程序包 cn.taketoday.core.io.buffer

类 DataBufferUtils

java.lang.Object
cn.taketoday.core.io.buffer.DataBufferUtils
public abstract class DataBufferUtils extends Object
Utility class for working with DataBuffers.
从以下版本开始:
4.0
作者:
Arjen Poutsma, Brian Clozel, Harry Yang

  • 字段详细资料

    • logger

      private static final Logger logger

    • RELEASE_CONSUMER

      private static final Consumer<DataBuffer> RELEASE_CONSUMER

    • DEFAULT_CHUNK_SIZE

      private static final int DEFAULT_CHUNK_SIZE
      另请参阅:

  • 构造器详细资料

    • DataBufferUtils

      public DataBufferUtils()

  • 方法详细资料

    • readInputStream

      public static reactor.core.publisher.Flux<DataBuffer> readInputStream(Callable<InputStream> inputStreamSupplier, DataBufferFactory bufferFactory, int bufferSize)
      Obtain an InputStream from the given supplier, and read it into a Flux of DataBuffers. Closes the input stream when the Flux is terminated.
      参数:
      inputStreamSupplier - the supplier for the input stream to read from
      bufferFactory - the factory to create data buffers with
      bufferSize - the maximum size of the data buffers
      返回:
      a Flux of data buffers read from the given channel

    • readByteChannel

      public static reactor.core.publisher.Flux<DataBuffer> readByteChannel(Callable<ReadableByteChannel> channelSupplier, DataBufferFactory bufferFactory, int bufferSize)
      Obtain a ReadableByteChannel from the given supplier, and read it into a Flux of DataBuffers. Closes the channel when the Flux is terminated.
      参数:
      channelSupplier - the supplier for the channel to read from
      bufferFactory - the factory to create data buffers with
      bufferSize - the maximum size of the data buffers
      返回:
      a Flux of data buffers read from the given channel

    • readAsynchronousFileChannel

      public static reactor.core.publisher.Flux<DataBuffer> readAsynchronousFileChannel(Callable<AsynchronousFileChannel> channelSupplier, DataBufferFactory bufferFactory, int bufferSize)
      Obtain a AsynchronousFileChannel from the given supplier, and read it into a Flux of DataBuffers. Closes the channel when the Flux is terminated.
      参数:
      channelSupplier - the supplier for the channel to read from
      bufferFactory - the factory to create data buffers with
      bufferSize - the maximum size of the data buffers
      返回:
      a Flux of data buffers read from the given channel

    • readAsynchronousFileChannel

      public static reactor.core.publisher.Flux<DataBuffer> readAsynchronousFileChannel(Callable<AsynchronousFileChannel> channelSupplier, long position, DataBufferFactory bufferFactory, int bufferSize)
      Obtain an AsynchronousFileChannel from the given supplier, and read it into a Flux of DataBuffers, starting at the given position. Closes the channel when the Flux is terminated.
      参数:
      channelSupplier - the supplier for the channel to read from
      position - the position to start reading from
      bufferFactory - the factory to create data buffers with
      bufferSize - the maximum size of the data buffers
      返回:
      a Flux of data buffers read from the given channel

    • read

      public static reactor.core.publisher.Flux<DataBuffer> read(Path path, DataBufferFactory bufferFactory, int bufferSize, OpenOption... options)
      Read bytes from the given file Path into a Flux of DataBuffers. The method ensures that the file is closed when the flux is terminated.
      参数:
      path - the path to read bytes from
      bufferFactory - the factory to create data buffers with
      bufferSize - the maximum size of the data buffers
      返回:
      a Flux of data buffers read from the given channel

    • read

      public static reactor.core.publisher.Flux<DataBuffer> read(Resource resource, DataBufferFactory bufferFactory, int bufferSize)
      Read the given Resource into a Flux of DataBuffers.

      If the resource is a file, it is read into an AsynchronousFileChannel and turned to Flux via readAsynchronousFileChannel(Callable, DataBufferFactory, int) or else fall back to readByteChannel(Callable, DataBufferFactory, int). Closes the channel when the flux is terminated.

      参数:
      resource - the resource to read from
      bufferFactory - the factory to create data buffers with
      bufferSize - the maximum size of the data buffers
      返回:
      a Flux of data buffers read from the given channel

    • read

      public static reactor.core.publisher.Flux<DataBuffer> read(Resource resource, long position, DataBufferFactory bufferFactory, int bufferSize)
      Read the given Resource into a Flux of DataBuffers starting at the given position.

      If the resource is a file, it is read into an AsynchronousFileChannel and turned to Flux via readAsynchronousFileChannel(Callable, DataBufferFactory, int) or else fall back on readByteChannel(Callable, DataBufferFactory, int). Closes the channel when the flux is terminated.

