Mailman 3 February 2004 - Python-checkins

python/dist/src/Lib urllib2.py,1.53.6.2,1.53.6.3
by loewis＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/dist/src/Lib In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv21163/Lib Modified Files: Tag: release23-maint urllib2.py Log Message: Patch #817379: Allow for absolute ftp paths. Index: urllib2.py =================================================================== RCS file: /cvsroot/python/python/dist/src/Lib/urllib2.py,v retrieving revision 1.53.6.2 retrieving revision 1.53.6.3 diff -C2 -d -r1.53.6.2 -r1.53.6.3 *** urllib2.py 4 Nov 2003 13:53:38 -0000 1.53.6.2 --- urllib2.py 15 Feb 2004 20:52:00 -0000 1.53.6.3 *************** *** 1022,1027 **** port = ftplib.FTP_PORT path, attrs = splitattr(req.get_selector()) - path = unquote(path) dirs = path.split('/') dirs, file = dirs[:-1], dirs[-1] if dirs and not dirs[0]: --- 1022,1027 ---- port = ftplib.FTP_PORT path, attrs = splitattr(req.get_selector()) dirs = path.split('/') + dirs = map(unquote, dirs) dirs, file = dirs[:-1], dirs[-1] if dirs and not dirs[0]:

1 0

python/dist/src/Lib urllib2.py,1.60,1.61
by loewis＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/dist/src/Lib In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv20962/Lib Modified Files: urllib2.py Log Message: Patch #817379: Allow for absolute ftp paths. Backported to 2.3. Index: urllib2.py =================================================================== RCS file: /cvsroot/python/python/dist/src/Lib/urllib2.py,v retrieving revision 1.60 retrieving revision 1.61 diff -C2 -d -r1.60 -r1.61 *** urllib2.py 18 Jan 2004 20:29:54 -0000 1.60 --- urllib2.py 15 Feb 2004 20:51:39 -0000 1.61 *************** *** 1100,1105 **** port = ftplib.FTP_PORT path, attrs = splitattr(req.get_selector()) - path = unquote(path) dirs = path.split('/') dirs, file = dirs[:-1], dirs[-1] if dirs and not dirs[0]: --- 1100,1105 ---- port = ftplib.FTP_PORT path, attrs = splitattr(req.get_selector()) dirs = path.split('/') + dirs = map(unquote, dirs) dirs, file = dirs[:-1], dirs[-1] if dirs and not dirs[0]:

1 0

python/dist/src/Lib modulefinder.py,1.9,1.10
by nascheme＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/dist/src/Lib In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv1761/Lib Modified Files: modulefinder.py Log Message: Fix typo. Index: modulefinder.py =================================================================== RCS file: /cvsroot/python/python/dist/src/Lib/modulefinder.py,v retrieving revision 1.9 retrieving revision 1.10 diff -C2 -d -r1.9 -r1.10 *** modulefinder.py 12 Feb 2004 17:35:06 -0000 1.9 --- modulefinder.py 15 Feb 2004 16:43:20 -0000 1.10 *************** *** 63,67 **** def __repr__(self): ! s = "Module(%r" % % (self.__name__,) if self.__file__ is not None: s = s + ", %r" % (self.__file__,) --- 63,67 ---- def __repr__(self): ! s = "Module(%r" % (self.__name__,) if self.__file__ is not None: s = s + ", %r" % (self.__file__,)

