Map a vertex buffer object to client memory. This allows you to modify its contents.
If planning to update VBO vertex data, make sure the VBO usage types are GL_DYNAMIC_* or GL_STREAM_* or else serious performance issues may arise.
Modifying buffer data must be done carefully, or else system stability may
be affected. Do not use the returned view ndarray outside of successive
unmapBuffer() calls. Do not use the mapped
buffer for rendering until after
unmapBuffer() is called.
vbo (VertexBufferInfo) – Vertex buffer to map to client memory.
start (int) – Initial index of the sub-range of the buffer to modify.
length (int or None) – Number of elements of the sub-array to map from offset. If None, all elements to from offset to the end of the array are mapped.
read (bool, optional) – Allow data to be read from the buffer (sets GL_MAP_READ_BIT). This is ignored if noSync is True.
write (bool, optional) – Allow data to be written to the buffer (sets GL_MAP_WRITE_BIT).
noSync (bool, optional) – If True, GL will not wait until the buffer is free (i.e. not being processed by the GPU) to map it (sets GL_MAP_UNSYNCHRONIZED_BIT). The contents of the previous storage buffer are discarded and the driver returns a new one. This prevents the CPU from stalling until the buffer is available.
View of the data. The type of the returned array is one which best matches the data type of the buffer.
Map a buffer and edit it:
arr = mapBuffer(vbo) arr[:, :] += 2.0 # add 2 to all values unmapBuffer(vbo) # call when done # Don't ever modify `arr` after calling `unmapBuffer`. Delete it if # necessary to prevent it form being used. del arr
Modify a sub-range of data by specifying start and length, indices correspond to values, not byte offsets:
arr = mapBuffer(vbo, start=12, end=24) arr[:, :] *= 10.0 unmapBuffer(vbo)