[Numpy-tickets] [NumPy] #982: Problem accessing data attribute of ndarray object created with certain sizes

[NumPy]

#982: Problem accessing data attribute of ndarray object created with certain
sizes
------------------------+---------------------------------------------------
 Reporter:  jtaylor2    |       Owner:  somebody
     Type:  defect      |      Status:  new     
 Priority:  normal      |   Milestone:  1.3.0   
Component:  numpy.core  |     Version:  none    
 Severity:  normal      |    Keywords:          
------------------------+---------------------------------------------------
 When creating a ndarray object on a 64 bit machine where the number of
 bytes of data in the array is described by the following rule:

 (size & 0x80000000) != 0 AND (size & 0xFFFFFFFF) != 0xFFFFFFFF

 Accessing the data attribute (the data buffer) of the ndarray object
 results in a ValueError exception.

 Note the following examples:

 {{{

 thor> python
 Python 2.5.1 (r251:54863, Jan 16 2008, 16:16:13)
 [GCC 3.4.6 20060404 (Red Hat 3.4.6-9)] on linux2
 Type "help", "copyright", "credits" or "license" for more information.
 >>> import numpy
 >>> numpy.__version__
 '1.3.0.dev6306'
 >>> a=numpy.zeros(0x7FFFFFFF,dtype='b1')
 >>> a.data
 <read-write buffer for 0xa054a0, size 2147483647, offset 0 at
 0x2a99e13e70>
 >>> a=numpy.zeros(0x80000000,dtype='b1')
 >>> a.data
 Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
 ValueError: size must be zero or positive
 >>> a=numpy.zeros(0xFFFFFFFE,dtype='b1')
 >>> a.data
 Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
 ValueError: size must be zero or positive
 >>> a=numpy.zeros(0xFFFFFFFF,dtype='b1')
 >>> a.data
 <read-write buffer for 0xa2e410, size -1, offset 0 at 0x2a99e13e30>
 >>> a=numpy.zeros(0x180000000,dtype='b1')
 >>> a.data
 Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
 ValueError: size must be zero or positive
 >>> a=numpy.zeros(0x180000001,dtype='b1')
 >>> a.data
 Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
 ValueError: size must be zero or positive
 >>> a=numpy.zeros(0x1FFFFFFFF,dtype='b1')
 >>> a.data
 <read-write buffer for 0xa2e590, size -1, offset 0 at 0x2a99e13e70>

 }}}

-- 
Ticket URL: <http://scipy.org/scipy/numpy/ticket/982>
NumPy <http://projects.scipy.org/scipy/numpy>
The fundamental package needed for scientific computing with Python.