DXR is a code search and navigation tool aimed at making sense of large projects. It supports full-text and regex searches as well as structural queries.

Header

Mercurial (409f3966645a)

VCS Links

Line Code
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146
//* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */

#include "ProtocolParser.h"
#include "LookupCache.h"
#include "nsNetCID.h"
#include "mozilla/Logging.h"
#include "prnetdb.h"
#include "prprf.h"

#include "nsUrlClassifierDBService.h"
#include "nsUrlClassifierUtils.h"
#include "nsPrintfCString.h"
#include "mozilla/Base64.h"
#include "RiceDeltaDecoder.h"
#include "mozilla/EndianUtils.h"
#include "mozilla/ErrorNames.h"
#include "mozilla/IntegerPrintfMacros.h"

// MOZ_LOG=UrlClassifierProtocolParser:5
mozilla::LazyLogModule gUrlClassifierProtocolParserLog("UrlClassifierProtocolParser");
#define PARSER_LOG(args) MOZ_LOG(gUrlClassifierProtocolParserLog, mozilla::LogLevel::Debug, args)

namespace mozilla {
namespace safebrowsing {

// Updates will fail if fed chunks larger than this
const uint32_t MAX_CHUNK_SIZE = (1024 * 1024);
// Updates will fail if the total number of tocuhed chunks is larger than this
const uint32_t MAX_CHUNK_RANGE = 1000000;

const uint32_t DOMAIN_SIZE = 4;

// Parse one stringified range of chunks of the form "n" or "n-m" from a
// comma-separated list of chunks.  Upon return, 'begin' will point to the
// next range of chunks in the list of chunks.
static bool
ParseChunkRange(nsACString::const_iterator& aBegin,
                const nsACString::const_iterator& aEnd,
                uint32_t* aFirst, uint32_t* aLast)
{
  nsACString::const_iterator iter = aBegin;
  FindCharInReadable(',', iter, aEnd);

  nsAutoCString element(Substring(aBegin, iter));
  aBegin = iter;
  if (aBegin != aEnd)
    aBegin++;

  uint32_t numRead = PR_sscanf(element.get(), "%u-%u", aFirst, aLast);
  if (numRead == 2) {
    if (*aFirst > *aLast) {
      uint32_t tmp = *aFirst;
      *aFirst = *aLast;
      *aLast = tmp;
    }
    return true;
  }

  if (numRead == 1) {
    *aLast = *aFirst;
    return true;
  }

  return false;
}

///////////////////////////////////////////////////////////////
// ProtocolParser implementation

ProtocolParser::ProtocolParser()
  : mUpdateStatus(NS_OK)
  , mUpdateWaitSec(0)
{
}

ProtocolParser::~ProtocolParser()
{
}

nsresult
ProtocolParser::Begin(const nsACString& aTable,
                      const nsTArray<nsCString>& aUpdateTables)
{
  // ProtocolParser objects should never be reused.
  MOZ_ASSERT(mPending.IsEmpty());
  MOZ_ASSERT(mTableUpdates.IsEmpty());
  MOZ_ASSERT(mForwards.IsEmpty());
  MOZ_ASSERT(mRequestedTables.IsEmpty());
  MOZ_ASSERT(mTablesToReset.IsEmpty());

  if (!aTable.IsEmpty()) {
    SetCurrentTable(aTable);
  }
  SetRequestedTables(aUpdateTables);

  return NS_OK;
}

RefPtr<TableUpdate>
ProtocolParser::GetTableUpdate(const nsACString& aTable)
{
  for (uint32_t i = 0; i < mTableUpdates.Length(); i++) {
    if (aTable.Equals(mTableUpdates[i]->TableName())) {
      return mTableUpdates[i];
    }
  }

  // We free automatically on destruction, ownership of these
  // updates can be transferred to DBServiceWorker, which passes
  // them back to Classifier when doing the updates, and that
  // will free them.
  RefPtr<TableUpdate> update = CreateTableUpdate(aTable);
  mTableUpdates.AppendElement(update);
  return update;
}

