public abstract class BaseBuffer extends Object implements PyBuffer
This class leaves undefined the storage mechanism for the bytes (typically byte[] or
java.nio.ByteBuffer), while remaining definite that it is an indexable sequence of
bytes. A concrete class that extends this one must provide elementary accessors
byteAtImpl(int), storeAtImpl(byte, int) that abstract this storage, a factory
getNIOByteBufferImpl() for ByteBuffers that wrap the storage, and a factory
for slices PyBuffer.getBufferSlice(int, int, int, int).
The sub-class constructor must specify the feature flags (see #BaseBuffer(int)), set
index0, shape and strides, and finally check the client capabilities
with checkRequestFlags(int). Sub-classes intended to represent slices of exporters that
must count their exports as part of a locking protocol, as does bytearray, must
override getRoot() so that a buffer view release() on a slice, propagates to
the buffer view that provided it.
Access methods provided here necessarily work with the abstracted byteAtImpl(int),
storeAtImpl(byte, int) interface, but subclasses are able to override them with more
efficient versions that employ knowledge of the particular storage type used.
This base implementation is writable only if PyBUF.WRITABLE is in the feature flags
passed to the constructor. Otherwise, all methods for write access raise a TypeError
and isReadonly() returns true. However, a client intending to write should
have presented PyBUF.WRITABLE in its client request flags when getting the buffer, and
been prevented by a BufferError exception at that point.
At the time of writing, only one-dimensional buffers of item size one are used in the Jython core.
PyBuffer.PointerANY_CONTIGUOUS, AS_ARRAY, C_CONTIGUOUS, CONTIG, CONTIG_RO, CONTIGUITY, F_CONTIGUOUS, FORMAT, FULL, FULL_RO, INDIRECT, IS_C_CONTIGUOUS, IS_F_CONTIGUOUS, MAX_NDIM, NAVIGATION, ND, RECORDS, RECORDS_RO, SIMPLE, STRIDED, STRIDED_RO, STRIDES, WRITABLE| Modifier and Type | Method and Description |
|---|---|
byte |
byteAt(int... indices)
Return the byte indexed from an N-dimensional buffer with item size one.
|
byte |
byteAt(int index)
Return the byte indexed from a one-dimensional buffer with item size one.
|
int |
byteIndex(int... indices)
Convert a multi-dimensional item index to an absolute byte index in the storage shared by the
exporter.
|
int |
byteIndex(int index)
Convert an item index (for a one-dimensional buffer) to an absolute byte index in the storage
shared by the exporter.
|
void |
close()
An alias for
PyBuffer.release() to satisfy AutoCloseable. |
void |
copyFrom(byte[] src,
int srcPos,
int destIndex,
int count)
Copy from a slice of a (Java) byte array into the buffer starting at a given destination
item-index.
|
void |
copyFrom(PyBuffer src)
Copy the whole of another
PyBuffer into this buffer. |
void |
copyTo(byte[] dest,
int destPos)
Copy the contents of the buffer to the destination byte array.
|
void |
copyTo(int srcIndex,
byte[] dest,
int destPos,
int count)
Copy a simple slice of the buffer-view to the destination byte array, defined by a starting
item-index in the source buffer and the
count of items to copy. |
PyBuffer.Pointer |
getBuf()
Return a structure describing the slice of a byte array that holds the data being exported to
the consumer.
|
PyBuffer |
getBuffer(int flags)
Method by which the consumer requests the buffer from the exporter.
|
BaseBuffer |
getBufferAgain(int flags)
Allow an exporter to re-use this object again even if it has been "finally" released.
|
PyBuffer |
getBufferSlice(int flags,
int start,
int count)
Equivalent to
PyBuffer.getBufferSlice(int, int, int, int) with stride 1. |
String |
getFormat()
A format string in the language of Python structs describing how the bytes of each item
should be interpreted.
|
int |
getItemsize()
The number of bytes stored in each indexable item.
|
int |
getLen()
The total number of bytes represented by the view, which will be the product of the elements of the
shape array, and the item size in bytes. |
int |
getNdim()
The number of dimensions to the buffer.
