Author: mattip <[email protected]>
Branch: numpy_broadcast
Changeset: r83712:16f4a95d75ee
Date: 2016-04-17 16:33 +0300
http://bitbucket.org/pypy/pypy/changeset/16f4a95d75ee/
Log: test, implement overflow checking; small cleanups
diff --git a/pypy/module/micronumpy/broadcast.py
b/pypy/module/micronumpy/broadcast.py
--- a/pypy/module/micronumpy/broadcast.py
+++ b/pypy/module/micronumpy/broadcast.py
@@ -1,6 +1,6 @@
import pypy.module.micronumpy.constants as NPY
from nditer import ConcreteIter, parse_op_flag, parse_op_arg
-from pypy.interpreter.error import OperationError
+from pypy.interpreter.error import OperationError, oefmt
from pypy.interpreter.gateway import interp2app
from pypy.interpreter.typedef import TypeDef, GetSetProperty
from pypy.module.micronumpy import support
@@ -8,6 +8,8 @@
from rpython.rlib import jit
from strides import calculate_broadcast_strides, shape_agreement_multiple
+def descr_new_broadcast(space, w_subtype, __args__):
+ return W_Broadcast(space, __args__.arguments_w)
class W_Broadcast(W_NumpyObject):
"""
@@ -15,15 +17,11 @@
This class is a simplified version of nditer.W_NDIter with fixed iteration
for broadcasted arrays.
"""
- @staticmethod
- def descr_new_broadcast(space, w_subtype, __args__):
- return W_Broadcast(space, __args__.arguments_w)
-
def __init__(self, space, args):
num_args = len(args)
if not (2 <= num_args <= NPY.MAXARGS):
- raise OperationError(space.w_ValueError,
- space.wrap("Need at least two and fewer than
(%d) array objects." % NPY.MAXARGS))
+ raise oefmt(space.w_ValueError,
+ "Need at least two and fewer than (%d) array
objects.", NPY.MAXARGS)
self.seq = [convert_to_array(space, w_elem)
for w_elem in args]
@@ -37,7 +35,10 @@
self.iters = []
self.index = 0
- self.size = support.product(self.shape)
+ try:
+ self.size = support.product_check(self.shape)
+ except OverflowError as e:
+ raise oefmt(space.w_ValueError, "broadcast dimensions too large.")
for i in range(len(self.seq)):
it = self.get_iter(space, i)
it.contiguous = False
@@ -99,7 +100,7 @@
W_Broadcast.typedef = TypeDef("numpy.broadcast",
-
__new__=interp2app(W_Broadcast.descr_new_broadcast),
+ __new__=interp2app(descr_new_broadcast),
__iter__=interp2app(W_Broadcast.descr_iter),
next=interp2app(W_Broadcast.descr_next),
shape=GetSetProperty(W_Broadcast.descr_get_shape),
diff --git a/pypy/module/micronumpy/test/test_broadcast.py
b/pypy/module/micronumpy/test/test_broadcast.py
--- a/pypy/module/micronumpy/test/test_broadcast.py
+++ b/pypy/module/micronumpy/test/test_broadcast.py
@@ -55,11 +55,19 @@
assert b == [(1, 4), (2, 5), (3, 6)]
assert b[0][0].dtype == x.dtype
- def test_broadcast_linear_unequal(self):
+ def test_broadcast_failures(self):
import numpy as np
+ import sys
x = np.array([1, 2, 3])
y = np.array([4, 5])
raises(ValueError, np.broadcast, x, y)
+ a = np.empty(2**16,dtype='int8')
+ a = a.reshape(-1, 1, 1, 1)
+ b = a.reshape(1, -1, 1, 1)
+ c = a.reshape(1, 1, -1, 1)
+ d = a.reshape(1, 1, 1, -1)
+ exc = raises(ValueError, np.broadcast, a, b, c, d)
+ assert exc.value[0] == ('broadcast dimensions too large.')
def test_broadcast_3_args(self):
import numpy as np
@@ -82,7 +90,8 @@
for j in range(35):
arrs = [arr] * j
if j < 2 or j > 32:
- raises(ValueError, np.broadcast, *arrs)
+ exc = raises(ValueError, np.broadcast, *arrs)
+ assert exc.value[0] == ('Need at least two and fewer than (32)
array objects.')
else:
mit = np.broadcast(*arrs)
assert mit.numiter == j
_______________________________________________
pypy-commit mailing list
[email protected]
https://mail.python.org/mailman/listinfo/pypy-commit
