What is the right way to create Datastore Entity and use it in Dataflow pipeline

Question

Which of these is the correct / preferred way to create Datastore Entity:

    // First, create a fake email for our fake user
    final String email = "[email protected]";

    // Now, create a 'key' for that user using the email
    final Key userKey =
            datastore.newKeyFactory().kind("user").newKey(email);

    // Now create a entity using that key adn add some fields to it
    final Entity newUser =
            Entity
                    .builder(userKey)
                    .set("name", "Faker")
                    .set("email", email)
                    .build();

or like it's done in DatastoreWordCount example ?

Entity from the first example is com.google.cloud.datastore.Entity.
Entity from the second example is com.google.datastore.v1.Entity.

I have code using com.google.cloud.datastore.Entity and I don't know how to store it Datastore as a part of Dataflow pipeline, since all code examples I found online suggest:

com.google.cloud.dataflow.sdk.io.datastore.DatastoreIO.v1().write()

but it's only working with com.google.datastore.v1.Entity.

I'm using com.google.cloud.dataflow:google-cloud-dataflow-java-sdk-all:1.7.0.

Kenn Knowles Kenn Knowles · Accepted Answer · 2016-10-11T20:55:16

You have analyzed the situation well - Dataflow's DatastoreIO.v1() works with com.google.datastore.v1.Entity.

What is the right way to create Datastore Entity and use it in Dataflow pipeline

1 Answers