Powerful tool for collection processing

Why Stream

  Assemble in Java Is the most widely used , Almost every Java Applications make and process collections , It can be said that set is the most basic unit of programming task . But such basic unit routine operation （ For example, data grouping 、 Screening 、 Sort, etc ）, did not API Support . And for the processing of set elements when there is no logical relationship , There is no parallel way to utilize multi-core mechanisms .

  that ,Java The designer of the language designed the stream , Make it easier for each engineer to deal with the set 、 Operation set .

Stream Detailed explanation

 Stream yes Java8 Provided API, Deal with data sets declaratively , More efficient processing , You can perform chain flow operations on collections .

 

Stream All operations of can be classified ：

• Intermediate operation

  Only one mark is recorded in the intermediate operation , Only the termination operation will trigger the actual calculation .

• No state ： It means that the element is not affected by the previous element , There is no association between elements

• A stateful ： Stateful intermediate operations must wait until all elements are processed before the final result is known

• Termination operation

  The termination operation is the operation to get the final calculation result .

• Short circuit operation ： You can return results without processing all the elements

• Non short circuit operation ： All elements must be processed to get the final result

Stream principle

 Stream Inside is through Pipeline The way to achieve , The basic idea is to perform as many operations as possible along the pipeline during iteration , Thus avoiding multiple iterations . Record every step of the user's operation in some way , When the user invokes the termination operation, the previous recording operations are superimposed , Perform all in another iteration .

among , The core issue is ：

• How to record user actions

• How operations stack

• How to perform the superimposed operation

How the operation records

  The photo Stream Inheritance relationship between related classes and interfaces ,Stream adopt Stage To describe a complete operation , And use some instantiated PipelineHelper To represent the Stage, Will have the order of each Stage Serial together , Make up the whole assembly line .

 

  Calling Stream() When the method is used , Will create a Head Instance to represent the header of the flow operation . Then call a series of intermediate operations , It's constantly generating new Stream. these Stream Objects are organized in the form of a two-way linked list , Make up the whole assembly line . Because each Stage All recorded the previous one Stage And this operation and destroy the function , Depending on this structure, all operations on the data source can be established .

How operations stack

  adopt Slin Interface . Every Stage All encapsulate their operations in one Slink in , Previous Stage Just call the latter Stage Of accept() The method can . actually Stream API The essence of internal implementation is how to rewrite Slink These four methods of the interface .

• First begin() The method tells Slink The number of elements participating in the operation , It is convenient to determine the size of the intermediate result container

• After through accept() Method to add elements to the intermediate results , At the end of execution, the caller will call the method continuously , Until you go through all the elements

• Last end() The method tells Slink All elements are traversed , Start the corresponding operation steps , After the operation is completed, the result is passed to the downstream Slink

• If the downstream Slink Time short circuit operation , Ask constantly when passing the results to the downstream cancellationrequested() Whether the processing can be ended

How to execute the operation after stacking

 Stream The designer of the class library set up a

Sink AbstractPipeline.opWrapSink(int flags, Sink downstream)

Method to get Sink, The function of this method is to return a new containing the current Stage Represents the operation and the ability to pass results to downstream Of Sink object .

  Use opWrapSink() You can connect the current operation with the downstream Sink（ Above downstream Parameters ） Combination and innovation Sink. Just imagine from the last one in the assembly line Stage Start , Keep calling the last one Stage Of opWrapSink() Until the beginning （ barring stage0, because stage0 For data sources , Operation not included ）, You can get a representation of all operations on the pipeline Sink. All operations from the beginning to the end of the assembly line are packed into one Sink in , Execute this Sink It's like executing the entire pipeline .