iterkeys, keys, iteritems, values, itervalues

bounter/count_min_sketch.py

@@ -63,10 +63,10 @@ def __init__(self, size_mb=64, width=None, depth=None, log_counting=None):
                 If width is not provided, the algorithm chooses the maximum width to fill the available size.
                 The more, the better, should be very large, preferably in the same order of magnitude as the cardinality
                 of the counted set.
-            log_counting (int): Use logarithmic counter value for reduced bucket size:
+            log_counting (int): Use logarithmic approximate counter value for reduced bucket size:
                 - None (default): 4B, no counter error
-                - 1024: 2B, counter error ~2% for values larger than 1024
-                - 8: 1B, counter error ~30% for values larger than 8
+                - 1024: 2B, value approximation error ~2% for values larger than 2048
+                - 8: 1B, value approximation error ~30% for values larger than 16
         """
 
         cell_size = CountMinSketch.cell_size(log_counting)

bounter/tests/hashtable/test_htc_iteration.py

@@ -0,0 +1,47 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+#
+# Author: Filip Stefanak <[email protected]>
+# Copyright (C) 2017 Rare Technologies
+#
+# This code is distributed under the terms and conditions
+# from the MIT License (MIT).
+
+import unittest
+
+from bounter import HashTable
+
+
+class HashTableIterationTest(unittest.TestCase):
+    """
+    Functional tests for additional HashTable iterations methods, which all produce an iterator
+    """
+
+    def setUp(self):
+        self.ht = HashTable(buckets=64)
+        self.ht.update([u"foo", u"bar", u"foo", "foo"])
+        self.keys = {u"foo", u"bar"}
+        self.values = {3, 1}
+        self.pairs = {(u"foo", 3), (u"bar", 1)}
+
+    def test_keys(self):
+        self.assertEqual(set(self.ht.keys()), self.keys)
+
+    def test_iterkeys(self):
+        self.assertEqual(set(self.ht.iterkeys()), self.keys)
+
+    def test_values(self):
+        self.assertEqual(set(self.ht.values()), self.values)
+
+    def test_itervalues(self):
+        self.assertEqual(set(self.ht.itervalues()), self.values)
+
+    def test_items(self):
+        self.assertEqual(set(self.ht.items()), self.pairs)
+
+    def test_iteritems(self):
+        self.assertEqual(set(self.ht.iteritems()), self.pairs)
+
+
+if __name__ == '__main__':
+    unittest.main()

cbounter/ht_common.c

@@ -797,6 +797,12 @@ PyObject* HT_VARIANT(_ITER_iternext)(HT_VARIANT(_ITER_TYPE) *self)
 
     if (i < buckets)
     {
+        if (self->result_type == ITER_RESULT_VALUES)
+        {
+            self->i = i + 1;
+            return Py_BuildValue("L", table[i].count);
+        }
+
         PyObject * result;
         char * current_key = table[i].key;
         PyObject * pkey;
@@ -915,6 +921,11 @@ HT_VARIANT(_HT_iter_K)(HT_TYPE *self)
     return HT_VARIANT(_make_iterator)(self, ITER_RESULT_KEYS);
 }
 
+static HT_VARIANT(_ITER_TYPE) *
+HT_VARIANT(_HT_iter_V)(HT_TYPE *self)
+{
+    return HT_VARIANT(_make_iterator)(self, ITER_RESULT_VALUES);
+}
 
 static PyMethodDef HT_VARIANT(_methods)[] = {
     {"increment", (PyCFunction)HT_VARIANT(_increment), METH_VARARGS,
@@ -929,6 +940,21 @@ static PyMethodDef HT_VARIANT(_methods)[] = {
     {"items", (PyCFunction)HT_VARIANT(_HT_iter_KV), METH_NOARGS,
      "Iterates over all key-value pairs."
     },
+    {"iteritems", (PyCFunction)HT_VARIANT(_HT_iter_KV), METH_NOARGS,
+     "Iterates over all key-value pairs."
+    },
+    {"keys", (PyCFunction)HT_VARIANT(_HT_iter_K), METH_NOARGS,
+     "Iterates over all keys."
+    },
+    {"iterkeys", (PyCFunction)HT_VARIANT(_HT_iter_K), METH_NOARGS,
+     "Iterates over all keys."
+    },
+    {"values", (PyCFunction)HT_VARIANT(_HT_iter_V), METH_NOARGS,
+     "Iterates over all non-zero counts."
+    },
+    {"itervalues", (PyCFunction)HT_VARIANT(_HT_iter_V), METH_NOARGS,
+     "Iterates over all non-zero counts."
+    },
     {"update", (PyCFunction)HT_VARIANT(_update), METH_VARARGS,
      "Adds all pairs from another counter, or adds all items from an iterable."
     },