      参数:
      resource - the resource to read from
      position - the position to start reading from
      bufferFactory - the factory to create data buffers with
      bufferSize - the maximum size of the data buffers
      返回:
      a Flux of data buffers read from the given channel

    • write

      public static reactor.core.publisher.Flux<DataBuffer> write(org.reactivestreams.Publisher<DataBuffer> source, OutputStream outputStream)
      Write the given stream of DataBuffers to the given OutputStream. Does not close the output stream when the flux is terminated, and does not release the data buffers in the source. If releasing is required, then subscribe to the returned Flux with a releaseConsumer().

      Note that the writing process does not start until the returned Flux is subscribed to.

      参数:
      source - the stream of data buffers to be written
      outputStream - the output stream to write to
      返回:
      a Flux containing the same buffers as in source, that starts the writing process when subscribed to, and that publishes any writing errors and the completion signal

    • write

      public static reactor.core.publisher.Flux<DataBuffer> write(org.reactivestreams.Publisher<DataBuffer> source, WritableByteChannel channel)
      Write the given stream of DataBuffers to the given WritableByteChannel. Does not close the channel when the flux is terminated, and does not release the data buffers in the source. If releasing is required, then subscribe to the returned Flux with a releaseConsumer().

      Note that the writing process does not start until the returned Flux is subscribed to.

      参数:
      source - the stream of data buffers to be written
      channel - the channel to write to
      返回:
      a Flux containing the same buffers as in source, that starts the writing process when subscribed to, and that publishes any writing errors and the completion signal

    • write

      public static reactor.core.publisher.Flux<DataBuffer> write(org.reactivestreams.Publisher<DataBuffer> source, AsynchronousFileChannel channel)
      Write the given stream of DataBuffers to the given AsynchronousFileChannel. Does not close the channel when the flux is terminated, and does not release the data buffers in the source. If releasing is required, then subscribe to the returned Flux with a releaseConsumer().

      Note that the writing process does not start until the returned Flux is subscribed to.

      参数:
      source - the stream of data buffers to be written
      channel - the channel to write to
      返回:
      a Flux containing the same buffers as in source, that starts the writing process when subscribed to, and that publishes any writing errors and the completion signal

    • write

      public static reactor.core.publisher.Flux<DataBuffer> write(org.reactivestreams.Publisher<? extends DataBuffer> source, AsynchronousFileChannel channel, long position)
      Write the given stream of DataBuffers to the given AsynchronousFileChannel. Does not close the channel when the flux is terminated, and does not release the data buffers in the source. If releasing is required, then subscribe to the returned Flux with a releaseConsumer().

      Note that the writing process does not start until the returned Flux is subscribed to.

      参数:
      source - the stream of data buffers to be written
      channel - the channel to write to
      position - the file position where writing is to begin; must be non-negative
      返回:
      a flux containing the same buffers as in source, that starts the writing process when subscribed to, and that publishes any writing errors and the completion signal

    • write

      public static reactor.core.publisher.Mono<Void> write(org.reactivestreams.Publisher<DataBuffer> source, Path destination, OpenOption... options)
      Write the given stream of DataBuffers to the given file Path. The optional options parameter specifies how the file is created or opened (defaults to CREATE, TRUNCATE_EXISTING, and WRITE).
      参数:
      source - the stream of data buffers to be written
      destination - the path to the file
      options - the options specifying how the file is opened
      返回:
      a Mono that indicates completion or error

    • checkWriteOptions

      private static Set<OpenOption> checkWriteOptions(OpenOption[] options)

    • closeChannel

      static void closeChannel(@Nullable Channel channel)

    • outputStreamPublisher

      public static org.reactivestreams.Publisher<DataBuffer> outputStreamPublisher(Consumer<OutputStream> outputStreamConsumer, DataBufferFactory bufferFactory, Executor executor)
      Create a new Publisher<DataBuffer> based on bytes written to a OutputStream.
      • The parameter outputStreamConsumer is invoked once per subscription of the returned Publisher, when the first item is requested.
      • OutputStream.write() invocations made by outputStreamConsumer are buffered until they exceed the default chunk size of 1024, or when the stream is flushed and then result in a published item if there is demand.
      • If there is no demand, OutputStream.write() will block until there is.
      • If the subscription is cancelled, OutputStream.write() will throw a IOException.
      • The subscription is completed when outputStreamHandler completes.
      • Any exceptions thrown from outputStreamHandler will be dispatched to the Subscriber.
      参数:
      outputStreamConsumer - invoked when the first buffer is requested
      executor - used to invoke the outputStreamHandler
      返回:
      a Publisher<DataBuffer> based on bytes written by outputStreamHandler