///////////////////////////////////////////////////////////////////////
// ProtocolParserV2

ProtocolParserV2::ProtocolParserV2()
  : mState(PROTOCOL_STATE_CONTROL)
  , mTableUpdate(nullptr)
{
}

ProtocolParserV2::~ProtocolParserV2()
{
}

void
ProtocolParserV2::SetCurrentTable(const nsACString& aTable)
{
  RefPtr<TableUpdate> update = GetTableUpdate(aTable);
  mTableUpdate = TableUpdate::Cast<TableUpdateV2>(update);
}

nsresult
ProtocolParserV2::AppendStream(const nsACString& aData)
{
  if (NS_FAILED(mUpdateStatus))
    return mUpdateStatus;

  nsresult rv;
  mPending.Append(aData);
#ifdef MOZ_SAFEBROWSING_DUMP_FAILED_UPDATES
  mRawUpdate.Append(aData);
#endif

  bool done = false;
  while (!done) {
    if (nsUrlClassifierDBService::ShutdownHasStarted()) {
      return NS_ERROR_ABORT;
    }

    if (mState == PROTOCOL_STATE_CONTROL) {
      rv = ProcessControl(&done);
    } else if (mState == PROTOCOL_STATE_CHUNK) {
      rv = ProcessChunk(&done);
    } else {
      NS_ERROR("Unexpected protocol state");
      rv = NS_ERROR_FAILURE;
    }
    if (NS_FAILED(rv)) {
      mUpdateStatus = rv;
      return rv;
    }
  }
  return NS_OK;
}

void
ProtocolParserV2::End()
{
  // Inbound data has already been processed in every AppendStream() call.
  mTableUpdate = nullptr;
}

nsresult
ProtocolParserV2::ProcessControl(bool* aDone)
{
  nsresult rv;

  nsAutoCString line;
  *aDone = true;
  while (NextLine(line)) {
    PARSER_LOG(("Processing %s\n", line.get()));

    if (StringBeginsWith(line, NS_LITERAL_CSTRING("i:"))) {
      // Set the table name from the table header line.
      SetCurrentTable(Substring(line, 2));
    } else if (StringBeginsWith(line, NS_LITERAL_CSTRING("n:"))) {
      if (PR_sscanf(line.get(), "n:%d", &mUpdateWaitSec) != 1) {
        PARSER_LOG(("Error parsing n: '%s' (%d)", line.get(), mUpdateWaitSec));
        return NS_ERROR_FAILURE;
      }
    } else if (line.EqualsLiteral("r:pleasereset")) {
      PARSER_LOG(("All tables will be reset."));
      mTablesToReset = mRequestedTables;
    } else if (StringBeginsWith(line, NS_LITERAL_CSTRING("u:"))) {
      rv = ProcessForward(line);
      NS_ENSURE_SUCCESS(rv, rv);
    } else if (StringBeginsWith(line, NS_LITERAL_CSTRING("a:")) ||
               StringBeginsWith(line, NS_LITERAL_CSTRING("s:"))) {
      rv = ProcessChunkControl(line);
      NS_ENSURE_SUCCESS(rv, rv);
      *aDone = false;
      return NS_OK;
    } else if (StringBeginsWith(line, NS_LITERAL_CSTRING("ad:")) ||
               StringBeginsWith(line, NS_LITERAL_CSTRING("sd:"))) {
      rv = ProcessExpirations(line);
      NS_ENSURE_SUCCESS(rv, rv);
    }
  }