|
ByteBuffer |
getNIOByteBuffer()
Obtain a
ByteBuffer giving access to the bytes that hold the data being
exported by the original object. |
BufferProtocol |
getObj()
Return the underlying exporting object (or
null if no object implementing the
BufferProtocol is in that role). |
PyBuffer.Pointer |
getPointer(int... indices)
Return a structure describing the position in a byte array of a single item from the data
being exported to the consumer, in the case that array may be multi-dimensional.
|
PyBuffer.Pointer |
getPointer(int index)
Return a structure describing the position in a byte array of a single item from the data
being exported to the consumer.
|
int[] |
getShape()
An array reporting the size of the buffer, considered as a multidimensional array, in each
dimension and (by its length) giving the number of dimensions.
|
int[] |
getStrides()
The
strides array gives the distance in the storage array between adjacent items
(in each dimension). |
int[] |
getSuboffsets()
The
suboffsets array is a further part of the support for interpreting the
buffer as an n-dimensional array of items, where the array potentially uses indirect
addressing (like a real Java array of arrays, in fact). |
boolean |
hasArray()
Report whether the exporter is able to offer direct access to the exported storage as a Java
byte array (through the API that involves class
PyBuffer.Pointer), or only supports the
abstract API. |
int |
intAt(int... indices)
Return the unsigned byte value indexed from an N-dimensional buffer with item size one.
|
int |
intAt(int index)
Return the unsigned byte value indexed from a one-dimensional buffer with item size one.
|
boolean |
isContiguous(char order)
Enquire whether the array is represented contiguously in the backing storage, according to C
or Fortran ordering.
|
boolean |
isReadonly()
Determine whether the consumer is entitled to write to the exported storage.
|
boolean |
isReleased()
True only if the buffer has been released with (the required number of calls to)
PyBuffer.release() or some equivalent operation. |
void |
release()
A buffer is (usually) a view onto to the internal state of an exporting object, and that
object may have to restrict its behaviour while the buffer exists.
|
void |
storeAt(byte value,
int... indices)
Store the given byte at the indexed location in of an N-dimensional buffer with item size
one.
|
void |
storeAt(byte value,
int index)
Store the given byte at the indexed location in of a one-dimensional buffer with item size
one.
|
String |
toString()
The toString() method of a buffer reproduces the values in the buffer (as unsigned integers)
as the character codes of a
String. |
equals, getClass, hashCode, notify, notifyAll, wait, wait, waitgetBufferSlicepublic boolean isReadonly()
PyBUFisReadonly in interface PyBUFpublic int getNdim()
PyBUFshape
array. The actual storage may be a linear array, but this is the number of dimensions in the
interpretation that the exporting object gives the data.public int[] getShape()
PyBUFshape array is
always returned (difference from CPython).public int getLen()
PyBUFshape array, and the item size in bytes.public final BufferProtocol getObj()
PyBuffernull if no object implementing the
BufferProtocol is in that role). This will often be a PyObject.public byte byteAt(int index)
throws IndexOutOfBoundsException
itemsize>1.
The BaseBuffer implementation delegates to byteAtImpl(int) via
byteAtImpl(byteIndex(index)).
byteAt in interface PyBufferindex - to retrieve fromIndexOutOfBoundsExceptionpublic int intAt(int index)
throws IndexOutOfBoundsException
itemsize>1.
The BaseBuffer implementation delegates to byteAtImpl(int) via
byteAtImpl(byteIndex(index)), cast unsigned to an int.
intAt in interface PyBufferindex - to retrieve from=0xff & byteAt(index)IndexOutOfBoundsExceptionpublic void storeAt(byte value,
int index)
throws IndexOutOfBoundsException,
PyException
itemsize>1.
The BaseBuffer implementation delegates to storeAtImpl(byte, int) via
storeAtImpl(value, byteIndex(index)).
storeAt in interface PyBuffervalue - to storeindex - to locationIndexOutOfBoundsExceptionPyExceptionpublic byte byteAt(int... indices)
throws IndexOutOfBoundsException
itemsize>1.
The BaseBuffer implementation delegates to byteAtImpl(int) via
byteAtImpl(byteIndex(indices)).
byteAt in interface PyBufferindices - specifying location to retrieve fromIndexOutOfBoundsExceptionpublic int intAt(int... indices)
throws IndexOutOfBoundsException
itemsize>1.