1 0

python/nondist/sandbox/statistics statistics.py, 1.3, 1.4
by rhettinger＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/nondist/sandbox/statistics In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv6713 Modified Files: statistics.py Log Message: More small improvements Index: statistics.py =================================================================== RCS file: /cvsroot/python/python/nondist/sandbox/statistics/statistics.py,v retrieving revision 1.3 retrieving revision 1.4 diff -C2 -d -r1.3 -r1.4 *** statistics.py 15 Feb 2004 07:22:33 -0000 1.3 --- statistics.py 15 Feb 2004 08:25:20 -0000 1.4 *************** *** 97,101 **** ua, nua = under.append, notunder.append it = iter(iterable) ! pivot = it.next() # must have at least one element for elem in it: if elem < pivot: --- 97,101 ---- ua, nua = under.append, notunder.append it = iter(iterable) ! pivot = it.next() # must have at least one element for elem in it: if elem < pivot: *************** *** 107,113 **** def select(data, n): """Find the nth rank ordered element (the least value has rank 0).""" ! if len(data) <= n: raise ValueError('not enough elements for the given rank') ! while 1: p, pivot, q = _partition(data) if n < len(p): --- 107,118 ---- def select(data, n): """Find the nth rank ordered element (the least value has rank 0).""" ! try: ! size = len(data) ! except TypeError: ! data = tuple(data) ! size = len(data) ! if size <= n: raise ValueError('not enough elements for the given rank') ! while True: p, pivot, q = _partition(data) if n < len(p): *************** *** 126,130 **** n = len(data) except TypeError: ! data = list(data) n = len(data) if n == 0: --- 131,135 ---- n = len(data) except TypeError: ! data = tuple(data) n = len(data) if n == 0: *************** *** 138,142 **** return reduce(operator.mul, iterable, 1) ! def nlargest(iterable, n=1): # XXX key= or cmp= argument """Find the n largest elements in the dataset""" ## When collections.fibheap is available, use it instead of heapq --- 143,147 ---- return reduce(operator.mul, iterable, 1) ! def nlargest(iterable, n=1): # XXX implement key= """Find the n largest elements in the dataset""" ## When collections.fibheap is available, use it instead of heapq

1 0

python/nondist/sandbox/statistics statistics.py, 1.2, 1.3
by rhettinger＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/nondist/sandbox/statistics In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv29711 Modified Files: statistics.py Log Message: Small cleanups. Index: statistics.py =================================================================== RCS file: /cvsroot/python/python/nondist/sandbox/statistics/statistics.py,v retrieving revision 1.2 retrieving revision 1.3 diff -C2 -d -r1.2 -r1.3 *** statistics.py 15 Feb 2004 06:45:09 -0000 1.2 --- statistics.py 15 Feb 2004 07:22:33 -0000 1.3 *************** *** 7,13 **** and make only a single pass over the data. This makes the tools suitable for use with generator expressions and keeps them as memory friendly as ! possible. Except for median(), the use of iterators means that the ! dataset can be processed one element at a time without bringing it all ! into memory. The functions, nlargest and nsmallest, are also designed to use the fewest --- 7,13 ---- and make only a single pass over the data. This makes the tools suitable for use with generator expressions and keeps them as memory friendly as ! possible. Except for median() and select(), the use of iterators means ! that the dataset can be processed one element at a time without bringing ! it all into memory. The functions, nlargest and nsmallest, are also designed to use the fewest *************** *** 15,21 **** as strings) as long they support __cmp__ or __lt__. - The code is designed to run at its best on Py2.4; however, - it being kept fully compatible with Py2.3. - Written and maintained by Raymond D. Hettinger <python at rcn.com>. Copyright (c) 2004 Python Software Foundation. All rights reserved. --- 15,18 ---- *************** *** 48,52 **** raise ValueError('data must have at least one element') - def stddev(data, sample=True): """Computes the standard deviation of the dataset. --- 45,48 ---- *************** *** 101,117 **** ua, nua = under.append, notunder.append it = iter(iterable) ! try: ! pivot = it.next() ! except StopIteration: ! pass for elem in it: if elem < pivot: ! under.append(elem) else: ! notunder.append(elem) return under, pivot, notunder def select(data, n): ! """Find the nth rank ordered element""" while 1: p, pivot, q = _partition(data) --- 97,112 ---- ua, nua = under.append, notunder.append it = iter(iterable) ! pivot = it.next() # must have at least one element for elem in it: if elem < pivot: ! ua(elem) else: ! nua(elem) return under, pivot, notunder def select(data, n): ! """Find the nth rank ordered element (the least value has rank 0).""" ! if len(data) <= n: ! raise ValueError('not enough elements for the given rank') while 1: p, pivot, q = _partition(data) *************** *** 120,127 **** elif n == len(p): return pivot ! else : data = q n -= len(p) + 1 - p = q = None # Free unused partition def median(data): --- 115,121 ---- elif n == len(p): return pivot ! else: data = q n -= len(p) + 1 def median(data): *************** *** 163,168 **** ## When collections.fibheap is available, use it instead of bisect it = iter(iterable) ! result = list(itertools.islice(it, 0, n)) ! result.sort() if len(result) == 0: raise ValueError('data must have at least one element') --- 157,161 ---- ## When collections.fibheap is available, use it instead of bisect it = iter(iterable) ! result = sorted(itertools.islice(it, 0, n)) if len(result) == 0: raise ValueError('data must have at least one element')

