Chromium Code Reviews| OLD | NEW |
|---|---|
| 1 // Copyright 2016 The Chromium Authors. All rights reserved. | 1 // Copyright 2016 The Chromium Authors. All rights reserved. |
| 2 // Use of this source code is governed by a BSD-style license that can be | 2 // Use of this source code is governed by a BSD-style license that can be |
| 3 // found in the LICENSE file. | 3 // found in the LICENSE file. |
| 4 | 4 |
| 5 #include <algorithm> | |
| 6 #include <vector> | |
| 7 | |
| 5 #include "base/logging.h" | 8 #include "base/logging.h" |
| 6 #include "base/numerics/safe_math.h" | 9 #include "base/numerics/safe_math.h" |
| 7 #include "base/strings/stringprintf.h" | 10 #include "base/strings/stringprintf.h" |
| 8 #include "components/safe_browsing_db/v4_rice.h" | 11 #include "components/safe_browsing_db/v4_rice.h" |
| 9 | 12 |
| 10 using ::google::protobuf::RepeatedField; | 13 using ::google::protobuf::RepeatedField; |
| 11 using ::google::protobuf::int32; | 14 using ::google::protobuf::int32; |
| 15 using ::google::protobuf::int64; | |
| 12 | 16 |
| 13 namespace safe_browsing { | 17 namespace safe_browsing { |
| 14 | 18 |
| 15 namespace { | 19 namespace { |
| 16 | 20 |
| 17 const int kBitsPerByte = 8; | 21 const int kBitsPerByte = 8; |
| 18 const unsigned int kMaxBitIndex = kBitsPerByte * sizeof(uint32_t); | 22 const unsigned int kMaxBitIndex = kBitsPerByte * sizeof(uint32_t); |
| 19 | 23 |
| 20 void GetBytesFromUInt32(uint32_t word, char* bytes) { | 24 uint32_t FlipEndianness(uint32_t word) { |
| 21 const size_t mask = 0xFF; | 25 return (word >> 24) | ((word << 8) & 0x00FF0000) | |
| 22 bytes[0] = (char)(word & mask); | 26 ((word >> 8) & 0x0000FF00) | (word << 24); |
| 23 bytes[1] = (char)((word >> 8) & mask); | |
| 24 bytes[2] = (char)((word >> 16) & mask); | |
| 25 bytes[3] = (char)((word >> 24) & mask); | |
| 26 } | 27 } |
| 27 | 28 |
| 28 } // namespace | 29 } // namespace |
| 29 | 30 |
| 30 // static | 31 // static |
| 31 V4DecodeResult V4RiceDecoder::ValidateInput(const int32 rice_parameter, | 32 V4DecodeResult V4RiceDecoder::ValidateInput(const int32 rice_parameter, |
| 32 const int32 num_entries, | 33 const int32 num_entries, |
| 33 const std::string& encoded_data) { | 34 const std::string& encoded_data) { |
| 34 if (num_entries < 0) { | 35 if (num_entries < 0) { |
| 35 NOTREACHED(); | 36 NOTREACHED(); |
| (...skipping 11 matching lines...) Expand all Loading... | |
| 47 | 48 |
| 48 if (encoded_data.empty()) { | 49 if (encoded_data.empty()) { |
| 49 NOTREACHED(); | 50 NOTREACHED(); |
| 50 return ENCODED_DATA_UNEXPECTED_EMPTY_FAILURE; | 51 return ENCODED_DATA_UNEXPECTED_EMPTY_FAILURE; |
| 51 } | 52 } |
| 52 | 53 |
| 53 return DECODE_SUCCESS; | 54 return DECODE_SUCCESS; |
| 54 } | 55 } |
| 55 | 56 |
| 56 // static | 57 // static |
| 57 V4DecodeResult V4RiceDecoder::DecodeIntegers(const int32 first_value, | 58 V4DecodeResult V4RiceDecoder::DecodeIntegers(const int64 first_value, |
| 58 const int32 rice_parameter, | 59 const int32 rice_parameter, |
| 59 const int32 num_entries, | 60 const int32 num_entries, |
| 60 const std::string& encoded_data, | 61 const std::string& encoded_data, |
| 61 RepeatedField<int32>* out) { | 62 RepeatedField<int32>* out) { |
| 62 DCHECK(out); | 63 DCHECK(out); |
| 63 | 64 |
| 64 V4DecodeResult result = | 65 V4DecodeResult result = |
