1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507
|
// Copyright 2017 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "components/cbor/reader.h"
#include <math.h>
#include <limits>
#include <map>
#include <utility>
#include "base/bit_cast.h"
#include "base/check_op.h"
#include "base/notreached.h"
#include "base/numerics/checked_math.h"
#include "base/numerics/safe_conversions.h"
#include "base/strings/string_util.h"
#include "components/cbor/constants.h"
#include "components/cbor/float_conversions.h"
namespace cbor {
namespace constants {
const char kUnsupportedMajorType[] = "Unsupported major type.";
}
namespace {
Value::Type GetMajorType(uint8_t initial_data_byte) {
return static_cast<Value::Type>(
(initial_data_byte & constants::kMajorTypeMask) >>
constants::kMajorTypeBitShift);
}
uint8_t GetAdditionalInfo(uint8_t initial_data_byte) {
return initial_data_byte & constants::kAdditionalInformationMask;
}
// Error messages that correspond to each of the error codes. There is 1
// exception: we declare |kUnsupportedMajorType| in constants.h in the
// `constants` namespace, because we use it in several files.
const char kNoError[] = "Successfully deserialized to a CBOR value.";
const char kUnknownAdditionalInfo[] =
"Unknown additional info format in the first byte.";
const char kIncompleteCBORData[] =
"Prematurely terminated CBOR data byte array.";
const char kIncorrectMapKeyType[] =
"Specified map key type is not supported by the current implementation.";
const char kTooMuchNesting[] = "Too much nesting.";
const char kInvalidUTF8[] =
"String encodings other than UTF-8 are not allowed.";
const char kExtraneousData[] = "Trailing data bytes are not allowed.";
const char kMapKeyOutOfOrder[] =
"Map keys must be strictly monotonically increasing based on byte length "
"and then by byte-wise lexical order.";
const char kNonMinimalCBOREncoding[] =
"Unsigned integers must be encoded with minimum number of bytes.";
const char kUnsupportedSimpleValue[] =
"Unsupported or unassigned simple value.";
const char kUnsupportedFloatingPointValue[] =
"Floating point numbers are not supported unless the "
"`allow_floating_point` configuration option is set.";
const char kOutOfRangeIntegerValue[] =
"Integer values must be between INT64_MIN and INT64_MAX.";
const char kMapKeyDuplicate[] = "Duplicate map keys are not allowed.";
const char kUnknownError[] = "An unknown error occured.";
} // namespace
Reader::Config::Config() = default;
Reader::Config::~Config() = default;
Reader::Reader(base::span<const uint8_t> data)
: rest_(data), error_code_(DecoderError::CBOR_NO_ERROR) {}
Reader::~Reader() = default;
// static
std::optional<Value> Reader::Read(base::span<uint8_t const> data,
DecoderError* error_code_out,
int max_nesting_level) {
Config config;
config.error_code_out = error_code_out;
config.max_nesting_level = max_nesting_level;
return Read(data, config);
}
// static
std::optional<Value> Reader::Read(base::span<uint8_t const> data,
size_t* num_bytes_consumed,
DecoderError* error_code_out,
int max_nesting_level) {
DCHECK(num_bytes_consumed);
Config config;
config.num_bytes_consumed = num_bytes_consumed;
config.error_code_out = error_code_out;
config.max_nesting_level = max_nesting_level;
return Read(data, config);
}
// static
std::optional<Value> Reader::Read(base::span<uint8_t const> data,
const Config& config) {
Reader reader(data);
std::optional<Value> value =
reader.DecodeCompleteDataItem(config, config.max_nesting_level);
auto error = reader.GetErrorCode();
const bool success = value.has_value();
DCHECK_EQ(success, error == DecoderError::CBOR_NO_ERROR);
if (config.num_bytes_consumed) {
*config.num_bytes_consumed =
success ? data.size() - reader.num_bytes_remaining() : 0;
} else if (success && reader.num_bytes_remaining() > 0) {
error = DecoderError::EXTRANEOUS_DATA;
value.reset();
}
if (config.error_code_out) {
*config.error_code_out = error;
}
return value;
}
std::optional<Value> Reader::DecodeCompleteDataItem(const Config& config,
int max_nesting_level) {
if (max_nesting_level < 0 || max_nesting_level > kCBORMaxDepth) {
error_code_ = DecoderError::TOO_MUCH_NESTING;
return std::nullopt;
}
std::optional<DataItemHeader> header = DecodeDataItemHeader();
if (!header.has_value()) {
return std::nullopt;
}
switch (header->type) {
case Value::Type::UNSIGNED:
return DecodeValueToUnsigned(header->value);
case Value::Type::NEGATIVE:
return DecodeValueToNegative(header->value);
case Value::Type::BYTE_STRING:
return ReadByteStringContent(*header);
case Value::Type::STRING:
return ReadStringContent(*header, config);
case Value::Type::ARRAY:
return ReadArrayContent(*header, config, max_nesting_level);
case Value::Type::MAP:
return ReadMapContent(*header, config, max_nesting_level);
case Value::Type::SIMPLE_VALUE:
case Value::Type::FLOAT_VALUE:
// Floating point values also go here since they are also type 7.
