components/rappor/rappor_metric.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/rappor_metric.h"
   6
   7 #include "base/logging.h"
   8 #include "base/rand_util.h"
   9
  10 namespace rappor {
  11
  12 RapporMetric::RapporMetric(const std::string& metric_name,
  13                            const RapporParameters& parameters,
  14                            int32_t cohort_seed)
  15     : metric_name_(metric_name),
  16       parameters_(parameters),
  17       sample_count_(0),
  18       bloom_filter_(parameters.bloom_filter_size_bytes,
  19                     parameters.bloom_filter_hash_function_count,
  20                     (cohort_seed % parameters.num_cohorts) *
  21                         parameters.bloom_filter_hash_function_count) {
  22   DCHECK_GE(cohort_seed, 0);
  23   DCHECK_LT(cohort_seed, RapporParameters::kMaxCohorts);
  24   // Since cohort_seed is in the range [0, kMaxCohorts), num_cohorts should
  25   // divide kMaxCohorts for each cohort to have equal weight.
  26   DCHECK_EQ(0, RapporParameters::kMaxCohorts % parameters.num_cohorts);
  27 }
  28
  29 RapporMetric::~RapporMetric() {}
  30
  31 void RapporMetric::AddSample(const std::string& str) {
  32   ++sample_count_;
  33   // Replace the previous sample with a 1 in sample_count_ chance so that each
  34   // sample has equal probability of being reported.
  35   if (base::RandGenerator(sample_count_) == 0) {
  36     bloom_filter_.SetString(str);
  37   }
  38 }
  39
  40 ByteVector RapporMetric::GetReport(const std::string& secret) const {
  41   // Generate a deterministically random mask of fake data using the
  42   // client's secret key + real data as a seed.  The inclusion of the secret
  43   // in the seed avoids correlations between real and fake data.
  44   // The seed isn't a human-readable string.
  45   const std::string personalization_string = metric_name_ +
  46       std::string(bytes().begin(), bytes().end());
  47   HmacByteVectorGenerator hmac_generator(bytes().size(), secret,
  48                                          personalization_string);
  49   const ByteVector fake_mask =
  50       hmac_generator.GetWeightedRandomByteVector(parameters().fake_prob);
  51   ByteVector fake_bits =
  52       hmac_generator.GetWeightedRandomByteVector(parameters().fake_one_prob);
  53
  54   // Redact most of the real data by replacing it with the fake data, hiding
  55   // and limiting the amount of information an individual client reports on.
  56   const ByteVector* fake_and_redacted_bits =
  57       ByteVectorMerge(fake_mask, bytes(), &fake_bits);
  58
  59   // Generate biased coin flips for each bit.
  60   ByteVectorGenerator coin_generator(bytes().size());
  61   const ByteVector zero_coins =
  62       coin_generator.GetWeightedRandomByteVector(parameters().zero_coin_prob);
  63   ByteVector one_coins =
  64       coin_generator.GetWeightedRandomByteVector(parameters().one_coin_prob);
  65
  66   // Create a randomized response report on the fake and redacted data, sending
  67   // the outcome of flipping a zero coin for the zero bits in that data, and of
  68   // flipping a one coin for the one bits in that data, as the final report.
  69   return *ByteVectorMerge(*fake_and_redacted_bits, zero_coins, &one_coins);
  70 }
  71
  72 void RapporMetric::SetBytesForTesting(const ByteVector& bytes) {
  73   bloom_filter_.SetBytesForTesting(bytes);
  74 }
  75
  76 }  // namespace rappor