Class EncapsulatedInputStream

java.lang.Object
java.io.InputStream
com.pixelmed.dicom.EncapsulatedInputStream
All Implemented Interfaces:
Closeable, AutoCloseable

public class EncapsulatedInputStream extends InputStream

A class that extends InputStream by adding a mechanism for unencapsulating an undefined length DICOM attribute, such as is used for compressed Pixel Data.

The read methods hide the fact that the data is encapsulated by removing the Items and Item and Sequence delimiter tags, as well as skipping any Basic Offset Table that may be present in the first Item.

Since an individual frame may be fragmented and padded beyond the JPEG EOI marker (0xffd9), and since the codec used for decoding may be "reading ahead" this class also removes any padding bytes at the end of any fragment, back as far as the EOI marker. Note that this means that theoretically frames could span fragments as long as there was no padding between them.

  • Constructor Details

    • EncapsulatedInputStream

      public EncapsulatedInputStream(BinaryInputStream i, boolean jpegEOIDetection, boolean oneFragmentPerFrame)

      Construct a byte ordered stream from the supplied stream.

      The byte order may be changed later.

      Parameters:
      i - the input stream to read from
      jpegEOIDetection - whether or not to detect JPEG EOI, to allow frames to span fragments
      oneFragmentPerFrame - if not using jpegEOIDetection, whether or not to assume one fragment per frame (e.g., for RLE), rather than all fragments in one frame (e.g., for MPEG)
    • EncapsulatedInputStream

      public EncapsulatedInputStream(BinaryInputStream i)

      Construct a byte ordered stream from the supplied stream.

      The byte order may be changed later.

      JPEG EOI detection is enabled, to allow frames to span fragments.

      Do not use for RLE encapsulated data, since it will assume all fragments are in one frame, rather than one fragment per frame.

      Parameters:
      i - the input stream to read from
  • Method Details

    • getBytesRead

      protected long getBytesRead()
    • getByteOffsetsOfItemsForCurrentFrame

      public long[] getByteOffsetsOfItemsForCurrentFrame(long startOfPixelData)
    • getItemLengthsForCurrentFrame

      public long[] getItemLengthsForCurrentFrame()
    • readSequenceDelimiter

      public void readSequenceDelimiter() throws IOException
      Throws:
      IOException - if an I/O error occurs
    • nextFrame

      public void nextFrame()

      Skip to the start of the first fragment for a frame, if not already there.

    • readUnsigned16

      public final void readUnsigned16(short[] w, int offset, int len) throws IOException

      Read an array of unsigned integer 16 bit values.

      Parameters:
      w - an array of sufficient size in which to return the values read
      offset - the offset in the array at which to begin storing values
      len - the number of 16 bit values to read
      Throws:
      IOException - if an I/O error occurs
    • readUnsigned32LittleEndian

      public final long readUnsigned32LittleEndian() throws IOException

      Read an unsigned integer 32 bit values little endian regardless of endianness of stream.

      Returns:
      the value read
      Throws:
      IOException - if an I/O error occurs
    • readUnsigned32LittleEndian

      public final void readUnsigned32LittleEndian(long[] w, int offset, int len) throws IOException

      Read an array of unsigned integer 32 bit values little endian regardless of endianness of stream.

      Parameters:
      w - an array of sufficient size in which to return the values read
      offset - the offset in the array at which to begin storing values
      len - the number of 32 bit values to read
      Throws:
      IOException - if an I/O error occurs
    • skipInsistently

      public void skipInsistently(long length) throws IOException

      Skip as many bytes as requested, unless an exception occurs.

      Parameters:
      length - number of bytes to read (no more and no less)
      Throws:
      IOException - if an I/O error occurs
    • read

      public final int read() throws IOException

      Extracts the next byte of data from the current or subsequent fragments.

      Specified by:
      read in class InputStream
      Returns:
      the next byte of data, or -1 if there is no more data because the end of the stream has been reached.
      Throws:
      IOException - if an I/O error occurs.
    • read

      public final int read(byte[] b) throws IOException

      Extracts byte.length bytes of data from the current or subsequent fragments.

      This method simply performs the call read(b, 0, b.length) and returns the result.

      Overrides:
      read in class InputStream
      Parameters:
      b - the buffer into which the data is read.
      Returns:
      the total number of bytes read into the buffer (always whatever was asked for), or -1 if there is no more data because the end of the stream has been reached.
      Throws:
      IOException - if an I/O error occurs.
      See Also:
    • read

      public final int read(byte[] b, int off, int len) throws IOException

      Extracts len bytes of data from the current or subsequent fragments.

      Overrides:
      read in class InputStream
      Parameters:
      b - the buffer into which the data is read.
      off - the start offset of the data.
      len - the number of bytes read.
      Returns:
      the total number of bytes read into the buffer (always whatever was asked for), or -1 if there is no more data because the end of a frame has been reached.
      Throws:
      IOException - if an I/O error occurs.
    • getOffsetOfNextByteToReadFromStartOfFragment

      public int getOffsetOfNextByteToReadFromStartOfFragment()
    • framesAreSeparated

      public boolean framesAreSeparated()

      Are frames separated (seperable)?

      This will be either because the stream contains recognizable end of frame markers (e.g., JPEG) or the Transfer Syntax specifies one fragment per frame (e.g., RLE).

      Returns:
      true if frames are separated rather than being lumped together in one byte stream/array