  *aDone = true;
  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessExpirations(const nsCString& aLine)
{
  if (!mTableUpdate) {
    NS_WARNING("Got an expiration without a table.");
    return NS_ERROR_FAILURE;
  }
  const nsACString& list = Substring(aLine, 3);
  nsACString::const_iterator begin, end;
  list.BeginReading(begin);
  list.EndReading(end);
  while (begin != end) {
    uint32_t first, last;
    if (ParseChunkRange(begin, end, &first, &last)) {
      if (last < first) return NS_ERROR_FAILURE;
      if (last - first > MAX_CHUNK_RANGE) return NS_ERROR_FAILURE;
      for (uint32_t num = first; num <= last; num++) {
        if (aLine[0] == 'a') {
          nsresult rv = mTableUpdate->NewAddExpiration(num);
          if (NS_FAILED(rv)) {
            return rv;
          }
        } else {
          nsresult rv = mTableUpdate->NewSubExpiration(num);
          if (NS_FAILED(rv)) {
            return rv;
          }
        }
      }
    } else {
      return NS_ERROR_FAILURE;
    }
  }
  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessChunkControl(const nsCString& aLine)
{
  if (!mTableUpdate) {
    NS_WARNING("Got a chunk before getting a table.");
    return NS_ERROR_FAILURE;
  }

  mState = PROTOCOL_STATE_CHUNK;
  char command;

  mChunkState.Clear();

  if (PR_sscanf(aLine.get(),
                "%c:%d:%d:%d",
                &command,
                &mChunkState.num, &mChunkState.hashSize, &mChunkState.length)
      != 4)
  {
    NS_WARNING(("PR_sscanf failed"));
    return NS_ERROR_FAILURE;
  }

  if (mChunkState.length > MAX_CHUNK_SIZE) {
    NS_WARNING("Invalid length specified in update.");
    return NS_ERROR_FAILURE;
  }

  if (!(mChunkState.hashSize == PREFIX_SIZE || mChunkState.hashSize == COMPLETE_SIZE)) {
    NS_WARNING("Invalid hash size specified in update.");
    return NS_ERROR_FAILURE;
  }

  if (StringEndsWith(mTableUpdate->TableName(),
                     NS_LITERAL_CSTRING("-shavar")) ||
      StringEndsWith(mTableUpdate->TableName(),
                     NS_LITERAL_CSTRING("-simple"))) {
    // Accommodate test tables ending in -simple for now.
    mChunkState.type = (command == 'a') ? CHUNK_ADD : CHUNK_SUB;
  } else if (StringEndsWith(mTableUpdate->TableName(),
    NS_LITERAL_CSTRING("-digest256"))) {
    mChunkState.type = (command == 'a') ? CHUNK_ADD_DIGEST : CHUNK_SUB_DIGEST;
  }
  nsresult rv;
  switch (mChunkState.type) {
    case CHUNK_ADD:
      rv = mTableUpdate->NewAddChunk(mChunkState.num);
      if (NS_FAILED(rv)) {
        return rv;
      }
      break;
    case CHUNK_SUB:
      rv = mTableUpdate->NewSubChunk(mChunkState.num);
      if (NS_FAILED(rv)) {
        return rv;
      }
      break;
    case CHUNK_ADD_DIGEST:
      rv = mTableUpdate->NewAddChunk(mChunkState.num);
      if (NS_FAILED(rv)) {
        return rv;
      }
      break;
    case CHUNK_SUB_DIGEST:
      rv = mTableUpdate->NewSubChunk(mChunkState.num);
      if (NS_FAILED(rv)) {
        return rv;
      }
      break;
  }

  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessForward(const nsCString& aLine)
{
  const nsACString& forward = Substring(aLine, 2);
  return AddForward(forward);
}

nsresult
ProtocolParserV2::AddForward(const nsACString& aUrl)
{
  if (!mTableUpdate) {
    NS_WARNING("Forward without a table name.");
    return NS_ERROR_FAILURE;
  }

  ForwardedUpdate *forward = mForwards.AppendElement();
  forward->table = mTableUpdate->TableName();
  forward->url.Assign(aUrl);

  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessChunk(bool* aDone)
{
  if (!mTableUpdate) {
    NS_WARNING("Processing chunk without an active table.");
    return NS_ERROR_FAILURE;
  }

  NS_ASSERTION(mChunkState.num != 0, "Must have a chunk number.");

  if (mPending.Length() < mChunkState.length) {
    *aDone = true;
    return NS_OK;
  }

