Merge pull request #4 from joseph-fox/fix-overflow-error

Replace xrange by count

Author: joseph-fox

CHANGES.txt

@@ -1,3 +1,9 @@
+Changes in 2.2
+==============
+Replaced the xrange by count, so it is a iterator now.
+This fixes overflow error when a large integer is passed
+the range_fun.
+
 Changes in 2.1
 ==============
 The tightening ratio is 0.9, and it is consistently used.

pybloom_live/pybloom.py

@@ -74,7 +74,7 @@ def make_hashfuncs(num_slices, num_bits):
     num_salts, extra = divmod(num_slices, len(fmt))
     if extra:
         num_salts += 1
-    salts = tuple(hashfn(hashfn(pack('I', i)).digest()) for i in range_fn(num_salts))
+    salts = tuple(hashfn(hashfn(pack('I', i)).digest()) for i in range_fn(0, num_salts))
 
     def _make_hashfuncs(key):
         if running_python_3:

pybloom_live/test_pybloom.py

@@ -82,7 +82,7 @@ def _run():
 
 class Serialization(unittest.TestCase):
     SIZE = 12345
-    EXPECTED = set([random.randint(0, 10000100) for _ in range_fn(SIZE)])
+    EXPECTED = set([random.randint(0, 10000100) for _ in range_fn(0, SIZE)])
 
     def test_serialization(self):
         for klass, args in [(BloomFilter, (self.SIZE,)),

pybloom_live/utils.py

@@ -1,4 +1,6 @@
 import sys
+import itertools
+
 try:
     import StringIO
     import cStringIO
@@ -8,11 +10,11 @@
 running_python_3 = sys.version_info[0] == 3
 
 
-def range_fn(*args):
+def range_fn(start=0, stop=None):
     if running_python_3:
-        return range(*args)
+        return range(start, stop)
     else:
-        return xrange(*args)
+        return iter(itertools.count(start).next, stop)
 
 
 def is_string_io(instance):

setup.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python
 from setuptools import setup
 
-VERSION = '2.1.0'
+VERSION = '2.2.0'
 DESCRIPTION = "Bloom filter: A Probabilistic data structure"
 LONG_DESCRIPTION = """
 This bloom filter is forked from pybloom, and its tightening ratio is changed to 0.9, and this ration is consistently used. Choosing r around 0.8 - 0.9 will result in better average space usage for wide range of growth, therefore the default value of model is set to LARGE_SET_GROWTH.