NonBlockingFileIO

public struct NonBlockingFileIO

NonBlockingFileIO is a helper that allows you to read files without blocking the calling thread.

It is worth noting that kqueue, epoll or poll returning claiming a file is readable does not mean that the data is already available in the kernel’s memory. In other words, a read from a file can still block even if reported as readable. This behaviour is also documented behaviour:

poll: Regular files shall always poll TRUE for reading and writing.
epoll: epoll is simply a faster poll(2), and can be used wherever the latter is used since it shares the same semantics.
kqueue: Returns when the file pointer is not at the end of file.

NonBlockingFileIO helps to work around this issue by maintaining its own thread pool that is used to read the data from the files into memory. It will then hand the (in-memory) data back which makes it available without the possibility of blocking.

defaultThreadPoolSize
The default and recommended size for NonBlockingFileIO‘s thread pool.
Declaration
Swift

public static let defaultThreadPoolSize: Int
defaultChunkSize
The default and recommended chunk size.
Declaration
Swift

public static let defaultChunkSize: Int
Error
NonBlockingFileIO errors.
See more
Declaration
Swift

public enum Error : Swift.Error


                    
                    
                    init(threadPool:)

Initialize a NonBlockingFileIO which uses the BlockingIOThreadPool.

Declaration

Swift

public init(threadPool: BlockingIOThreadPool)

Parameters


                                threadPool

The BlockingIOThreadPool that will be used for all the IO.


                    
                    
                    readChunked(fileRegion:chunkSize:allocator:eventLoop:chunkHandler:)

Read a FileRegion in chunks of chunkSize bytes on NonBlockingFileIO‘s private thread pool which is separate from any EventLoop thread.

chunkHandler will be called on eventLoop for every chunk that was read. Assuming fileRegion.readableBytes is greater than zero and there are enough bytes available chunkHandler will be called 1 + |_ fileRegion.readableBytes / chunkSize _| times, delivering chunkSize bytes each time. If less than fileRegion.readableBytes bytes can be read from the file, chunkHandler will be called less often with the last invocation possibly being of less than chunkSize bytes.

The allocation and reading of a subsequent chunk will only be attempted when chunkHandler succeeds.

Declaration

Swift

public func readChunked(fileRegion: FileRegion,
                        chunkSize: Int = NonBlockingFileIO.defaultChunkSize,
                        allocator: ByteBufferAllocator,
                        eventLoop: EventLoop,
                        chunkHandler: @escaping (ByteBuffer) -> EventLoopFuture<Void>) -> EventLoopFuture<Void>

Parameters

`fileRegion`	The file region to read.
`chunkSize`	The size of the individual chunks to deliver.
`allocator`	A `ByteBufferAllocator` used to allocate space for the chunks.
`eventLoop`	The `EventLoop` to call `chunkHandler` on.
`chunkHandler`	Called for every chunk read. The next chunk will be read upon successful completion of the returned `EventLoopFuture`. If the returned `EventLoopFuture` fails, the overall operation is aborted.

Return Value

An EventLoopFuture which is the result of the overall operation. If either the reading of fileHandle or chunkHandler fails, the EventLoopFuture will fail too. If the reading of fileHandle as well as chunkHandler always succeeded, the EventLoopFuture will succeed too.


                    
                    
                    readChunked(fileHandle:byteCount:chunkSize:allocator:eventLoop:chunkHandler:)

Read byteCount bytes in chunks of chunkSize bytes from fileHandle in NonBlockingFileIO‘s private thread pool which is separate from any EventLoop thread.

chunkHandler will be called on eventLoop for every chunk that was read. Assuming byteCount is greater than zero and there are enough bytes available chunkHandler will be called 1 + |_ byteCount / chunkSize _| times, delivering chunkSize bytes each time. If less than byteCount bytes can be read from descriptor, chunkHandler will be called less often with the last invocation possibly being of less than chunkSize bytes.

The allocation and reading of a subsequent chunk will only be attempted when chunkHandler succeeds.

Note

readChunked(fileRegion:chunkSize:allocator:eventLoop:chunkHandler:) should be preferred as it uses FileRegion object instead of raw FileHandles.