  // Pull the chunk out of the pending stream data.
  nsAutoCString chunk;
  chunk.Assign(Substring(mPending, 0, mChunkState.length));
  mPending.Cut(0, mChunkState.length);

  *aDone = false;
  mState = PROTOCOL_STATE_CONTROL;

  if (StringEndsWith(mTableUpdate->TableName(),
                     NS_LITERAL_CSTRING("-shavar"))) {
    return ProcessShaChunk(chunk);
  }
  if (StringEndsWith(mTableUpdate->TableName(),
             NS_LITERAL_CSTRING("-digest256"))) {
    return ProcessDigestChunk(chunk);
  }
  return ProcessPlaintextChunk(chunk);
}

/**
 * Process a plaintext chunk (currently only used in unit tests).
 */
nsresult
ProtocolParserV2::ProcessPlaintextChunk(const nsACString& aChunk)
{
  if (!mTableUpdate) {
    NS_WARNING("Chunk received with no table.");
    return NS_ERROR_FAILURE;
  }

  PARSER_LOG(("Handling a %d-byte simple chunk", aChunk.Length()));

  nsTArray<nsCString> lines;
  ParseString(PromiseFlatCString(aChunk), '\n', lines);

  // non-hashed tables need to be hashed
  for (uint32_t i = 0; i < lines.Length(); i++) {
    nsCString& line = lines[i];

    if (mChunkState.type == CHUNK_ADD) {
      if (mChunkState.hashSize == COMPLETE_SIZE) {
        Completion hash;
        hash.FromPlaintext(line);
        nsresult rv = mTableUpdate->NewAddComplete(mChunkState.num, hash);
        if (NS_FAILED(rv)) {
          return rv;
        }
      } else {
        NS_ASSERTION(mChunkState.hashSize == 4, "Only 32- or 4-byte hashes can be used for add chunks.");
        Prefix hash;
        hash.FromPlaintext(line);
        nsresult rv = mTableUpdate->NewAddPrefix(mChunkState.num, hash);
        if (NS_FAILED(rv)) {
          return rv;
        }
      }
    } else {
      nsCString::const_iterator begin, iter, end;
      line.BeginReading(begin);
      line.EndReading(end);
      iter = begin;
      uint32_t addChunk;
      if (!FindCharInReadable(':', iter, end) ||
          PR_sscanf(lines[i].get(), "%d:", &addChunk) != 1) {
        NS_WARNING("Received sub chunk without associated add chunk.");
        return NS_ERROR_FAILURE;
      }
      iter++;

      if (mChunkState.hashSize == COMPLETE_SIZE) {
        Completion hash;
        hash.FromPlaintext(Substring(iter, end));
        nsresult rv = mTableUpdate->NewSubComplete(addChunk, hash, mChunkState.num);
        if (NS_FAILED(rv)) {
          return rv;
        }
      } else {
        NS_ASSERTION(mChunkState.hashSize == 4, "Only 32- or 4-byte hashes can be used for add chunks.");
        Prefix hash;
        hash.FromPlaintext(Substring(iter, end));
        nsresult rv = mTableUpdate->NewSubPrefix(addChunk, hash, mChunkState.num);
        if (NS_FAILED(rv)) {
          return rv;
        }
      }
    }
  }

  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessShaChunk(const nsACString& aChunk)
{
  uint32_t start = 0;
  while (start < aChunk.Length()) {
    // First four bytes are the domain key.
    Prefix domain;
    domain.Assign(Substring(aChunk, start, DOMAIN_SIZE));
    start += DOMAIN_SIZE;

    // Then a count of entries.
    uint8_t numEntries = static_cast<uint8_t>(aChunk[start]);
    start++;