return DecodeToSimpleValueOrFloat(*header, config);
case Value::Type::TAG: // We explicitly don't support TAG.
case Value::Type::NONE:
case Value::Type::INVALID_UTF8:
break;
}
error_code_ = DecoderError::UNSUPPORTED_MAJOR_TYPE;
return std::nullopt;
}
std::optional<Reader::DataItemHeader> Reader::DecodeDataItemHeader() {
const std::optional<uint8_t> initial_byte = ReadByte();
if (!initial_byte) {
return std::nullopt;
}
const auto major_type = GetMajorType(initial_byte.value());
const uint8_t additional_info = GetAdditionalInfo(initial_byte.value());
std::optional<uint64_t> value =
ReadVariadicLengthInteger(major_type, additional_info);
return value ? std::make_optional(
DataItemHeader{major_type, additional_info, value.value()})
: std::nullopt;
}
std::optional<uint64_t> Reader::ReadVariadicLengthInteger(
Value::Type type,
uint8_t additional_info) {
uint8_t additional_bytes = 0;
if (additional_info < 24) {
return std::make_optional(additional_info);
} else if (additional_info == 24) {
additional_bytes = 1;
} else if (additional_info == 25) {
additional_bytes = 2;
} else if (additional_info == 26) {
additional_bytes = 4;
} else if (additional_info == 27) {
additional_bytes = 8;
} else {
error_code_ = DecoderError::UNKNOWN_ADDITIONAL_INFO;
return std::nullopt;
}
const std::optional<base::span<const uint8_t>> bytes =
ReadBytes(additional_bytes);
if (!bytes) {
return std::nullopt;
}
uint64_t int_data = 0;
for (const uint8_t b : bytes.value()) {
int_data <<= 8;
int_data |= b;
}
if (type == Value::Type::SIMPLE_VALUE && additional_info >= 25 &&
additional_info <= 27) {
// This is a floating point value and so `additional_bytes` should not be
// treated as an integer by minimality checking.
return std::make_optional(int_data);
}
return IsEncodingMinimal(additional_bytes, int_data)
? std::make_optional(int_data)
: std::nullopt;
}
std::optional<Value> Reader::DecodeValueToNegative(uint64_t value) {
auto negative_value = -base::CheckedNumeric<int64_t>(value) - 1;
if (!negative_value.IsValid()) {
error_code_ = DecoderError::OUT_OF_RANGE_INTEGER_VALUE;
return std::nullopt;
}
return Value(static_cast<int64_t>(negative_value.ValueOrDie()));
}
std::optional<Value> Reader::DecodeValueToUnsigned(uint64_t value) {
auto unsigned_value = base::CheckedNumeric<int64_t>(value);
if (!unsigned_value.IsValid()) {
error_code_ = DecoderError::OUT_OF_RANGE_INTEGER_VALUE;
return std::nullopt;
}
return Value(static_cast<int64_t>(unsigned_value.ValueOrDie()));
}
std::optional<Value> Reader::DecodeToSimpleValueOrFloat(
const DataItemHeader& header,
const Config& config) {
// ReadVariadicLengthInteger provides this bound.