1 0

python/nondist/sandbox/statistics statistics.py, 1.1, 1.2
by rhettinger＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/nondist/sandbox/statistics In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv25075 Modified Files: statistics.py Log Message: Add remaining docstrings. Index: statistics.py =================================================================== RCS file: /cvsroot/python/python/nondist/sandbox/statistics/statistics.py,v retrieving revision 1.1 retrieving revision 1.2 diff -C2 -d -r1.1 -r1.2 *** statistics.py 15 Feb 2004 06:03:57 -0000 1.1 --- statistics.py 15 Feb 2004 06:45:09 -0000 1.2 *************** *** 29,33 **** import bisect ! __all__ = ['mean', 'stddev', 'product', 'nlargest', 'nsmallest', 'select'] def mean(data): --- 29,34 ---- import bisect ! __all__ = ['mean', 'median', 'nlargest', 'nsmallest', ! 'product', 'select', 'stddev'] def mean(data): *************** *** 89,94 **** return (s / k) ** 0.5 # population standard deviation ! def partition(iterable): ! # Support routine for select under, notunder = [], [] ua, nua = under.append, notunder.append --- 90,101 ---- return (s / k) ** 0.5 # population standard deviation ! def _partition(iterable): ! """ Private function supporting select() ! ! Chooses the first element as pivot and returns tuple with a ! list of values less than the pivot, the pivot itself, and a ! a list of values not less than the pivot. ! """ ! under, notunder = [], [] ua, nua = under.append, notunder.append *************** *** 106,112 **** def select(data, n): ! 'Find the nth rank ordered element' while 1: ! p, pivot, q = partition(data) if n < len(p): data = p --- 113,119 ---- def select(data, n): ! """Find the nth rank ordered element""" while 1: ! p, pivot, q = _partition(data) if n < len(p): data = p *************** *** 116,122 **** data = q n -= len(p) + 1 ! p = q = None def median(data): try: n = len(data) --- 123,132 ---- data = q n -= len(p) + 1 ! p = q = None # Free unused partition def median(data): + """Find a value such that half of the data elements are above + the value and half are below. + """ try: n = len(data) *************** *** 131,139 **** def product(iterable): ! "Compute the product of the data elements" return reduce(operator.mul, iterable, 1) def nlargest(iterable, n=1): # XXX key= or cmp= argument ! "Compute the n largest elements in the dataset" ## When collections.fibheap is available, use it instead of heapq it = iter(iterable) --- 141,149 ---- def product(iterable): ! """Compute the product of the data elements""" return reduce(operator.mul, iterable, 1) def nlargest(iterable, n=1): # XXX key= or cmp= argument ! """Find the n largest elements in the dataset""" ## When collections.fibheap is available, use it instead of heapq it = iter(iterable) *************** *** 150,154 **** def nsmallest(iterable, n=1): ! "Compute the n largest elements in the dataset" ## When collections.fibheap is available, use it instead of bisect it = iter(iterable) --- 160,164 ---- def nsmallest(iterable, n=1): ! """Find the n smallest elements in the dataset""" ## When collections.fibheap is available, use it instead of bisect it = iter(iterable)