The BaseBuffer implementation delegates to byteAtImpl(int) via
byteAtImpl(byteIndex(indices)), cast unsigned to an int.
intAt in interface PyBufferindices - specifying location to retrieve from=0xff & byteAt(index)IndexOutOfBoundsExceptionpublic void storeAt(byte value,
int... indices)
throws IndexOutOfBoundsException,
PyException
itemsize>1.
The BaseBuffer implementation delegates to storeAtImpl(byte, int) via
storeAtImpl(value, byteIndex(indices)).
storeAt in interface PyBuffervalue - to storeindices - specifying location to store atIndexOutOfBoundsExceptionPyExceptionpublic int byteIndex(int index)
throws IndexOutOfBoundsException
PyBufferPyBuffer is a linearly-indexed
sequence of bytes, although it may not actually be a heap-allocated Java byte[]
object. The purpose of this method is to allow the exporter to define the relationship
between the item index (as used in PyBuffer.byteAt(int)) and the byte-index (as used with the
ByteBuffer returned by PyBuffer.getNIOByteBuffer()). See
PyBuffer.byteIndex(int[]) for discussion of the multi-dimensional case.byteIndex in interface PyBufferindex - item-index from consumerIndexOutOfBoundsExceptionpublic int byteIndex(int... indices)
throws IndexOutOfBoundsException
PyBufferPyBuffer is a linearly-indexed sequence of
bytes, although it may not actually be a heap-allocated Java byte[] object. The
purpose of this method is to allow the exporter to define the relationship between the item
index (as used in PyBuffer.byteAt(int...) and the byte-index (as used with the
ByteBuffer returned by PyBuffer.getNIOByteBuffer()).byteIndex in interface PyBufferindices - n-dimensional item-index from consumerIndexOutOfBoundsExceptionpublic void copyTo(byte[] dest,
int destPos)
throws IndexOutOfBoundsException
PyBUF.getLen(), and the order is the storage order in the exporter.
(Note: Correct ordering for multidimensional arrays, including those with indirection needs
further study.)
The default implementation in BaseBuffer deals with the general one-dimensional
case of arbitrary item size and stride.
copyTo in interface PyBufferdest - destination byte arraydestPos - byte-index in the destination array of the byte [0]IndexOutOfBoundsException - if the destination cannot hold itpublic void copyTo(int srcIndex,
byte[] dest,
int destPos,
int count)
throws IndexOutOfBoundsException,
PyException
count of items to copy. This may validly
be done only for a one-dimensional buffer, as the meaning of the starting item-index is
otherwise not defined. count*itemsize bytes will be occupied in the destination.
The default implementation in BaseBuffer deals with the general one-dimensional
case of arbitrary item size and stride, but is unable to optimise access to sequential bytes.
copyTo in interface PyBuffersrcIndex - starting item-index in the source bufferdest - destination byte arraydestPos - byte-index in the destination array of the source item [0,...]count - number of items to copyIndexOutOfBoundsException - if access out of bounds in source or destinationPyExceptionpublic void copyFrom(byte[] src,
int srcPos,
int destIndex,
int count)
throws IndexOutOfBoundsException,
PyException
count*itemsize bytes will be read
from the source.
The default implementation in BaseBuffer deals with the general one-dimensional
case of arbitrary item size and stride, but is unable to optimise access to sequential bytes.
copyFrom in interface PyBuffersrc - source byte arraysrcPos - location in source of first byte to copydestIndex - starting item-index in the destination (i.e. this)count - number of items to copy inIndexOutOfBoundsException - if access out of bounds in source or destinationPyException - (TypeError) if read-only bufferpublic void copyFrom(PyBuffer src) throws IndexOutOfBoundsException, PyException
PyBuffer into this buffer. This may validly be done
only for buffers that are consistent in their dimensions. When it is necessary to copy
partial buffers, this may be achieved using a buffer slice on the source or destination.
The default implementation in BaseBuffer deals with the general one-dimensional
case of arbitrary item size and stride, but is unable to optimise access to sequential bytes.
copyFrom in interface PyBuffersrc - source bufferIndexOutOfBoundsException - if access out of bounds in source or destinationPyException - (TypeError) if read-only bufferpublic PyBuffer getBuffer(int flags)
PyBufferPyBuffer.release() on the buffer it
obtained, since some objects alter their behaviour while buffers are exported.