    PARSER_LOG(("Handling a %d-byte shavar chunk containing %u entries"
                " for domain %X", aChunk.Length(), numEntries,
                domain.ToUint32()));

    nsresult rv;
    if (mChunkState.type == CHUNK_ADD && mChunkState.hashSize == PREFIX_SIZE) {
      rv = ProcessHostAdd(domain, numEntries, aChunk, &start);
    } else if (mChunkState.type == CHUNK_ADD && mChunkState.hashSize == COMPLETE_SIZE) {
      rv = ProcessHostAddComplete(numEntries, aChunk, &start);
    } else if (mChunkState.type == CHUNK_SUB && mChunkState.hashSize == PREFIX_SIZE) {
      rv = ProcessHostSub(domain, numEntries, aChunk, &start);
    } else if (mChunkState.type == CHUNK_SUB && mChunkState.hashSize == COMPLETE_SIZE) {
      rv = ProcessHostSubComplete(numEntries, aChunk, &start);
    } else {
      NS_WARNING("Unexpected chunk type/hash size!");
      PARSER_LOG(("Got an unexpected chunk type/hash size: %s:%d",
           mChunkState.type == CHUNK_ADD ? "add" : "sub",
           mChunkState.hashSize));
      return NS_ERROR_FAILURE;
    }
    NS_ENSURE_SUCCESS(rv, rv);
  }

  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessDigestChunk(const nsACString& aChunk)
{
  PARSER_LOG(("Handling a %d-byte digest256 chunk", aChunk.Length()));

  if (mChunkState.type == CHUNK_ADD_DIGEST) {
    return ProcessDigestAdd(aChunk);
  }
  if (mChunkState.type == CHUNK_SUB_DIGEST) {
    return ProcessDigestSub(aChunk);
  }
  return NS_ERROR_UNEXPECTED;
}

nsresult
ProtocolParserV2::ProcessDigestAdd(const nsACString& aChunk)
{
  MOZ_ASSERT(mTableUpdate);
  // The ABNF format for add chunks is (HASH)+, where HASH is 32 bytes.
  MOZ_ASSERT(aChunk.Length() % 32 == 0,
             "Chunk length in bytes must be divisible by 4");
  uint32_t start = 0;
  while (start < aChunk.Length()) {
    Completion hash;
    hash.Assign(Substring(aChunk, start, COMPLETE_SIZE));
    start += COMPLETE_SIZE;
    nsresult rv = mTableUpdate->NewAddComplete(mChunkState.num, hash);
    if (NS_FAILED(rv)) {
      return rv;
    }
  }
  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessDigestSub(const nsACString& aChunk)
{
  MOZ_ASSERT(mTableUpdate);
  // The ABNF format for sub chunks is (ADDCHUNKNUM HASH)+, where ADDCHUNKNUM
  // is a 4 byte chunk number, and HASH is 32 bytes.
  MOZ_ASSERT(aChunk.Length() % 36 == 0,
             "Chunk length in bytes must be divisible by 36");
  uint32_t start = 0;
  while (start < aChunk.Length()) {
    // Read ADDCHUNKNUM
    const nsACString& addChunkStr = Substring(aChunk, start, 4);
    start += 4;

    uint32_t addChunk;
    memcpy(&addChunk, addChunkStr.BeginReading(), 4);
    addChunk = PR_ntohl(addChunk);

    // Read the hash
    Completion hash;
    hash.Assign(Substring(aChunk, start, COMPLETE_SIZE));
    start += COMPLETE_SIZE;

    nsresult rv = mTableUpdate->NewSubComplete(addChunk, hash, mChunkState.num);
    if (NS_FAILED(rv)) {
      return rv;
    }
  }
  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessHostAdd(const Prefix& aDomain, uint8_t aNumEntries,
                               const nsACString& aChunk, uint32_t* aStart)
{
  MOZ_ASSERT(mTableUpdate);
  NS_ASSERTION(mChunkState.hashSize == PREFIX_SIZE,
               "ProcessHostAdd should only be called for prefix hashes.");

  if (aNumEntries == 0) {
    nsresult rv = mTableUpdate->NewAddPrefix(mChunkState.num, aDomain);
    if (NS_FAILED(rv)) {
      return rv;
    }
    return NS_OK;
  }