Declaration

Swift

public func readChunked(fileHandle: FileHandle,
                        byteCount: Int,
                        chunkSize: Int = NonBlockingFileIO.defaultChunkSize,
                        allocator: ByteBufferAllocator,
                        eventLoop: EventLoop, chunkHandler: @escaping (ByteBuffer) -> EventLoopFuture<Void>) -> EventLoopFuture<Void>

Parameters

`fileHandle`	The `FileHandle` to read from.
`byteCount`	The number of bytes to read from `fileHandle`.
`chunkSize`	The size of the individual chunks to deliver.
`allocator`	A `ByteBufferAllocator` used to allocate space for the chunks.
`eventLoop`	The `EventLoop` to call `chunkHandler` on.
`chunkHandler`	Called for every chunk read. The next chunk will be read upon successful completion of the returned `EventLoopFuture`. If the returned `EventLoopFuture` fails, the overall operation is aborted.

Return Value


                    
                    
                    read(fileRegion:allocator:eventLoop:)

Read a FileRegion in NonBlockingFileIO‘s private thread pool which is separate from any EventLoop thread.

The returned ByteBuffer will not have less than fileRegion.readableBytes unless we hit end-of-file in which case the ByteBuffer will contain the bytes available to read.

Note

Only use this function for small enough FileRegions as it will need to allocate enough memory to hold fileRegion.readableBytes bytes.

Note

In most cases you should prefer one of the readChunked functions.

Declaration

Swift

public func read(fileRegion: FileRegion, allocator: ByteBufferAllocator, eventLoop: EventLoop) -> EventLoopFuture<ByteBuffer>

Parameters

`fileRegion`	The file region to read.
`allocator`	A `ByteBufferAllocator` used to allocate space for the returned `ByteBuffer`.
`eventLoop`	The `EventLoop` to create the returned `EventLoopFuture` from.

Return Value

An EventLoopFuture which delivers a ByteBuffer if the read was successful or a failure on error.


                    
                    
                    read(fileHandle:byteCount:allocator:eventLoop:)

Read byteCount bytes from fileHandle in NonBlockingFileIO‘s private thread pool which is separate from any EventLoop thread.

The returned ByteBuffer will not have less than byteCount bytes unless we hit end-of-file in which case the ByteBuffer will contain the bytes available to read.

Note

Only use this function for small enough byteCounts as it will need to allocate enough memory to hold byteCount bytes.

Note

read(fileRegion:allocator:eventLoop:) should be preferred as it uses FileRegion object instead of raw FileHandles.

Declaration

Swift

public func read(fileHandle: FileHandle, byteCount: Int, allocator: ByteBufferAllocator, eventLoop: EventLoop) -> EventLoopFuture<ByteBuffer>

Parameters

`fileHandle`	The `FileHandle` to read.
`byteCount`	The number of bytes to read from `fileHandle`.
`allocator`	A `ByteBufferAllocator` used to allocate space for the returned `ByteBuffer`.
`eventLoop`	The `EventLoop` to create the returned `EventLoopFuture` from.

Return Value

An EventLoopFuture which delivers a ByteBuffer if the read was successful or a failure on error.


                    
                    
                    write(fileHandle:buffer:eventLoop:)

Write buffer to fileHandle in NonBlockingFileIO‘s private thread pool which is separate from any EventLoop thread.

Declaration

Swift

public func write(fileHandle: FileHandle,
                  buffer: ByteBuffer,
                  eventLoop: EventLoop) -> EventLoopFuture<()>

Parameters

`fileHandle`	The `FileHandle` to write to.
`buffer`	The `ByteBuffer` to write.
`eventLoop`	The `EventLoop` to create the returned `EventLoopFuture` from.

Return Value

An EventLoopFuture which is fulfilled if the write was successful or fails on error.


                    
                    
                    openFile(path:eventLoop:)

Open the file at path on a private thread pool which is separate from any EventLoop thread.

This function will return (a future) of the FileHandle associated with the file opened and a FileRegion comprising of the whole file. The caller must close the returned FileHandle when it’s no longer needed.

Note

The reason this returns the FileHandle and the FileRegion is that both the opening of a file as well as the querying of its size are blocking.

Declaration

Swift

public func openFile(path: String, eventLoop: EventLoop) -> EventLoopFuture<(FileHandle, FileRegion)>

Parameters

`path`	The path of the file to be opened.
`eventLoop`	The `EventLoop` on which the returned `EventLoopFuture` will fire.

Return Value

An EventLoopFuture containing the FileHandle and the FileRegion comprising the whole file.