When a PyBuffer is the target, the same checks are carried out on the consumer
flags, and a return will normally be a reference to that buffer. A Jython
PyBuffer keeps count of these re-exports in order to match them with the number
of calls to PyBuffer.release(). When the last matching release() arrives it is
considered "final", and release actions may then take place on the exporting object. After
the final release of a buffer, a call to getBuffer should raise an exception.
getBuffer in interface BufferProtocolgetBuffer in interface PyBufferflags - specifying features demanded and the navigational capabilities of the consumerpublic BaseBuffer getBufferAgain(int flags)
BaseBytes can be re-used even after a final release by consumers,
simply by incrementing the exports count again: the navigation arrays and the
buffer view of the exporter's state all remain valid. We do not let consumers do this through
the PyBuffer interface: from their perspective, calling PyBuffer.release()
should mean the end of their access, although we can't stop them holding a reference to the
PyBuffer. Only the exporting object, which handles the implementation type is trusted to know
when re-use is safe.
An exporter will use this method as part of its implementation of
BufferProtocol.getBuffer(int). On return from that, the buffer and the exporting
object must then be in effectively the same state as if the buffer had just been
constructed by that method. Exporters that destroy related resources on final release of
their buffer (by overriding releaseAction()), or permit themselves structural change
invalidating the buffer, must either reconstruct the missing resources or avoid
getBufferAgain.
public void release()
BufferProtocol.getBuffer(int) or PyBuffer.getBuffer(int) should make a
matching call to PyBuffer.release(). The consumer may be sharing the PyBuffer
with other consumers and the buffer uses the pairing of getBuffer and
release to manage the lock on behalf of the exporter. It is an error to make
more than one call to release for a single call to getBuffer.
When the final matching release occurs (that is the number of release calls
equals the number of getBuffer calls), the implementation here calls
releaseAction(), which the implementer of a specific buffer type should override if
it needs specific actions to take place.
Note that, when this is a sliced view obtained from another PyBuffer the
implementation in BaseBuffer automatically sends one release()
Sub-classes should not propagate the release themselves when overriding
releaseAction().
public void close()
PyBufferPyBuffer.release() to satisfy AutoCloseable.close in interface AutoCloseableclose in interface PyBufferpublic boolean isReleased()
PyBufferPyBuffer.release() or some equivalent operation. The consumer may be sharing the reference
with other consumers and the buffer only achieves the released state when all consumers who
called getBuffer have called release.isReleased in interface PyBufferpublic PyBuffer getBufferSlice(int flags, int start, int count)
PyBufferPyBuffer.getBufferSlice(int, int, int, int) with stride 1.getBufferSlice in interface PyBufferflags - specifying features demanded and the navigational capabilities of the consumerstart - index in the current buffercount - number of items in the required slicepublic ByteBuffer getNIOByteBuffer()
PyBufferByteBuffer giving access to the bytes that hold the data being
exported by the original object. The position of the buffer is at the first byte of the item
with zero index (quite possibly not the lowest valid byte-index), the limit of the buffer is
beyond the largest valid byte index, and the mark is undefined.
For a one-dimensional contiguous buffer, the limit is one byte beyond the last item, so that
consecutive reads from the ByteBuffer return the data in order. Assuming the
following client code where obj has type BufferProtocol:
PyBuffer a = obj.getBuffer(PyBUF.SIMPLE); int itemsize = a.getItemsize(); ByteBuffer bb = a.getNIOBuffer();the item with index
k is in bb at positions
bb.pos()+k*itemsize to bb.pos()+(k+1)*itemsize - 1 inclusive. And
if itemsize==1, the item is simply the byte at position bb.pos()+k.