  if (*aStart + (PREFIX_SIZE * aNumEntries) > aChunk.Length()) {
    NS_WARNING("Chunk is not long enough to contain the expected entries.");
    return NS_ERROR_FAILURE;
  }

  for (uint8_t i = 0; i < aNumEntries; i++) {
    Prefix hash;
    hash.Assign(Substring(aChunk, *aStart, PREFIX_SIZE));
    PARSER_LOG(("Add prefix %X", hash.ToUint32()));
    nsresult rv = mTableUpdate->NewAddPrefix(mChunkState.num, hash);
    if (NS_FAILED(rv)) {
      return rv;
    }
    *aStart += PREFIX_SIZE;
  }

  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessHostSub(const Prefix& aDomain, uint8_t aNumEntries,
                               const nsACString& aChunk, uint32_t *aStart)
{
  MOZ_ASSERT(mTableUpdate);
  NS_ASSERTION(mChunkState.hashSize == PREFIX_SIZE,
               "ProcessHostSub should only be called for prefix hashes.");

  if (aNumEntries == 0) {
    if ((*aStart) + 4 > aChunk.Length()) {
      NS_WARNING("Received a zero-entry sub chunk without an associated add.");
      return NS_ERROR_FAILURE;
    }

    const nsACString& addChunkStr = Substring(aChunk, *aStart, 4);
    *aStart += 4;

    uint32_t addChunk;
    memcpy(&addChunk, addChunkStr.BeginReading(), 4);
    addChunk = PR_ntohl(addChunk);

    PARSER_LOG(("Sub prefix (addchunk=%u)", addChunk));
    nsresult rv = mTableUpdate->NewSubPrefix(addChunk, aDomain, mChunkState.num);
    if (NS_FAILED(rv)) {
      return rv;
    }
    return NS_OK;
  }

  if (*aStart + ((PREFIX_SIZE + 4) * aNumEntries) > aChunk.Length()) {
    NS_WARNING("Chunk is not long enough to contain the expected entries.");
    return NS_ERROR_FAILURE;
  }

  for (uint8_t i = 0; i < aNumEntries; i++) {
    const nsACString& addChunkStr = Substring(aChunk, *aStart, 4);
    *aStart += 4;

    uint32_t addChunk;
    memcpy(&addChunk, addChunkStr.BeginReading(), 4);
    addChunk = PR_ntohl(addChunk);

    Prefix prefix;
    prefix.Assign(Substring(aChunk, *aStart, PREFIX_SIZE));
    *aStart += PREFIX_SIZE;

    PARSER_LOG(("Sub prefix %X (addchunk=%u)", prefix.ToUint32(), addChunk));
    nsresult rv = mTableUpdate->NewSubPrefix(addChunk, prefix, mChunkState.num);
    if (NS_FAILED(rv)) {
      return rv;
    }
  }

  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessHostAddComplete(uint8_t aNumEntries,
                                       const nsACString& aChunk, uint32_t* aStart)
{
  MOZ_ASSERT(mTableUpdate);
  NS_ASSERTION(mChunkState.hashSize == COMPLETE_SIZE,
               "ProcessHostAddComplete should only be called for complete hashes.");

  if (aNumEntries == 0) {
    // this is totally comprehensible.
    // My sarcasm detector is going off!
    NS_WARNING("Expected > 0 entries for a 32-byte hash add.");
    return NS_OK;
  }

  if (*aStart + (COMPLETE_SIZE * aNumEntries) > aChunk.Length()) {
    NS_WARNING("Chunk is not long enough to contain the expected entries.");
    return NS_ERROR_FAILURE;
  }

  for (uint8_t i = 0; i < aNumEntries; i++) {
    Completion hash;
    hash.Assign(Substring(aChunk, *aStart, COMPLETE_SIZE));
    nsresult rv = mTableUpdate->NewAddComplete(mChunkState.num, hash);
    if (NS_FAILED(rv)) {
      return rv;
    }
    *aStart += COMPLETE_SIZE;
  }