CHECK_LE(header.additional_info, 27);
// Floating point numbers.
if (header.additional_info > 24) {
if (header.additional_info >= 28) {
error_code_ = DecoderError::UNSUPPORTED_SIMPLE_VALUE;
return std::nullopt;
}
if (!config.allow_floating_point) {
error_code_ = DecoderError::UNSUPPORTED_FLOATING_POINT_VALUE;
return std::nullopt;
}
switch (header.additional_info) {
case 25:
return Value(DecodeHalfPrecisionFloat(header.value));
case 26: {
double result =
base::bit_cast<float>(static_cast<uint32_t>(header.value));
if (!std::isfinite(result) ||
result ==
DecodeHalfPrecisionFloat(EncodeHalfPrecisionFloat(result))) {
// This could have been encoded as a 16 bit float.
// Note that we use `isfinite()` here to handle NaN since infinity
// and NaN can both be encoded in 16 bits but NaN doesn't compare
// with equality.
error_code_ = DecoderError::NON_MINIMAL_CBOR_ENCODING;
return std::nullopt;
}
return Value(result);
}
case 27: {
double result = base::bit_cast<double>(header.value);
float result_32 = result;
if (result == result_32) {
// This could have been encoded as a 32 bit float.
error_code_ = DecoderError::NON_MINIMAL_CBOR_ENCODING;
return std::nullopt;
}
return Value(result);
}
default:
NOTREACHED();
}
}
// Since |header.additional_info| <= 24, ReadVariadicLengthInteger also
// provides this bound for |header.value|.
CHECK_LE(header.value, 255u);
// |SimpleValue| is an enum class and so the underlying type is specified to
// be |int|. So this cast is safe.
Value::SimpleValue possibly_unsupported_simple_value =
static_cast<Value::SimpleValue>(static_cast<int>(header.value));
switch (possibly_unsupported_simple_value) {
case Value::SimpleValue::FALSE_VALUE:
case Value::SimpleValue::TRUE_VALUE:
case Value::SimpleValue::NULL_VALUE:
case Value::SimpleValue::UNDEFINED:
return Value(possibly_unsupported_simple_value);
}
error_code_ = DecoderError::UNSUPPORTED_SIMPLE_VALUE;
return std::nullopt;
}
std::optional<Value> Reader::ReadStringContent(
const Reader::DataItemHeader& header,
const Config& config) {
uint64_t num_bytes = header.value;
const std::optional<base::span<const uint8_t>> bytes = ReadBytes(num_bytes);
if (!bytes) {
return std::nullopt;
}
std::string cbor_string(bytes->begin(), bytes->end());
if (base::IsStringUTF8(cbor_string)) {
return Value(std::move(cbor_string));
}
if (config.allow_invalid_utf8) {
return Value(*bytes, Value::Type::INVALID_UTF8);
}
error_code_ = DecoderError::INVALID_UTF8;
return std::nullopt;
}
std::optional<Value> Reader::ReadByteStringContent(
const Reader::DataItemHeader& header) {
uint64_t num_bytes = header.value;
const std::optional<base::span<const uint8_t>> bytes = ReadBytes(num_bytes);
if (!bytes) {
return std::nullopt;
}
std::vector<uint8_t> cbor_byte_string(bytes->begin(), bytes->end());
return Value(std::move(cbor_byte_string));
}
std::optional<Value> Reader::ReadArrayContent(
const Reader::DataItemHeader& header,
const Config& config,
int max_nesting_level) {
const uint64_t length = header.value;
Value::ArrayValue cbor_array;
for (uint64_t i = 0; i < length; ++i) {
std::optional<Value> cbor_element =
DecodeCompleteDataItem(config, max_nesting_level - 1);
if (!cbor_element.has_value()) {
return std::nullopt;
}
cbor_array.push_back(std::move(cbor_element.value()));
}
return Value(std::move(cbor_array));
}
std::optional<Value> Reader::ReadMapContent(
const Reader::DataItemHeader& header,
const Config& config,
int max_nesting_level) {
const uint64_t length = header.value;
std::map<Value, Value, Value::Less> cbor_map;
for (uint64_t i = 0; i < length; ++i) {
std::optional<Value> key =
DecodeCompleteDataItem(config, max_nesting_level - 1);
std::optional<Value> value =
DecodeCompleteDataItem(config, max_nesting_level - 1);
if (!key.has_value() || !value.has_value()) {
return std::nullopt;
}
switch (key.value().type()) {
case Value::Type::UNSIGNED:
case Value::Type::NEGATIVE:
case Value::Type::STRING:
case Value::Type::BYTE_STRING:
break;
case Value::Type::INVALID_UTF8:
error_code_ = DecoderError::INVALID_UTF8;
return std::nullopt;
default:
error_code_ = DecoderError::INCORRECT_MAP_KEY_TYPE;
return std::nullopt;
}
if (IsDuplicateKey(key.value(), cbor_map))
return std::nullopt;
if (!config.allow_and_canonicalize_out_of_order_keys &&
!IsKeyInOrder(key.value(), cbor_map)) {
return std::nullopt;
}
cbor_map.emplace(std::move(key.value()), std::move(value.value()));
}
Value::MapValue map;
map.reserve(cbor_map.size());
// TODO(crbug.com/40205788): when Chromium switches to C++17, this code can be
// optimized using std::map::extract().
