[ 
https://issues.apache.org/jira/browse/BEAM-5413?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Jeff Klukas updated BEAM-5413:
------------------------------
    Description: 
Defining a composite transform today requires writing a full named subclass of 
PTransform (as [the programming guide 
documents|[https://beam.apache.org/documentation/programming-guide/#composite-transforms]]
 but there are cases where users may want to define a fairly trivial composite 
transform using a less verbose Java 8 lambda expression.

Consider an example where the user has defined MyDeserializationTransform that 
attempts to deserialize byte arrays into some object, returning a 
PCollectionTuple  with tags for successfully deserialized records (mainTag) and 
for errors (errorTag).

If we introduce a PTransform::compose method that takes in a 
SerializableFunction, the user can handle errors in a small lambda expression:

 
{code:java}
byteArrays
    .apply("attempt to deserialize messages", 
        new MyDeserializationTransform())
    .apply("write deserialization errors",
        PTransform.compose((PCollectionTuple input) -> {
            input
              .get(errorTag)
              .apply(new MyErrorOutputTransform());
            return input.get(mainTag);
        })
    .apply("more processing on the deserialized messages", 
         new MyOtherTransform())
{code}
This style allows a more concise and fluent pipeline definition than is 
currently possible.

 

  was:
Defining a composite transform today requires writing a full named subclass of 
PTransform (as [the programming guide 
documents|[https://beam.apache.org/documentation/programming-guide/#composite-transforms]),]
 but there are cases where users may want to define a fairly trivial composite 
transform using a less verbose Java 8 lambda expression.

Consider an example where the user has defined MyDeserializationTransform that 
attempts to deserialize byte arrays into some object, returning a 
PCollectionTuple  with tags for successfully deserialized records (mainTag) and 
for errors (errorTag).

If we introduce a PTransform::compose method that takes in a 
SerializableFunction, the user can handle errors in a small lambda expression:

 
{code:java}
byteArrays
    .apply("attempt to deserialize messages", 
        new MyDeserializationTransform())
    .apply("write deserialization errors",
        PTransform.compose((PCollectionTuple input) -> {
            input
              .get(errorTag)
              .apply(new MyErrorOutputTransform());
            return input.get(mainTag);
        })
    .apply("more processing on the deserialized messages", 
         new MyOtherTransform())
{code}
This style allows a more concise and fluent pipeline definition than is 
currently possible.

 


> Add method for defining composite transforms as lambda expressions
> ------------------------------------------------------------------
>
>                 Key: BEAM-5413
>                 URL: https://issues.apache.org/jira/browse/BEAM-5413
>             Project: Beam
>          Issue Type: Improvement
>          Components: sdk-java-core
>            Reporter: Jeff Klukas
>            Assignee: Kenneth Knowles
>            Priority: Minor
>             Fix For: 2.8.0
>
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> Defining a composite transform today requires writing a full named subclass 
> of PTransform (as [the programming guide 
> documents|[https://beam.apache.org/documentation/programming-guide/#composite-transforms]]
>  but there are cases where users may want to define a fairly trivial 
> composite transform using a less verbose Java 8 lambda expression.
> Consider an example where the user has defined MyDeserializationTransform 
> that attempts to deserialize byte arrays into some object, returning a 
> PCollectionTuple  with tags for successfully deserialized records (mainTag) 
> and for errors (errorTag).
> If we introduce a PTransform::compose method that takes in a 
> SerializableFunction, the user can handle errors in a small lambda expression:
>  
> {code:java}
> byteArrays
>     .apply("attempt to deserialize messages", 
>         new MyDeserializationTransform())
>     .apply("write deserialization errors",
>         PTransform.compose((PCollectionTuple input) -> {
>             input
>               .get(errorTag)
>               .apply(new MyErrorOutputTransform());
>             return input.get(mainTag);
>         })
>     .apply("more processing on the deserialized messages", 
>          new MyOtherTransform())
> {code}
> This style allows a more concise and fluent pipeline definition than is 
> currently possible.
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to