  return NS_OK;
}

nsresult
ProtocolParserV2::ProcessHostSubComplete(uint8_t aNumEntries,
                                         const nsACString& aChunk, uint32_t* aStart)
{
  MOZ_ASSERT(mTableUpdate);
  NS_ASSERTION(mChunkState.hashSize == COMPLETE_SIZE,
               "ProcessHostSubComplete should only be called for complete hashes.");

  if (aNumEntries == 0) {
    // this is totally comprehensible.
    NS_WARNING("Expected > 0 entries for a 32-byte hash sub.");
    return NS_OK;
  }

  if (*aStart + ((COMPLETE_SIZE + 4) * aNumEntries) > aChunk.Length()) {
    NS_WARNING("Chunk is not long enough to contain the expected entries.");
    return NS_ERROR_FAILURE;
  }

  for (uint8_t i = 0; i < aNumEntries; i++) {
    Completion hash;
    hash.Assign(Substring(aChunk, *aStart, COMPLETE_SIZE));
    *aStart += COMPLETE_SIZE;

    const nsACString& addChunkStr = Substring(aChunk, *aStart, 4);
    *aStart += 4;

    uint32_t addChunk;
    memcpy(&addChunk, addChunkStr.BeginReading(), 4);
    addChunk = PR_ntohl(addChunk);

    nsresult rv = mTableUpdate->NewSubComplete(addChunk, hash, mChunkState.num);
    if (NS_FAILED(rv)) {
      return rv;
    }
  }

  return NS_OK;
}

bool
ProtocolParserV2::NextLine(nsACString& aLine)
{
  int32_t newline = mPending.FindChar('\n');
  if (newline == kNotFound) {
    return false;
  }
  aLine.Assign(Substring(mPending, 0, newline));
  mPending.Cut(0, newline + 1);
  return true;
}

RefPtr<TableUpdate>
ProtocolParserV2::CreateTableUpdate(const nsACString& aTableName) const
{
  return new TableUpdateV2(aTableName);
}

///////////////////////////////////////////////////////////////////////
// ProtocolParserProtobuf

ProtocolParserProtobuf::ProtocolParserProtobuf()
{
}

ProtocolParserProtobuf::~ProtocolParserProtobuf()
{
}

void
ProtocolParserProtobuf::SetCurrentTable(const nsACString& aTable)
{
  // Should never occur.
  MOZ_ASSERT_UNREACHABLE("SetCurrentTable shouldn't be called");
}


RefPtr<TableUpdate>
ProtocolParserProtobuf::CreateTableUpdate(const nsACString& aTableName) const
{
  return new TableUpdateV4(aTableName);
}

nsresult
ProtocolParserProtobuf::AppendStream(const nsACString& aData)
{
  // Protobuf data cannot be parsed progressively. Just save the incoming data.
  mPending.Append(aData);
  return NS_OK;
}

void
ProtocolParserProtobuf::End()
{
  // mUpdateStatus will be updated to success as long as not all
  // the responses are invalid.
  mUpdateStatus = NS_ERROR_FAILURE;

  FetchThreatListUpdatesResponse response;
  if (!response.ParseFromArray(mPending.get(), mPending.Length())) {
    NS_WARNING("ProtocolParserProtobuf failed parsing data.");
    return;
  }

  auto minWaitDuration = response.minimum_wait_duration();
  mUpdateWaitSec = minWaitDuration.seconds() +
                   minWaitDuration.nanos() / 1000000000;

  for (int i = 0; i < response.list_update_responses_size(); i++) {
    auto r = response.list_update_responses(i);
    nsAutoCString listName;
    nsresult rv = ProcessOneResponse(r, listName);
    if (NS_SUCCEEDED(rv)) {
      mUpdateStatus = rv;
    } else {
      nsAutoCString errorName;
      mozilla::GetErrorName(rv, errorName);
      NS_WARNING(nsPrintfCString("Failed to process one response for '%s': %s",
                                 listName.get(), errorName.get()).get());
      if (!listName.IsEmpty()) {
        PARSER_LOG(("Table %s will be reset.", listName.get()));
        mTablesToReset.AppendElement(listName);
      }
    }
  }
}

nsresult
ProtocolParserProtobuf::ProcessOneResponse(const ListUpdateResponse& aResponse,
                                           nsACString& aListName)
{
  MOZ_ASSERT(aListName.IsEmpty());

