org.apache.beam.sdk.io

Class PubsubIO.Read

java.lang.Object

org.apache.beam.sdk.io.PubsubIO.Read

Enclosing class:

PubsubIO

public static class PubsubIO.Read
extends Object

A PTransform that continuously reads from a Cloud Pub/Sub stream and returns a PCollection of Strings containing the items from the stream.

When running with a PipelineRunner that only supports bounded PCollections (such as DirectRunner), only a bounded portion of the input Pub/Sub stream can be processed. As such, either PubsubIO.Read.Bound.maxNumRecords(int) or PubsubIO.Read.Bound.maxReadTime(Duration) must be set.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	PubsubIO.Read.Bound<T> A PTransform that reads from a Cloud Pub/Sub source and returns a unbounded PCollection containing the items from the stream.

Method Summary

Modifier and Type	Method and Description
static PubsubIO.Read.Bound<String>	idLabel(String idLabel) Creates and returns a transform for reading from Cloud Pub/Sub where unique record identifiers are expected to be provided as Pub/Sub message attributes.
static PubsubIO.Read.Bound<String>	maxNumRecords(int maxNumRecords) Creates and returns a transform for reading from Cloud Pub/Sub with a maximum number of records that will be read.
static PubsubIO.Read.Bound<String>	maxReadTime(Duration maxReadTime) Creates and returns a transform for reading from Cloud Pub/Sub with a maximum number of duration during which records will be read.
static PubsubIO.Read.Bound<String>	subscription(String subscription) Creates and returns a transform for reading from a specific Cloud Pub/Sub subscription.
static PubsubIO.Read.Bound<String>	timestampLabel(String timestampLabel) Creates and returns a transform reading from Cloud Pub/Sub where record timestamps are expected to be provided as Pub/Sub message attributes.
static PubsubIO.Read.Bound<String>	topic(String topic) Creates and returns a transform for reading from a Cloud Pub/Sub topic.
static <T> PubsubIO.Read.Bound<T>	withCoder(Coder<T> coder) Creates and returns a transform for reading from Cloud Pub/Sub that uses the given Coder to decode Pub/Sub messages into a value of type T.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

topic

public static PubsubIO.Read.Bound<String> topic(String topic)

Creates and returns a transform for reading from a Cloud Pub/Sub topic. Mutually exclusive with subscription(String).

See PubsubIO.PubsubTopic.fromPath(String) for more details on the format of the topic string.

Dataflow will start reading data published on this topic from the time the pipeline is started. Any data published on the topic before the pipeline is started will not be read by Dataflow.

subscription

public static PubsubIO.Read.Bound<String> subscription(String subscription)

Creates and returns a transform for reading from a specific Cloud Pub/Sub subscription. Mutually exclusive with topic(String).

See PubsubIO.PubsubSubscription.fromPath(String) for more details on the format of the subscription string.

timestampLabel

public static PubsubIO.Read.Bound<String> timestampLabel(String timestampLabel)

Creates and returns a transform reading from Cloud Pub/Sub where record timestamps are expected to be provided as Pub/Sub message attributes. The timestampLabel parameter specifies the name of the attribute that contains the timestamp.

The timestamp value is expected to be represented in the attribute as either:

• a numerical value representing the number of milliseconds since the Unix epoch. For example, if using the Joda time classes, Instant.getMillis() returns the correct value for this attribute.
• a String in RFC 3339 format. For example, 2015-10-29T23:41:41.123Z. The sub-second component of the timestamp is optional, and digits beyond the first three (i.e., time units smaller than milliseconds) will be ignored.

If timestampLabel is not provided, the system will generate record timestamps the first time it sees each record. All windowing will be done relative to these timestamps.

By default, windows are emitted based on an estimate of when this source is likely done producing data for a given timestamp (referred to as the Watermark; see AfterWatermark for more details). Any late data will be handled by the trigger specified with the windowing strategy – by default it will be output immediately.

Note that the system can guarantee that no late data will ever be seen when it assigns timestamps by arrival time (i.e. timestampLabel is not provided).

See Also:

RFC 3339

idLabel

public static PubsubIO.Read.Bound<String> idLabel(String idLabel)

Creates and returns a transform for reading from Cloud Pub/Sub where unique record identifiers are expected to be provided as Pub/Sub message attributes. The idLabel parameter specifies the attribute name. The value of the attribute can be any string that uniquely identifies this record.

If idLabel is not provided, Dataflow cannot guarantee that no duplicate data will be delivered on the Pub/Sub stream. In this case, deduplication of the stream will be strictly best effort.

withCoder

public static <T> PubsubIO.Read.Bound<T> withCoder(Coder<T> coder)

Creates and returns a transform for reading from Cloud Pub/Sub that uses the given Coder to decode Pub/Sub messages into a value of type T.

By default, uses StringUtf8Coder, which just returns the text lines as Java strings.

Type Parameters:

T - the type of the decoded elements, and the elements of the resulting PCollection.

maxNumRecords

public static PubsubIO.Read.Bound<String> maxNumRecords(int maxNumRecords)

Creates and returns a transform for reading from Cloud Pub/Sub with a maximum number of records that will be read. The transform produces a bounded PCollection.

Either this option or maxReadTime(Duration) must be set in order to create a bounded source.

maxReadTime

public static PubsubIO.Read.Bound<String> maxReadTime(Duration maxReadTime)

Creates and returns a transform for reading from Cloud Pub/Sub with a maximum number of duration during which records will be read. The transform produces a bounded PCollection.

Either this option or maxNumRecords(int) must be set in order to create a bounded source.