In the disjunctive case this means that as data flows through the stream if it becomes a
Left it will no longer be
processed by the disjunctive stages and will pass through continuing further down stream.
Add a dependency in your
val fleamVersion = "7.0.0" ... libraryDependencies ++= Seq( "com.nike.fleam" %% "fleam" % fleamVersion, "com.nike.fleam" %% "fleam-aws-sqs" % fleamVersion, "com.nike.fleam" %% "fleam-aws-cloudwatch" % fleamVersion)
eitherMap- Applies a function to items that are Right where only the right-hand value is passed and the result is placed back into the Either
R => R1
eitherFlatMap- Applies a function to items that are Right where only the right-hand value is passed and the result is a new either.
R => Either[L, R1]
eitherFlatMapAsyncUnordered- Future based versions of
eitherFlatMapthat preserve item order or not.
flatten- On an Either turns a Right Iterable into individual Rights
flatMapConcat- Takes a function from
Either[L, R]that creates a Source of eithers to be introduced into the stream.
broadcastMerge- process an item through a collection of flows in parallel and merge the results back into the stream
joinRight- Join nested eithers to the right a stream of
Either[L, [Either[L, R]]becomes
joinLeft- Join nested eithers to the left a stream of
Either[Either[L, R], R]becomes
eitherMap, but takes a flow instead of a function
viaLeft- process left values through a flow
biVia- Process Lefts and Rights through different flow and merge the results back into the stream
StreamDaemona class that manages starting and cleanly stopping a stream.
- Type classes for stream based metrics logging whereby data is created as part of the stream.
tickingGroupedWithin- Like a normal
groupedWithinexcept will emit an empty Seq after the elapsed
withinduration even if nothing has been received.
SerializedByKeyBidi- a BidiFlow that limits items by a key to serial processing. For example an items of key
Awill while have to complete processing before another item of key
Awill be processed. This helps prevent concurrent operations for a key.
Valve- Slows processing during failed downstream services instead of failing fast.
ContainsCount- typeclass to require the ability to extract a count for an item. Used to track repeated trips through a processing stream.
- Case class based configuration for common parameters,
- Enrichments to convert a function that returns a future into a flow
- Enrichments to help with processing tuples through flows
Fleam SQS is a library of classes to aid in processing AWS SQS messages in a functional manner. In practice this means providing operations that are complete and explicitly handling retry and dead-letter scenarios instead of relying on message timeouts.
ContainsRetrievedMessage- typeclass to require an SQS message is extractable in an item
ToMessage- typeclass to turn an object into a new SQS Message
MessageAttributes- provides an instance of
ContainsCountthat stores a count in an SQS message's messageAttributes
RetrievedTime- typeclass which requires a retrieved time for the message and calculates the elapsed time.
Sourcefor reading messages from an SQS Queue.
SqsDelete- flow to delete SQS message individually or in batches.
SqsEnqueue- flow to enqueue SQS messages individually or in batches.
SqsReduce- combines messages within a grouping window with the same key into a single message and deletes the duplicate messages.
SqsRetry- explicit handling of retry and dead-letters. Takes two partial functions to define each group. Provides options for back-off, retry count, timeout, error inclusion in the message, and message duplication id modification.
- Case class based Configuration
toMessageAttributes- extension on tuples to create maps of MessageAttributeValue
Showinstances for logging SqsRetry Errors
Provides a class to create a flow which logs a count to Cloudwatch as part of the stream. Often used to create a metric of items processed.
Enhancements to streams can be imported using
SQS specific enhancements can be imported using
- Turning Async Function into Flows
- Using Tuple-2 Flow Helpers
- Using Valves to slow processing when external system are unavailable
- Processing SQS pipelines
- Logging stream metrics
- Logging with disjunction flows
- Using SqsRetry to manage your retry and dead-letter policies
See our Contributing guidelines.