Improve extended slicing support in builtin types and classes. Specifically:

- Specialcase extended slices that amount to a shallow copy the same way as
   is done for simple slices, in the tuple, string and unicode case.

 - Specialcase step-1 extended slices to optimize the common case for all
   involved types.

 - For lists, allow extended slice assignment of differing lengths as long
   as the step is 1. (Previously, 'l[:2:1] = []' failed even though
   'l[:2] = []' and 'l[:2:None] = []' do not.)

 - Implement extended slicing for buffer, array, structseq, mmap and
   UserString.UserString.

 - Implement slice-object support (but not non-step-1 slice assignment) for
   UserString.MutableString.

 - Add tests for all new functionality.
This commit is contained in:
Thomas Wouters 2007-08-28 15:28:19 +00:00
parent 0f4a14b56f
commit 3ccec68a05
16 changed files with 730 additions and 120 deletions

View file

@ -149,15 +149,41 @@ def __init__(self, string=""):
def __hash__(self):
raise TypeError, "unhashable type (it is mutable)"
def __setitem__(self, index, sub):
if index < 0:
index += len(self.data)
if index < 0 or index >= len(self.data): raise IndexError
self.data = self.data[:index] + sub + self.data[index+1:]
if isinstance(index, slice):
if isinstance(sub, UserString):
sub = sub.data
elif not isinstance(sub, basestring):
sub = str(sub)
start, stop, step = index.indices(len(self.data))
if step == -1:
start, stop = stop+1, start+1
sub = sub[::-1]
elif step != 1:
# XXX(twouters): I guess we should be reimplementing
# the extended slice assignment/deletion algorithm here...
raise TypeError, "invalid step in slicing assignment"
start = min(start, stop)
self.data = self.data[:start] + sub + self.data[stop:]
else:
if index < 0:
index += len(self.data)
if index < 0 or index >= len(self.data): raise IndexError
self.data = self.data[:index] + sub + self.data[index+1:]
def __delitem__(self, index):
if index < 0:
index += len(self.data)
if index < 0 or index >= len(self.data): raise IndexError
self.data = self.data[:index] + self.data[index+1:]
if isinstance(index, slice):
start, stop, step = index.indices(len(self.data))
if step == -1:
start, stop = stop+1, start+1
elif step != 1:
# XXX(twouters): see same block in __setitem__
raise TypeError, "invalid step in slicing deletion"
start = min(start, stop)
self.data = self.data[:start] + self.data[stop:]
else:
if index < 0:
index += len(self.data)
if index < 0 or index >= len(self.data): raise IndexError
self.data = self.data[:index] + self.data[index+1:]
def __setslice__(self, start, end, sub):
start = max(start, 0); end = max(end, 0)
if isinstance(sub, UserString):

View file

@ -179,8 +179,10 @@ def test_setslice(self):
self.assertEqual(a, self.type2test(range(10)))
self.assertRaises(TypeError, a.__setslice__, 0, 1, 5)
self.assertRaises(TypeError, a.__setitem__, slice(0, 1, 5))
self.assertRaises(TypeError, a.__setslice__)
self.assertRaises(TypeError, a.__setitem__)
def test_delslice(self):
a = self.type2test([0, 1])

View file

@ -912,7 +912,6 @@ def test_subscript(self):
self.checkequal(u'abc', 'abc', '__getitem__', slice(0, 1000))
self.checkequal(u'a', 'abc', '__getitem__', slice(0, 1))
self.checkequal(u'', 'abc', '__getitem__', slice(0, 0))
# FIXME What about negative indices? This is handled differently by [] and __getitem__(slice)
self.checkraises(TypeError, 'abc', '__getitem__', 'def')
@ -926,10 +925,21 @@ def test_slice(self):
self.checkequal('', 'abc', '__getslice__', 1000, 1000)
self.checkequal('', 'abc', '__getslice__', 2000, 1000)
self.checkequal('', 'abc', '__getslice__', 2, 1)
# FIXME What about negative indizes? This is handled differently by [] and __getslice__
self.checkraises(TypeError, 'abc', '__getslice__', 'def')
def test_extended_getslice(self):
# Test extended slicing by comparing with list slicing.
s = string.ascii_letters + string.digits
indices = (0, None, 1, 3, 41, -1, -2, -37)
for start in indices:
for stop in indices:
# Skip step 0 (invalid)
for step in indices[1:]:
L = list(s)[start:stop:step]
self.checkequal(u"".join(L), s, '__getitem__',
slice(start, stop, step))
def test_mul(self):
self.checkequal('', 'abc', '__mul__', -1)
self.checkequal('', 'abc', '__mul__', 0)