| 65 ValidateInput(rice_parameter, num_entries, encoded_data); | 66 ValidateInput(rice_parameter, num_entries, encoded_data); |
| 66 if (result != DECODE_SUCCESS) { | 67 if (result != DECODE_SUCCESS) { |
| 67 return result; | 68 return result; |
| (...skipping 20 matching lines...) Expand all Loading... | |
| 88 return DECODED_INTEGER_OVERFLOW_FAILURE; | 89 return DECODED_INTEGER_OVERFLOW_FAILURE; |
| 89 } | 90 } |
| 90 | 91 |
| 91 out->Add(last_value.ValueOrDie()); | 92 out->Add(last_value.ValueOrDie()); |
| 92 } | 93 } |
| 93 | 94 |
| 94 return DECODE_SUCCESS; | 95 return DECODE_SUCCESS; |
| 95 } | 96 } |
| 96 | 97 |
| 97 // static | 98 // static |
| 98 V4DecodeResult V4RiceDecoder::DecodeBytes(const int32 first_value, | 99 V4DecodeResult V4RiceDecoder::DecodeBytes(const int64 first_value, |
| 99 const int32 rice_parameter, | 100 const int32 rice_parameter, |
| 100 const int32 num_entries, | 101 const int32 num_entries, |
| 101 const std::string& encoded_data, | 102 const std::string& encoded_data, |
| 102 std::string* out) { | 103 std::vector<uint32_t>* out) { |
| 103 DCHECK(out); | 104 DCHECK(out); |
| 104 | 105 |
| 105 V4DecodeResult result = | 106 V4DecodeResult result = |
| 106 ValidateInput(rice_parameter, num_entries, encoded_data); | 107 ValidateInput(rice_parameter, num_entries, encoded_data); |
| 107 if (result != DECODE_SUCCESS) { | 108 if (result != DECODE_SUCCESS) { |
| 108 return result; | 109 return result; |
| 109 } | 110 } |
| 110 out->reserve((num_entries + 1) * 4); | 111 out->reserve((num_entries + 1)); |
| 111 | 112 |
| 112 // Cast to unsigned since we don't look at the sign bit as a sign bit. | 113 base::CheckedNumeric<uint32_t> last_value(first_value); |
| 113 // first_value should have been an unsigned to begin with but proto don't | 114 out->push_back(FlipEndianness(last_value.ValueOrDie())); |
| 114 // allow that. | |
| 115 uint32_t first_value_unsigned = static_cast<uint32_t>(first_value); | |
| 116 base::CheckedNumeric<uint32_t> last_value(first_value_unsigned); | |
| 117 char bytes[4]; | |
| 118 GetBytesFromUInt32(last_value.ValueOrDie(), bytes); | |
| 119 out->append(bytes, 4); | |
| 120 | 115 |
| 121 if (num_entries == 0) { | 116 if (num_entries > 0) { |
| 122 return DECODE_SUCCESS; | 117 V4RiceDecoder decoder(rice_parameter, num_entries, encoded_data); |
| 118 while (decoder.HasAnotherValue()) { | |
| 119 uint32_t offset; | |
| 120 result = decoder.GetNextValue(&offset); | |
| 121 if (result != DECODE_SUCCESS) { | |
| 122 return result; | |
| 123 } | |
| 124 | |
| 125 last_value += offset; | |
| 126 if (!last_value.IsValid()) { | |
| 127 NOTREACHED(); | |
| 128 return DECODED_INTEGER_OVERFLOW_FAILURE; | |
| 129 } | |
| 130 | |
| 131 // This flipping is mandated by the protocol. For details, see: | |
| 132 // https://developers.google.com/safe-browsing/v4/compression | |
|
Scott Hess - ex-Googler
2016/08/15 21:18:15
The protocol says the data is in network byte orde
vakh (use Gerrit instead)
2016/08/16 00:23:03
My machine is little-endian and it does require by
| |
| 133 out->push_back(FlipEndianness(last_value.ValueOrDie())); | |
| 134 } | |
| 123 } | 135 } |
| 136 std::sort(out->begin(), out->end()); | |
|
Scott Hess - ex-Googler
2016/08/15 21:18:15
Whoa! I could see ASSERT_TRUE(std::is_sorted(...)