    • outputStreamPublisher

      public static org.reactivestreams.Publisher<DataBuffer> outputStreamPublisher(Consumer<OutputStream> outputStreamConsumer, DataBufferFactory bufferFactory, Executor executor, int chunkSize)
      Creates a new Publisher<DataBuffer> based on bytes written to a OutputStream.
      • The parameter outputStreamConsumer is invoked once per subscription of the returned Publisher, when the first item is requested.
      • OutputStream.write() invocations made by outputStreamHandler are buffered until they reach or exceed chunkSize, or when the stream is flushed and then result in a published item if there is demand.
      • If there is no demand, OutputStream.write() will block until there is.
      • If the subscription is cancelled, OutputStream.write() will throw a IOException.
      • The subscription is completed when outputStreamHandler completes.
      • Any exceptions thrown from outputStreamHandler will be dispatched to the Subscriber.
      参数:
      outputStreamConsumer - invoked when the first buffer is requested
      executor - used to invoke the outputStreamHandler
      chunkSize - minimum size of the buffer produced by the publisher
      返回:
      a Publisher<DataBuffer> based on bytes written by outputStreamHandler

    • takeUntilByteCount

      public static <T extends DataBuffer> reactor.core.publisher.Flux<T> takeUntilByteCount(org.reactivestreams.Publisher<T> publisher, long maxByteCount)
      Relay buffers from the given Publisher until the total byte count reaches the given maximum byte count, or until the publisher is complete.
      参数:
      publisher - the publisher to filter
      maxByteCount - the maximum byte count
      返回:
      a flux whose maximum byte count is maxByteCount

    • skipUntilByteCount

      public static <T extends DataBuffer> reactor.core.publisher.Flux<T> skipUntilByteCount(org.reactivestreams.Publisher<T> publisher, long maxByteCount)
      Skip buffers from the given Publisher until the total byte count reaches the given maximum byte count, or until the publisher is complete.
      参数:
      publisher - the publisher to filter
      maxByteCount - the maximum byte count
      返回:
      a flux with the remaining part of the given publisher

    • retain

      public static <T extends DataBuffer> T retain(T dataBuffer)
      Retain the given data buffer, if it is a PooledDataBuffer.
      参数:
      dataBuffer - the data buffer to retain
      返回:
      the retained buffer

    • touch

      public static <T extends DataBuffer> T touch(T dataBuffer, Object hint)
      Associate the given hint with the data buffer if it is a pooled buffer and supports leak tracking.
      参数:
      dataBuffer - the data buffer to attach the hint to
      hint - the hint to attach
      返回:
      the input buffer

    • release

      public static boolean release(@Nullable DataBuffer dataBuffer)
      Release the given data buffer. If it is a PooledDataBuffer and has been allocated, this method will call PooledDataBuffer.release(). If it is a CloseableDataBuffer, this method will call CloseableDataBuffer.close().
      参数:
      dataBuffer - the data buffer to release
      返回:
      true if the buffer was released; false otherwise.

    • releaseConsumer

      public static Consumer<DataBuffer> releaseConsumer()
      Return a consumer that calls release(DataBuffer) on all passed data buffers.

    • join

      public static reactor.core.publisher.Mono<DataBuffer> join(org.reactivestreams.Publisher<? extends DataBuffer> dataBuffers)
      Return a new DataBuffer composed of joining together the given dataBuffers elements. Depending on the DataBuffer type, the returned buffer may be a single buffer containing all data of the provided buffers, or it may be a zero-copy, composite with references to the given buffers.

      If dataBuffers produces an error or if there is a cancel signal, then all accumulated buffers will be released.

      Note that the given data buffers do not have to be released. They will be released as part of the returned composite.

      参数:
      dataBuffers - the data buffers that are to be composed
      返回:
      a buffer that is composed of the dataBuffers argument

    • join

      public static reactor.core.publisher.Mono<DataBuffer> join(org.reactivestreams.Publisher<? extends DataBuffer> buffers, int maxByteCount)
      Variant of join(Publisher) that behaves the same way up until the specified max number of bytes to buffer. Once the limit is exceeded, DataBufferLimitException is raised.
      参数:
      buffers - the data buffers that are to be composed
      maxByteCount - the max number of bytes to buffer, or -1 for unlimited
      返回:
      a buffer with the aggregated content, possibly an empty Mono if the max number of bytes to buffer is exceeded.
      抛出:
      DataBufferLimitException - if maxByteCount is exceeded

    • matcher

      public static DataBufferUtils.Matcher matcher(byte[] delimiter)
      Return a DataBufferUtils.Matcher for the given delimiter. The matcher can be used to find the delimiters in a stream of data buffers.
      参数:
      delimiter - the delimiter bytes to find
      返回:
      the matcher

    • matcher

      public static DataBufferUtils.Matcher matcher(byte[]... delimiters)
      Return a DataBufferUtils.Matcher for the given delimiters. The matcher can be used to find the delimiters in a stream of data buffers.
      参数:
      delimiters - the delimiters bytes to find
      返回:
      the matcher

    • createMatcher

      private static DataBufferUtils.NestedMatcher createMatcher(byte[] delimiter)