View file

@ -474,6 +474,18 @@ def test_getslice(self):
array.array(self.typecode)
)
def test_extended_getslice(self):
# Test extended slicing by comparing with list slicing
# (Assumes list conversion works correctly, too)
a = array.array(self.typecode, self.example)
indices = (0, None, 1, 3, 19, 100, -1, -2, -31, -100)
for start in indices:
for stop in indices:
# Everything except the initial 0 (invalid step)
for step in indices[1:]:
self.assertEqual(list(a[start:stop:step]),
list(a)[start:stop:step])
def test_setslice(self):
a = array.array(self.typecode, self.example)
a[:1] = a
@ -557,12 +569,34 @@ def test_setslice(self):
a = array.array(self.typecode, self.example)
self.assertRaises(TypeError, a.__setslice__, 0, 0, None)
self.assertRaises(TypeError, a.__setitem__, slice(0, 0), None)
self.assertRaises(TypeError, a.__setitem__, slice(0, 1), None)
b = array.array(self.badtypecode())
self.assertRaises(TypeError, a.__setslice__, 0, 0, b)
self.assertRaises(TypeError, a.__setitem__, slice(0, 0), b)
self.assertRaises(TypeError, a.__setitem__, slice(0, 1), b)
def test_extended_set_del_slice(self):
indices = (0, None, 1, 3, 19, 100, -1, -2, -31, -100)
for start in indices:
for stop in indices:
# Everything except the initial 0 (invalid step)
for step in indices[1:]:
a = array.array(self.typecode, self.example)
L = list(a)
# Make sure we have a slice of exactly the right length,
# but with (hopefully) different data.
data = L[start:stop:step]
data.reverse()
L[start:stop:step] = data
a[start:stop:step] = array.array(self.typecode, data)
self.assertEquals(a, array.array(self.typecode, L))
del L[start:stop:step]
del a[start:stop:step]
self.assertEquals(a, array.array(self.typecode, L))
def test_index(self):
example = 2*self.example
a = array.array(self.typecode, example)

29
Lib/test/test_buffer.py Normal file
View file

@ -0,0 +1,29 @@
"""Unit tests for buffer objects.
For now, tests just new or changed functionality.
"""
import unittest
from test import test_support
class BufferTests(unittest.TestCase):
def test_extended_getslice(self):
# Test extended slicing by comparing with list slicing.
s = "".join(chr(c) for c in list(range(255, -1, -1)))
b = buffer(s)
indices = (0, None, 1, 3, 19, 300, -1, -2, -31, -300)
for start in indices:
for stop in indices:
# Skip step 0 (invalid)
for step in indices[1:]:
self.assertEqual(b[start:stop:step],
s[start:stop:step])
def test_main():
test_support.run_unittest(BufferTests)
if __name__ == "__main__":
test_main()

View file

@ -306,6 +306,40 @@ def test_anonymous(self):
m[x] = ch = chr(x & 255)
self.assertEqual(m[x], ch)
def test_extended_getslice(self):
# Test extended slicing by comparing with list slicing.
s = "".join(chr(c) for c in reversed(range(256)))
m = mmap.mmap(-1, len(s))
m[:] = s
self.assertEqual(m[:], s)
indices = (0, None, 1, 3, 19, 300, -1, -2, -31, -300)
for start in indices:
for stop in indices:
# Skip step 0 (invalid)
for step in indices[1:]:
self.assertEqual(m[start:stop:step],
s[start:stop:step])
def test_extended_set_del_slice(self):
# Test extended slicing by comparing with list slicing.
s = "".join(chr(c) for c in reversed(range(256)))
m = mmap.mmap(-1, len(s))
indices = (0, None, 1, 3, 19, 300, -1, -2, -31, -300)
for start in indices:
for stop in indices:
# Skip invalid step 0
for step in indices[1:]:
m[:] = s
self.assertEqual(m[:], s)
L = list(s)
# Make sure we have a slice of exactly the right length,
# but with different data.
data = L[start:stop:step]
data = "".join(reversed(data))
L[start:stop:step] = data
m[start:stop:step] = data
self.assertEquals(m[:], "".join(L))
def test_main():
run_unittest(MmapTests)

View file

@ -97,6 +97,18 @@ def test_reduce(self):
t = time.gmtime()
x = t.__reduce__()
def test_extended_getslice(self):
# Test extended slicing by comparing with list slicing.
t = time.gmtime()
L = list(t)
indices = (0, None, 1, 3, 19, 300, -1, -2, -31, -300)
for start in indices:
for stop in indices:
# Skip step 0 (invalid)
for step in indices[1:]:
self.assertEqual(list(t[start:stop:step]),
L[start:stop:step])
def test_main():
test_support.run_unittest(StructSeqTest)

View file

@ -3,6 +3,7 @@
# UserString instances should behave similar to builtin string objects.
import unittest
import string
from test import test_support, string_tests
from UserString import UserString, MutableString
@ -88,6 +89,28 @@ def test_delslice(self):
del s[-1:10]
self.assertEqual(s, "fo")
def test_extended_set_del_slice(self):
indices = (0, None, 1, 3, 19, 100, -1, -2, -31, -100)
orig = string.ascii_letters + string.digits
for start in indices:
for stop in indices:
# Use indices[1:] when MutableString can handle real
# extended slices
for step in (None, 1, -1):
s = self.type2test(orig)
L = list(orig)
# Make sure we have a slice of exactly the right length,
# but with (hopefully) different data.
data = L[start:stop:step]
data.reverse()
L[start:stop:step] = data
s[start:stop:step] = "".join(data)
self.assertEquals(s, "".join(L))
del L[start:stop:step]
del s[start:stop:step]
self.assertEquals(s, "".join(L))
def test_immutable(self):
s = self.type2test("foobar")
s2 = s.immutable()