If the buffer is multidimensional or non-contiguous (strided), the buffer position is still
the (first byte of) the item at index [0] or [0,...,0]. However, it
is necessary to navigate bb using the shape, strides
and maybe suboffsets provided by the API.
getNIOByteBuffer in interface PyBufferByteBuffer onto the exported data contents.public boolean hasArray()
PyBufferPyBuffer.Pointer), or only supports the
abstract API. See also PyBUF.AS_ARRAY.public PyBuffer.Pointer getBuf()
PyBufferobj has type BufferProtocol:
PyBuffer a = obj.getBuffer(PyBUF.SIMPLE); int itemsize = a.getItemsize(); PyBuffer.Pointer b = a.getBuf();the item with index
k is in the array b.storage at index
[b.offset + k*itemsize] to [b.offset + (k+1)*itemsize - 1]
inclusive. And if itemsize==1, the item is simply the byte
b.storage[b.offset + k]
If the buffer is multidimensional or non-contiguous, storage[offset] is still
the (first byte of) the item at index [0] or [0,...,0]. However, it is necessary to navigate
b.storage using the shape, strides and maybe
suboffsets provided by the API.
public PyBuffer.Pointer getPointer(int index) throws IndexOutOfBoundsException
PyBufferobj has type BufferProtocol:
int k = ... ; PyBuffer a = obj.getBuffer(PyBUF.FULL); int itemsize = a.getItemsize(); PyBuffer.Pointer b = a.getPointer(k);the item with index
k is in the array b.storage at index
[b.offset] to [b.offset + itemsize - 1] inclusive. And if
itemsize==1, the item is simply the byte b.storage[b.offset]
Essentially this is a method for computing the offset of a particular index. The client is
free to navigate the underlying buffer b.storage without respecting these
boundaries.
getPointer in interface PyBufferindex - in the buffer to position the pointerIndexOutOfBoundsExceptionpublic PyBuffer.Pointer getPointer(int... indices) throws IndexOutOfBoundsException
PyBufferobj
has type BufferProtocol:
int i, j, k; // ... calculation that assigns i, j, k PyBuffer a = obj.getBuffer(PyBUF.FULL); int itemsize = a.getItemsize(); PyBuffer.Pointer b = a.getPointer(i,j,k);the item with index
[i,j,k] is in the array b.storage at index
[b.offset] to [b.offset + itemsize - 1] inclusive. And if
itemsize==1, the item is simply the byte b.storage[b.offset]
Essentially this is a method for computing the offset of a particular index. The client is
free to navigate the underlying buffer b.storage without respecting these
boundaries. If the buffer is non-contiguous, the above description is still valid (since a
multi-byte item must itself be contiguously stored), but in any additional navigation of
b.storage[] to other items, the client must use the shape, strides and
sub-offsets provided by the API. Normally one starts b = a.getBuf() in order to
establish the offset of index [0,...,0].
getPointer in interface PyBufferindices - multidimensional index at which to position the pointerIndexOutOfBoundsExceptionpublic int[] getStrides()
PyBUFstrides array gives the distance in the storage array between adjacent items
(in each dimension). In the rawest parts of the buffer API, the consumer of the buffer is
able to navigate the exported storage. The "strides" array is part of the support for
interpreting the buffer as an n-dimensional array of items. It provides the coefficients of
the "addressing polynomial". (More on this in the CPython documentation.) The consumer must
not modify this array. A valid strides array is always returned (difference from
CPython).getStrides in interface PyBUFpublic int[] getSuboffsets()
PyBUFsuboffsets array is a further part of the support for interpreting the
buffer as an n-dimensional array of items, where the array potentially uses indirect
addressing (like a real Java array of arrays, in fact). This is only applicable when there is
more than 1 dimension, and it works in conjunction with the strides array. (More
on this in the CPython documentation.) When used, suboffsets[k] is an integer
index, not a byte offset as in CPython. The consumer must not modify this array. When not
needed for navigation null is returned (as in CPython).getSuboffsets in interface PyBUFnull if not necessary for navigationpublic boolean isContiguous(char order)
PyBUFisContiguous in interface PyBUForder - 'C', 'F' or 'A', as the storage order is C, Fortran or either.public String getFormat()
PyBufferPyBUF.FORMAT bit in the consumer's call to
getBuffer, a valid format string is always returned (difference
from CPython).
Jython only implements "B" so far, and it is debatable whether anything fancier than "<n>B" can be supported in Java.
public int getItemsize()
PyBUFgetItemsize in interface PyBUF