components/rappor/bloom_filter.cc

   1 // Copyright 2014 The Chromium Authors. All rights reserved.
   2 // Use of this source code is governed by a BSD-style license that can be
   3 // found in the LICENSE file.
   4
   5 #include "components/rappor/bloom_filter.h"
   6
   7 #include "base/logging.h"
   8 #include "third_party/smhasher/src/City.h"
   9
  10 namespace rappor {
  11
  12 BloomFilter::BloomFilter(uint32_t bytes_size,
  13                          uint32_t hash_function_count,
  14                          uint32_t hash_seed_offset)
  15     : bytes_(bytes_size),
  16       hash_function_count_(hash_function_count),
  17       hash_seed_offset_(hash_seed_offset) {
  18   DCHECK_GT(bytes_size, 0u);
  19 }
  20
  21 BloomFilter::~BloomFilter() {}
  22
  23 void BloomFilter::SetString(const std::string& str) {
  24   for (size_t i = 0; i < bytes_.size(); ++i) {
  25     bytes_[i] = 0;
  26   }
  27   for (size_t i = 0; i < hash_function_count_; ++i) {
  28     // Using CityHash here because we have support for it in Dremel.  Many hash
  29     // functions, such as MD5, SHA1, or Murmur, would probably also work.
  30     uint32_t index =
  31         CityHash64WithSeed(str.data(), str.size(), hash_seed_offset_ + i);
  32     // Note that the "bytes" are uint8_t, so they are always 8-bits.
  33     uint32_t byte_index = (index / 8) % bytes_.size();
  34     uint32_t bit_index = index % 8;
  35     bytes_[byte_index] |= 1 << bit_index;
  36   }
  37 }
  38
  39 void BloomFilter::SetBytesForTesting(const ByteVector& bytes) {
  40   DCHECK_EQ(bytes_.size(), bytes.size());
  41   for (size_t i = 0; i < bytes_.size(); ++i) {
  42     bytes_[i] = bytes[i];
  43   }
  44 }
  45
  46 }  // namespace rappor