1 0

python/nondist/sandbox/statistics statistics.py, NONE, 1.1 test_statistics.py, NONE, 1.1
by rhettinger＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/nondist/sandbox/statistics In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv20081 Added Files: statistics.py test_statistics.py Log Message: Load a draft version of a statistics module. --- NEW FILE: statistics.py --- """statistics.py A collection of functions for summarizing data The module focuses on everyday data reduction functions and omits more sophisticated statistical tools. Unless otherwise noted, each function is designed to accept any iterable and make only a single pass over the data. This makes the tools suitable for use with generator expressions and keeps them as memory friendly as possible. Except for median(), the use of iterators means that the dataset can be processed one element at a time without bringing it all into memory. The functions, nlargest and nsmallest, are also designed to use the fewest possible data comparisions and the data elements may be non-numerical (such as strings) as long they support __cmp__ or __lt__. The code is designed to run at its best on Py2.4; however, it being kept fully compatible with Py2.3. Written and maintained by Raymond D. Hettinger <python at rcn.com>. Copyright (c) 2004 Python Software Foundation. All rights reserved. """ from __future__ import division import operator import itertools import heapq import bisect __all__ = ['mean', 'stddev', 'product', 'nlargest', 'nsmallest', 'select'] def mean(data): """Compute the arithmetic mean of a data sample""" try: cnt = len(data) # do it the fast way if possible except TypeError: cum = 0 for cnt, elem in enumerate(data): cum += elem cnt += 1 # enumerate counts from zero else: cum = sum(data) try: return cum / cnt # relies on true division except ZeroDivisionError: raise ValueError('data must have at least one element') def stddev(data, sample=True): """Computes the standard deviation of the dataset. If sample is True, computes the standard deviation with an n-1 divisor for an unbiased estimator for a data sample. If sample is False, computes with a divisor of n, giving the standard deviation for a complete population. """ # Formula for recurrence taken from Seminumerical Algorithms, Knuth, 4.2.2 # It has substantially better numerical performance than computing the # result from sum(data) and sum(x*x for x in data). it = iter(data) try: m = it.next() # running mean except StopIteration: raise ValueError('data must have at least one element') s = 0 # running sum((x-mean)**2 for x in data) k = 1 # number of items dm = 0 # carried forward error term for x in it: k += 1 # This block computes: newm = (x-m)/k # The use of a cumulative error term improves accuracy when the mean is # much larger than the standard deviation. Also, it makes the formula # less sensitive to data order (with sorted data resulting is larger # relative errors according to experiments by Chris Reedy). adjm = (x-m)/k - dm # relies on true division newm = m + adjm dm = (newm - m) - adjm s += (x-m)*(x-newm) m = newm if sample: try: return (s / (k-1)) ** 0.5 # sample standard deviation except ZeroDivisionError: raise ValueError('sample deviation requires at least two elements') else: return (s / k) ** 0.5 # population standard deviation def partition(iterable): # Support routine for select under, notunder = [], [] ua, nua = under.append, notunder.append it = iter(iterable) try: pivot = it.next() except StopIteration: pass for elem in it: if elem < pivot: under.append(elem) else: notunder.append(elem) return under, pivot, notunder def select(data, n): 'Find the nth rank ordered element' while 1: p, pivot, q = partition(data) if n < len(p): data = p elif n == len(p): return pivot else : data = q n -= len(p) + 1 p = q = None def median(data): try: n = len(data) except TypeError: data = list(data) n = len(data) if n == 0: raise ValueError('data must have at least one element') if n & 1 == 1: # if length is an odd number return select(data, n//2) return (select(data, n//2) + select(data, n//2-1)) / 2 def product(iterable): "Compute the product of the data elements" return reduce(operator.mul, iterable, 1) def nlargest(iterable, n=1): # XXX key= or cmp= argument "Compute the n largest elements in the dataset" ## When collections.fibheap is available, use it instead of heapq it = iter(iterable) result = list(itertools.islice(it, n)) heapq.heapify(result) if len(result) == 0: raise ValueError('data must have at least one element') subst = heapq.heapreplace for elem in it: if (elem > result[0]): subst(result, elem) result.sort(reverse=True) return result def nsmallest(iterable, n=1): "Compute the n largest elements in the dataset" ## When collections.fibheap is available, use it instead of bisect it = iter(iterable) result = list(itertools.islice(it, 0, n)) result.sort() if len(result) == 0: raise ValueError('data must have at least one element') insort = bisect.insort pop = result.pop for elem in it: if (elem < result[-1]): insort(result, elem) pop() return result XXX = """ Other possible functions include data groupers for binning, counting, and splitting into equivalence classes. """ --- NEW FILE: test_statistics.py --- from __future__ import division from statistics import mean, stddev, product, nlargest, nsmallest, select, median ### UNITTESTS ################################################# import unittest import random def g(n): "iterator substitute for xrange (without defining __len__)" for i in xrange(n): yield i class TestStats(unittest.TestCase): def test_mean(self): self.assertEqual(mean(range(6)), 15/6.0) self.assertEqual(mean(g(6)), 15/6.0) self.assertEqual(mean([10]), 10) self.assertRaises(ValueError, mean, []) self.assertRaises(TypeError, mean, 'abc') def test_stddev(self): self.assertEqual(stddev([10,15,20]), 5) self.assertEqual(round(stddev([11.1,4,9,13]), 3), 3.878) self.assertEqual(round(stddev([11.1,4,9,13], False), 3), 3.358) self.assertEqual(stddev([10], False), 0.0) self.assertRaises(ValueError, stddev, [1]) self.assertRaises(ValueError, stddev, [], False) def test_median(self): data = range(10) random.shuffle(data) copy = data[:] self.assertEqual(median(data), 4.5) self.assertEqual(data, copy) self.assertEqual(median(g(10)), 4.5) data.insert(1, 10) self.assertEqual(median(data), 5) self.assertEqual(median([-50,10,2,11]), 6) self.assertEqual(median([30,10,2]), 10) self.assertRaises(ValueError, median, []) def test_product(self): self.assertEqual(product(range(1, 5)), 24) def g(lo, hi): for i in xrange(lo, hi): yield i self.assertEqual(product(g(1, 5)), 24) def test_select(self): n = 200 a = range(n) random.shuffle(a) for i in xrange(100): nth = random.randrange(n) self.assertEqual(select(a, nth), nth) def test_nlargest(self): n = 10 data = range(n) random.shuffle(data) copy = data[:] self.assertEqual(nlargest(data), [n-1]) self.assertRaises(ValueError, nlargest, data, 0) for i in xrange(1, n+1): self.assertEqual(nlargest(data, i), range(n-1, n-i-1, -1)) self.assertEqual(data, copy) self.assertEqual(nlargest('abcde', 3), list('edc')) def test_nsmallest(self): n = 10 data = range(n) random.shuffle(data) copy = data[:] self.assertEqual(nsmallest(data), [0]) self.assertRaises(ValueError, nlargest, data, 0) for i in xrange(1, n+1): self.assertEqual(nsmallest(data, i), range(i)) self.assertEqual(data, copy) self.assertEqual(nsmallest('abcde', 3), list('abc')) if __name__ == '__main__': suite = unittest.TestSuite() suite.addTest(unittest.makeSuite(TestStats)) unittest.TextTestRunner(verbosity=2).run(suite)

