com.neurosys.q.service

Interface BatchService<B extends Batch<T>,T extends Token<T>>

Type Parameters:

T - as Token

All Superinterfaces:

TokenContainerService<T,B>

public interface BatchService<B extends Batch<T>,T extends Token<T>>
extends TokenContainerService<T,B>

A generic definitions on the operations that are supported on a Batch

Author:

Arjun Dhar

Method Summary

Modifier and Type	Method and Description
List<T>	add(B batch, List<T> token, Consumer<T> consumer, TokenEvent<T> tokenEvent)
B	determineRoute(T token) Determine Batch for a given Token
Batch<T>	getBatch(String id) Fetch a Batch by Id
List<T>	getInProcess(B tokenContainer) Tokens that out of the Q but being processed
default Integer	getPermissableBatchSize() How many tokens to support per Batch
List<T>	getPreProcess(B tokenContainer) Tokens that are not yet eligible to be in Q but are being considered.
List<T>	getProcessed(B tokenContainer) Tokens that out of the Q and processed
List<T>	getReadyToBeProcessed(B tokenContainer, T lastToken) Get tokens ready to be processed; from lastToken (excluding it)
List<T>	getTokens(B tokenContainer) Get all tokens ever
List<T>	getTokens(B tokenContainer, T lastToken, Collection states) Fetch tokens across multiple states @ once
T	poll(B batch) Poll then next token READY to be served.
default Number	size(B batch) # of tokens Tokens that are in the TokenContainer.

Methods inherited from interface com.neurosys.q.service.TokenContainerService

getAssignedContainer, getMaxResultSize, getPositions, getQStatesStrategy, getSequence, getType, hasNext, init, linkTokens, lookup, lookupById, lookupBySequence, lookupBySequence, move, poll, poll, poll, remove, save, update, updateStatus, updateStatus

Method Detail

getBatch

Batch<T> getBatch(String id)

Fetch a Batch by Id

poll

T poll(B batch)

Poll then next token READY to be served. A State of READY is defined by QStatesStrategy.getReadyToBeProcessed()

Specified by:

poll in interface TokenContainerService<T extends Token<T>,B extends Batch<T>>

Parameters:

batch - as container type

Returns:

next Token within the container. null if none

See Also:

Queue.poll()

add

List<T> add(B batch,
            List<T> token,
            Consumer<T> consumer,
            TokenEvent<T> tokenEvent)
     throws TokenProcessingException

Specified by:

add in interface TokenContainerService<T extends Token<T>,B extends Batch<T>>

Parameters:

batch - as container type

token - as List of Tokens

consumer - as Consumer to apply an anonymous function at runtime over the Token elements iterated; after they are added. nullable true

tokenEvent - as TokenEvent

Returns:

List of Tokens that were added

Throws:

TokenProcessingException

See Also:

Queue.add(Object)

determineRoute

B determineRoute(T token)
          throws TokenProcessingException

Determine Batch for a given Token

Parameters:

token -

Returns:

Throws:

TokenProcessingException

getReadyToBeProcessed

List<T> getReadyToBeProcessed(B tokenContainer,
                              T lastToken)

Get tokens ready to be processed; from lastToken (excluding it)

Parameters:

tokenContainer - as container type

lastToken - as Token type , as last token the calling API received. This is optional. Since a Q is constantly @ risk of being modified this ensures error correction.

Returns:

List of Tokens

See Also:

QStatesStrategy

getPreProcess

List<T> getPreProcess(B tokenContainer)

Tokens that are not yet eligible to be in Q but are being considered.

Parameters:

tokenContainer - as container type

Returns:

List of Tokens

See Also:

QStatesStrategy

getInProcess

List<T> getInProcess(B tokenContainer)

Tokens that out of the Q but being processed

Parameters:

tokenContainer - as container type

Returns:

List of Tokens

See Also:

QStatesStrategy, #getProcessed(TokenContainer)

getProcessed

List<T> getProcessed(B tokenContainer)

Tokens that out of the Q and processed

Parameters:

tokenContainer - as container type

Returns:

List of Tokens

See Also:

QStatesStrategy, #getInProcess(TokenContainer)

getTokens

List<T> getTokens(B tokenContainer,
                  T lastToken,
                  Collection states)

Fetch tokens across multiple states @ once

Parameters:

tokenContainer - as container type

lastToken - as Token type , as last token the calling API received. This is optional. Since a Q is constantly @ risk of being modified this ensures error correction.

states - as Collection of STATEs

Returns:

List of Tokens

getTokens

List<T> getTokens(B tokenContainer)

Get all tokens ever

Parameters:

tokenContainer - as container type

Returns:

List of Tokens

size

default Number size(B batch)

# of tokens Tokens that are in the TokenContainer. If the container is a finite state container then it will be the actual size of it, else if a stream then the count of tokens in #getReadyToBeProcessed(TokenContainer, Token). This is an idempotent operation.

Specified by:

size in interface TokenContainerService<T extends Token<T>,B extends Batch<T>>

Parameters:

batch - as TokenContainer type

Returns:

Number

getPermissableBatchSize

default Integer getPermissableBatchSize()

How many tokens to support per Batch

Can be null:

true, means all possible

Default value:

10