for (auto& it : cbor_map)
map.emplace_hint(map.end(), it.first.Clone(), std::move(it.second));
return Value(std::move(map));
}
std::optional<uint8_t> Reader::ReadByte() {
const std::optional<base::span<const uint8_t>> bytes = ReadBytes(1);
return bytes ? std::make_optional(bytes.value()[0]) : std::nullopt;
}
std::optional<base::span<const uint8_t>> Reader::ReadBytes(uint64_t num_bytes) {
if (base::strict_cast<uint64_t>(rest_.size()) < num_bytes) {
error_code_ = DecoderError::INCOMPLETE_CBOR_DATA;
return std::nullopt;
}
// The `uint64_t` => `size_t` conversion below will always succeed
// because the `if` condition above implies that `num_bytes` fits into a
// `size_t`.
size_t size = base::checked_cast<size_t>(num_bytes);
const base::span<const uint8_t> ret = rest_.first(size);
rest_ = rest_.subspan(size);
return ret;
}
bool Reader::IsEncodingMinimal(uint8_t additional_bytes, uint64_t uint_data) {
if ((additional_bytes == 1 && uint_data < 24) ||
uint_data <= (1ULL << 8 * (additional_bytes >> 1)) - 1) {
error_code_ = DecoderError::NON_MINIMAL_CBOR_ENCODING;
return false;
}
return true;
}
bool Reader::IsKeyInOrder(const Value& new_key,
const std::map<Value, Value, Value::Less>& map) {
if (map.empty()) {
return true;
}
const auto& max_current_key = map.rbegin()->first;
const auto less = map.key_comp();
if (!less(max_current_key, new_key)) {
error_code_ = DecoderError::OUT_OF_ORDER_KEY;
return false;
}
return true;
}
bool Reader::IsDuplicateKey(const Value& new_key,
const std::map<Value, Value, Value::Less>& map) {
if (map.find(new_key) == map.end()) {
return false;
}
error_code_ = DecoderError::DUPLICATE_KEY;
return true;
}
// static
const char* Reader::ErrorCodeToString(DecoderError error) {
switch (error) {
case DecoderError::CBOR_NO_ERROR:
return kNoError;
case DecoderError::UNSUPPORTED_MAJOR_TYPE:
return constants::kUnsupportedMajorType;
case DecoderError::UNKNOWN_ADDITIONAL_INFO:
return kUnknownAdditionalInfo;
case DecoderError::INCOMPLETE_CBOR_DATA:
return kIncompleteCBORData;
case DecoderError::INCORRECT_MAP_KEY_TYPE:
return kIncorrectMapKeyType;
case DecoderError::TOO_MUCH_NESTING:
return kTooMuchNesting;
case DecoderError::INVALID_UTF8:
return kInvalidUTF8;
case DecoderError::EXTRANEOUS_DATA:
return kExtraneousData;
case DecoderError::OUT_OF_ORDER_KEY:
return kMapKeyOutOfOrder;
case DecoderError::NON_MINIMAL_CBOR_ENCODING:
return kNonMinimalCBOREncoding;
case DecoderError::UNSUPPORTED_SIMPLE_VALUE:
return kUnsupportedSimpleValue;
case DecoderError::UNSUPPORTED_FLOATING_POINT_VALUE:
return kUnsupportedFloatingPointValue;
case DecoderError::OUT_OF_RANGE_INTEGER_VALUE:
return kOutOfRangeIntegerValue;
case DecoderError::DUPLICATE_KEY:
return kMapKeyDuplicate;
case DecoderError::UNKNOWN_ERROR:
return kUnknownError;
default:
NOTREACHED();
}
}
} // namespace cbor
|