vakh (use Gerrit instead)
2016/08/16 00:23:03
These aren't already sorted. That's because of the
| |
| 124 | 137 |
| 125 V4RiceDecoder decoder(rice_parameter, num_entries, encoded_data); | 138 // This flipping is done so that when the vector is interpreted as a string, |
| 126 while (decoder.HasAnotherValue()) { | 139 // the bytes are in the correct order. |
| 127 uint32_t offset; | 140 for (unsigned i = 0; i < out->size(); i++) { |
| 128 result = decoder.GetNextValue(&offset); | 141 (*out)[i] = FlipEndianness((*out)[i]); |
| 129 if (result != DECODE_SUCCESS) { | |
| 130 return result; | |
| 131 } | |
| 132 | |
| 133 last_value += offset; | |
| 134 if (!last_value.IsValid()) { | |
| 135 NOTREACHED(); | |
| 136 return DECODED_INTEGER_OVERFLOW_FAILURE; | |
| 137 } | |
| 138 | |
| 139 GetBytesFromUInt32(last_value.ValueOrDie(), bytes); | |
| 140 out->append(bytes, 4); | |
| 141 } | 142 } |
| 142 | 143 |
| 143 return DECODE_SUCCESS; | 144 return DECODE_SUCCESS; |
| 144 } | 145 } |
| 145 | 146 |
| 146 V4RiceDecoder::V4RiceDecoder(const int rice_parameter, | 147 V4RiceDecoder::V4RiceDecoder(const int rice_parameter, |
| 147 const int num_entries, | 148 const int num_entries, |
| 148 const std::string& encoded_data) | 149 const std::string& encoded_data) |
| 149 : rice_parameter_(rice_parameter), | 150 : rice_parameter_(rice_parameter), |
| 150 num_entries_(num_entries), | 151 num_entries_(num_entries), |
| (...skipping 80 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 231 return result; | 232 return result; |
| 232 } | 233 } |
| 233 } | 234 } |
| 234 | 235 |
| 235 unsigned int num_bits_left_in_current_word = | 236 unsigned int num_bits_left_in_current_word = |
| 236 kMaxBitIndex - current_word_bit_index_; | 237 kMaxBitIndex - current_word_bit_index_; |
| 237 if (num_bits_left_in_current_word >= num_requested_bits) { | 238 if (num_bits_left_in_current_word >= num_requested_bits) { |
| 238 // All the bits that we need are in |current_word_|. | 239 // All the bits that we need are in |current_word_|. |
| 239 GetBitsFromCurrentWord(num_requested_bits, x); | 240 GetBitsFromCurrentWord(num_requested_bits, x); |
| 240 } else { | 241 } else { |
| 241 // |current_word_| contains fewer bits than we need so we store the current | 242 // |current_word_| contains fewer bits than we need so read the remaining |
| 242 // value of |current_word_| in |lower|, then read in a new |current_word_|, | 243 // bits from |current_word_| into |lower|, and then call GetNextBits on the |
| 243 // and then pick the remaining bits from the new value of |current_word_|. | 244 // remaining number of bits, which will read in a new word into |
| 244 uint32_t lower = current_word_; | 245 // |current_word_|. |
| 245 // Bits we need after we read all of |current_word_| | 246 uint32_t lower; |
| 247 V4DecodeResult result = GetNextBits(num_bits_left_in_current_word, &lower); | |
| 248 if (result != DECODE_SUCCESS) { | |
| 249 return result; | |
| 250 } | |
| 251 | |
| 246 unsigned int num_bits_from_next_word = | 252 unsigned int num_bits_from_next_word = |
| 247 num_requested_bits - num_bits_left_in_current_word; | 253 num_requested_bits - num_bits_left_in_current_word; |
| 248 uint32_t upper; | 254 uint32_t upper; |
| 249 V4DecodeResult result = GetNextWord(¤t_word_); | 255 result = GetNextBits(num_bits_from_next_word, &upper); |
| 250 if (result != DECODE_SUCCESS) { | 256 if (result != DECODE_SUCCESS) { |
| 251 return result; | 257 return result; |
| 252 } | 258 } |
| 253 GetBitsFromCurrentWord(num_bits_from_next_word, &upper); | 259 *x = (upper << num_bits_left_in_current_word) | lower; |
| 254 *x = (upper << (num_requested_bits - num_bits_from_next_word)) | lower; | |
| 255 } | 260 } |
| 256 return DECODE_SUCCESS; | 261 return DECODE_SUCCESS; |
| 257 } | 262 } |
| 258 | 263 |
| 259 void V4RiceDecoder::GetBitsFromCurrentWord(unsigned int num_requested_bits, | 264 void V4RiceDecoder::GetBitsFromCurrentWord(unsigned int num_requested_bits, |
| 260 uint32_t* x) { | 265 uint32_t* x) { |
| 261 uint32_t mask = 0xFFFFFFFF >> (kMaxBitIndex - num_requested_bits); | 266 uint32_t mask = 0xFFFFFFFF >> (kMaxBitIndex - num_requested_bits); |
| 262 *x = current_word_ & mask; | 267 *x = current_word_ & mask; |
| 263 current_word_ = current_word_ >> num_requested_bits; | 268 current_word_ = current_word_ >> num_requested_bits; |
| 264 current_word_bit_index_ += num_requested_bits; | 269 current_word_bit_index_ += num_requested_bits; |
| (...skipping 11 matching lines...) Expand all Loading... | |
| 276 bits_read, current_word_, data_byte_index_, current_word_bit_index_, | 281 bits_read, current_word_, data_byte_index_, current_word_bit_index_, |
| 277 rice_parameter_); | 282 rice_parameter_); |
| 278 } | 283 } |
| 279 | 284 |
| 280 std::ostream& operator<<(std::ostream& os, const V4RiceDecoder& rice_decoder) { | 285 std::ostream& operator<<(std::ostream& os, const V4RiceDecoder& rice_decoder) { |
| 281 os << rice_decoder.DebugString(); | 286 os << rice_decoder.DebugString(); |
| 282 return os; | 287 return os; |
| 283 } | 288 } |
| 284 | 289 |
| 285 } // namespace safe_browsing | 290 } // namespace safe_browsing |
| OLD | NEW |