList.class))
                        .unmarshal(bindyOneBody)
                    .end()

                    // now do something with the unmarshalled objects in Camel-headers "header" (type VoLTEHeader), "footer" (type VoLTEFooter) and "bodyList" (type List<VoLTEBody>)
                    .log("VoLTEHeader: ${header.header}")
                    .log("VoLTEBody*: ${header.bodyList}")
                    .log("VoLTEFooter: ${header.footer}")
                    ;
            }
        };
    }

    @Test
    public void test() throws Exception {
        producer.sendBody("1VOLTE20190515     32010123456782019051520101234567820190516201012345678201905173AAAA.DAT           ");
    }
}

@FixedLengthRecord(length = 20, paddingChar = ' ')
public class VoLTEHeader {
    @DataField(pos = 1, length = 1, trim = true)
    String record_type;

    @DataField(pos = 2, length = 5, trim = true)
    String service_type;

    @DataField(pos = 7, length = 8, trim = true)
    String creation_date;

    @DataField(pos = 15, length = 6, trim = true, align = "R")
    String custom_flag;

    @Override
    public String toString() {
        return String.format("VoLTEHeader[record_type=%s, service_type=%s, creation_date=%s, custom_flag=%s]", record_type, service_type, creation_date, custom_flag);
    }
}

@FixedLengthRecord(length = 20)
public class VoLTEBody {
    @DataField(pos = 1, length = 1, trim = true)
    String record_type;

    @DataField(pos = 2, length = 11, trim = true)
    String mobile_number;

    @DataField(pos = 13, length = 8, trim = true)
    String call_start_date;

    @Override
    public String toString() {
        return String.format("VoLTEBody[record_type=%s, mobile_number=%s, call_start_date=%s]", record_type, mobile_number, call_start_date);
    }
}

@FixedLengthRecord(length = 20, paddingChar = ' ')
public class VoLTEFooter {
    @DataField(pos = 1, length = 1, trim = true)
    String record_type;

    @DataField(pos = 2, length = 19, trim = true)
    String file_name;

    @Override
    public String toString() {
        return String.format("VoLTEFooter[record_type=%s, file_name=%s]", record_type, file_name);
    }
}

输出:

[main] INFO route1 - VoLTEHeader: VoLTEHeader[record_type=1, service_type=VOLTE, creation_date=20190515, custom_flag=3]
[main] INFO route1 - VoLTEBody*: [VoLTEBody[record_type=2, mobile_number=01012345678, call_start_date=20190515], VoLTEBody[record_type=2, mobile_number=01012345678, call_start_date=20190516], VoLTEBody[record_type=2, mobile_number=01012345678, call_start_date=20190517]]
[main] INFO route1 - VoLTEFooter: VoLTEFooter[record_type=3, file_name=AAAA.DAT           ]

在此路线的结尾,您应该在骆驼头“ header”中有一个VoLTEHeader类型的对象,在骆驼头“ footer”中有一个VoLTEFooter类型的对象,并且在骆驼头“ bodyList”中有一个VoLTEBody列表。 。您现在就可以使用它们进行处理了。

答案 1 :(得分:0)

谢谢。我根据您的指导实施了它。 我认为Bindy组件支持页眉,页脚和正文。

所以,我考虑了以下逻辑。

传入文件->拆分(20个长度)->汇总拆分文件->解组

我使用带有页眉,页脚和正文的聚合文件。

下面是代码。

  from("direct:start")
                    // assuming the raw record is in the body, we keep a copy in a Camel-header
                    .setHeader("record", body())
                    .setBody().groovy("request.body.split('(?<=\\\\G.{20})')")
                    .split(body(), AggregationStrategies.flexible().storeInHeader("bodyList").accumulateInCollection(ArrayList.class))
                    .unmarshal(bindyOneBody)
                    .end()
                    .log("VoLTEBody*: ${header.bodyList}")
                   ...

VoLTEBody

       @FixedLengthRecord(length=20, header=VoLTEHeader.class, footer=VoLTETailer.class)
       public class VoLTEBody {

       @DataField(pos=1, length=1,trim=true)
       String record_type;

但是有如下错误。

       Stacktrace
       ------------------------------------------------------------------------------ 
       --------------------------------------------------------- 
       java.lang.IllegalArgumentException: No records have been defined in the file
               at org.apache.camel.dataformat.bindy.fixed.BindyFixedLengthDataFormat.unmarshal(BindyFixedLengthDataFormat.java:250) ~[camel-bindy-2.23.2.jar:2.23.2]
               at org.apache.camel.processor.UnmarshalProcessor.process(UnmarshalProcessor.java:69) ~[camel-core-2.23.2.jar:2.23.2]
               at org.apache.camel.processor.RedeliveryErrorHandler.process(RedeliveryErrorHandler.java:548) [camel-core-2.23.2.jar:2.23.2]
               at org.apache.camel.processor.CamelInternalProcessor.process(CamelInternalProcessor.java:201) [camel-core-2.23.2.jar:2.23.2]
               at org.apache.camel.processor.RedeliveryErrorHandler.process(RedeliveryErrorHandler.java:548) [camel-core-2.23.2.jar:2.23.2]

如果在VoLTEBody类中正确定义了有关页眉和页脚的注释,我认为此过程很好。

如何处理此问题?

此外,我执行了另一个测试。 在这种情况下,不会拆分。

传入文件->解组(bindyOneBody)

路由器

    from("direct:start")
                    // assuming the raw record is in the body, we keep a copy in a Camel-header
                    .setHeader("record", body())
                    //.setBody().groovy("request.body.split('(?<=\\\\G.{20})')")
                    //.split(body(), AggregationStrategies.flexible().storeInHeader("bodyList").accumulateInCollection(ArrayList.class))
                    .unmarshal(bindyOneBody)

有以下结果。 在此结果中,我们找不到VoLTEHeader和VoLTEFooter结构。很好。

     2019-05-16 15:22:15,798 DEBUG BindyFixedLengthDataFormat - Graph of objects created: {camel.dataformat.volte.VoLTEBody_sample=VoLTEBody[record_type=2, mobile_number=01012345678, call_start_date=20190517]} 
     2019-05-16 15:22:15,798 INFO  route1 - [VoLTEBody[record_type=2, mobile_number=01012345678, call_start_date=20190515], VoLTEBody[record_type=2, mobile_number=01012345678, call_start_date=20190516], VoLTEBody[record_type=2, mobile_number=01012345678, call_start_date=20190517]] 
     2019-05-16 15:22:15,798 INFO  MyRouteBuilderTest - ********************************************************************************