1 0

python/nondist/sandbox/statistics - New directory
by rhettinger＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/nondist/sandbox/statistics In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv19938/statistics Log Message: Directory /cvsroot/python/python/nondist/sandbox/statistics added to the repository

1 0

python/dist/src/Python newcompile.c, 1.1.2.72, 1.1.2.73
by jhylton＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/dist/src/Python In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv6712 Modified Files: Tag: ast-branch newcompile.c Log Message: Add comment from Tim about basic blocks representation. Index: newcompile.c =================================================================== RCS file: /cvsroot/python/python/dist/src/Python/Attic/newcompile.c,v retrieving revision 1.1.2.72 retrieving revision 1.1.2.73 diff -C2 -d -r1.1.2.72 -r1.1.2.73 *** newcompile.c 15 Feb 2004 04:18:13 -0000 1.1.2.72 --- newcompile.c 15 Feb 2004 04:19:28 -0000 1.1.2.73 *************** *** 45,48 **** --- 45,55 ---- of function `PyParser_SimpleParseString' #: compile.h::b_return is only set, never used + + ISSUES: + + Tim mentioned that it's more common for a basic blocks representation + to use real pointers for jump targets rather than indexes into an + array of basic blocks. + */

1 0

python/dist/src/Python newcompile.c, 1.1.2.71, 1.1.2.72
by jhylton＠users.sourceforge.net Feb. 15, 2004

Feb. 15, 2004

Update of /cvsroot/python/python/dist/src/Python In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv6574 Modified Files: Tag: ast-branch newcompile.c Log Message: Fixed. Index: newcompile.c =================================================================== RCS file: /cvsroot/python/python/dist/src/Python/Attic/newcompile.c,v retrieving revision 1.1.2.71 retrieving revision 1.1.2.72 diff -C2 -d -r1.1.2.71 -r1.1.2.72 *** newcompile.c 15 Feb 2004 04:14:03 -0000 1.1.2.71 --- newcompile.c 15 Feb 2004 04:18:13 -0000 1.1.2.72 *************** *** 41,45 **** #: ref leaks in interpreter when press return on empty line #: yield or return outside a function don't raise a SyntaxError - #: LOAD_NAME is output instead of LOAD_GLOBAL #: line numbers are off a bit (may just need to add calls to set lineno) #: Modules/parsermodule.c:496: warning: implicit declaration --- 41,44 ----

1 0