From e341600ffb46733e588dcbda8a1600f583248acd Mon Sep 17 00:00:00 2001
From: Davis King <davis@dlib.net>
Date: Mon, 21 Jan 2013 17:06:27 -0500
Subject: [PATCH] Added count_bits() and hamming_distance()

---
 dlib/general_hash/count_bits.h          | 62 +++++++++++++++++++++++++
 dlib/general_hash/count_bits_abstract.h | 48 +++++++++++++++++++
 dlib/hash.h                             |  1 +
 dlib/test/hash.cpp                      | 51 ++++++++++++++++++++
 4 files changed, 162 insertions(+)
 create mode 100644 dlib/general_hash/count_bits.h
 create mode 100644 dlib/general_hash/count_bits_abstract.h

diff --git a/dlib/general_hash/count_bits.h b/dlib/general_hash/count_bits.h
new file mode 100644
index 000000000..4de771ae0
--- /dev/null
+++ b/dlib/general_hash/count_bits.h
@@ -0,0 +1,62 @@
+// Copyright (C) 2013  Davis E. King (davis@dlib.net)
+// License: Boost Software License   See LICENSE.txt for the full license.
+#ifndef DLIB_COUNT_BiTS_H__
+#define DLIB_COUNT_BiTS_H__
+
+#include "../algs.h"
+#include <climits>
+
+namespace dlib
+{
+
+// ----------------------------------------------------------------------------------------
+
+    template <
+        typename T
+        >
+    T count_bits (
+        T v
+    )
+    /*!
+        requires
+            - T is an unsigned integral type
+        ensures
+            - returns the number of bits in v which are set to 1.
+    !*/
+    {
+        COMPILE_TIME_ASSERT(is_unsigned_type<T>::value && sizeof(T) <= 8);
+
+        // This bit of bit trickery is from:
+        // http://graphics.stanford.edu/~seander/bithacks.html#CountBitsSet64
+
+        v = v - ((v >> 1) & (T)~(T)0/3);                           
+        v = (v & (T)~(T)0/15*3) + ((v >> 2) & (T)~(T)0/15*3);      
+        v = (v + (v >> 4)) & (T)~(T)0/255*15;                      
+        return (T)(v * ((T)~(T)0/255)) >> (sizeof(T) - 1) * CHAR_BIT; 
+    }
+
+// ----------------------------------------------------------------------------------------
+
+    template <
+        typename T
+        >
+    T hamming_distance (
+        const T& a,
+        const T& b
+    )
+    /*!
+        requires
+            - T is an unsigned integral type
+        ensures
+            - returns the number of bits which differ between a and b.
+    !*/
+    {
+        return count_bits(a^b);
+    }
+
+// ----------------------------------------------------------------------------------------
+
+}
+
+#endif // DLIB_COUNT_BiTS_H__
+
diff --git a/dlib/general_hash/count_bits_abstract.h b/dlib/general_hash/count_bits_abstract.h
new file mode 100644
index 000000000..ba3dae583
--- /dev/null
+++ b/dlib/general_hash/count_bits_abstract.h
@@ -0,0 +1,48 @@
+// Copyright (C) 2013  Davis E. King (davis@dlib.net)
+// License: Boost Software License   See LICENSE.txt for the full license.
+#undef DLIB_COUNT_BiTS_ABSTRACT_H__
+#ifdef DLIB_COUNT_BiTS_ABSTRACT_H__
+
+
+namespace dlib
+{
+
+// ----------------------------------------------------------------------------------------
+
+    template <
+        typename T
+        >
+    T count_bits (
+        T v
+    );
+    /*!
+        requires
+            - T is an unsigned integral type
+        ensures
+            - returns the number of bits in v which are set to 1.
+    !*/
+
+// ----------------------------------------------------------------------------------------
+
+    template <
+        typename T
+        >
+    T hamming_distance (
+        const T& a,
+        const T& b
+    );
+    /*!
+        requires
+            - T is an unsigned integral type
+        ensures
+            - returns the number of bits which differ between a and b.  (I.e. returns
+              count_bits(a^b).)
+    !*/
+
+// ----------------------------------------------------------------------------------------
+
+}
+
+#endif // DLIB_COUNT_BiTS_ABSTRACT_H__
+
+
diff --git a/dlib/hash.h b/dlib/hash.h
index 82115c698..5a018b438 100644
--- a/dlib/hash.h
+++ b/dlib/hash.h
@@ -6,6 +6,7 @@
 
 #include "general_hash/hash.h"
 #include "general_hash/random_hashing.h"
+#include "general_hash/count_bits.h"
 
 
 #endif // DLIB_HASh_
diff --git a/dlib/test/hash.cpp b/dlib/test/hash.cpp
index 10fb50606..4db326f23 100644
--- a/dlib/test/hash.cpp
+++ b/dlib/test/hash.cpp
@@ -205,6 +205,55 @@ namespace
         }
     }
 
+// ----------------------------------------------------------------------------------------
+
+    uint64  slow_count_bits ( uint64 v)
+    {
+        uint64 count = 0;
+        for (int i = 0; i < 64; ++i)
+        {
+            if (v&1)
+                ++count;
+            v >>= 1;
+        }
+        return count;
+    }
+
+
+    uint32  slow_count_bits ( uint32 v)
+    {
+        uint32 count = 0;
+        for (int i = 0; i < 32; ++i)
+        {
+            if (v&1)
+                ++count;
+            v >>= 1;
+        }
+        return count;
+    }
+
+
+// ----------------------------------------------------------------------------------------
+
+    void test_hamming_stuff()
+    {
+        dlib::rand rnd;
+        for (int i = 0; i < 10000; ++i)
+        {
+            uint32 v = rnd.get_random_32bit_number();
+            uint64 v2 = rnd.get_random_64bit_number();
+            DLIB_TEST(slow_count_bits(v) == count_bits(v));
+            DLIB_TEST(slow_count_bits(v2) == count_bits(v2));
+        }
+
+        DLIB_TEST(hamming_distance((uint32)0x1F, (uint32)0x0F) == 1);
+        DLIB_TEST(hamming_distance((uint32)0x1F, (uint32)0x1F) == 0);
+        DLIB_TEST(hamming_distance((uint32)0x1F, (uint32)0x19) == 2);
+        DLIB_TEST(hamming_distance((uint32)0x2F, (uint32)0x19) == 4);
+    }
+
+// ----------------------------------------------------------------------------------------
+
     class test_hash : public tester
     {
     public:
@@ -219,6 +268,8 @@ namespace
         {
             print_spinner();
 
+            test_hamming_stuff();
+
             murmur_hash_test();
             murmur_hash_128_test();