  // A response must have a threat type.
  if (!aResponse.has_threat_type()) {
    NS_WARNING("Threat type not initialized. This seems to be an invalid response.");
    return NS_ERROR_UC_PARSER_MISSING_PARAM;
  }

  // Convert threat type to list name.
  nsCOMPtr<nsIUrlClassifierUtils> urlUtil =
    do_GetService(NS_URLCLASSIFIERUTILS_CONTRACTID);
  nsCString possibleListNames;
  nsresult rv = urlUtil->ConvertThreatTypeToListNames(aResponse.threat_type(),
                                                      possibleListNames);
  if (NS_FAILED(rv)) {
    PARSER_LOG(("Threat type to list name conversion error: %d",
                aResponse.threat_type()));
    return NS_ERROR_UC_PARSER_UNKNOWN_THREAT;
  }

  // Match the table name we received with one of the ones we requested.
  // We ignore the case where a threat type matches more than one list
  // per provider and return the first one. See bug 1287059."
  nsTArray<nsCString> possibleListNameArray;
  Classifier::SplitTables(possibleListNames, possibleListNameArray);
  for (auto possibleName : possibleListNameArray) {
    if (mRequestedTables.Contains(possibleName)) {
      aListName = possibleName;
      break;
    }
  }

  if (aListName.IsEmpty()) {
    PARSER_LOG(("We received an update for a list we didn't ask for. Ignoring it."));
    return NS_ERROR_FAILURE;
  }

  // Test if this is a full update.
  bool isFullUpdate = false;
  if (aResponse.has_response_type()) {
    isFullUpdate =
      aResponse.response_type() == ListUpdateResponse::FULL_UPDATE;
  } else {
    NS_WARNING("Response type not initialized.");
    return NS_ERROR_UC_PARSER_MISSING_PARAM;
  }

  // Warn if there's no new state.
  if (!aResponse.has_new_client_state()) {
    NS_WARNING("New state not initialized.");
    return NS_ERROR_UC_PARSER_MISSING_PARAM;
  }

  auto tu = GetTableUpdate(aListName);
  auto tuV4 = TableUpdate::Cast<TableUpdateV4>(tu);
  NS_ENSURE_TRUE(tuV4, NS_ERROR_FAILURE);

  nsCString state(aResponse.new_client_state().c_str(),
                  aResponse.new_client_state().size());
  tuV4->SetNewClientState(state);

  if (aResponse.has_checksum()) {
    tuV4->NewChecksum(aResponse.checksum().sha256());
  }

  PARSER_LOG(("==== Update for threat type '%d' ====", aResponse.threat_type()));
  PARSER_LOG(("* aListName: %s\n", PromiseFlatCString(aListName).get()));
  PARSER_LOG(("* newState: %s\n", aResponse.new_client_state().c_str()));
  PARSER_LOG(("* isFullUpdate: %s\n", (isFullUpdate ? "yes" : "no")));
  PARSER_LOG(("* hasChecksum: %s\n", (aResponse.has_checksum() ? "yes" : "no")));
  PARSER_LOG(("* additions: %d\n", aResponse.additions().size()));
  PARSER_LOG(("* removals: %d\n", aResponse.removals().size()));

  tuV4->SetFullUpdate(isFullUpdate);

  rv = ProcessAdditionOrRemoval(*tuV4, aResponse.additions(), true /*aIsAddition*/);
  NS_ENSURE_SUCCESS(rv, rv);
  rv = ProcessAdditionOrRemoval(*tuV4, aResponse.removals(), false);
  NS_ENSURE_SUCCESS(rv, rv);

  PARSER_LOG(("\n\n"));

  return NS_OK;
}

nsresult
ProtocolParserProtobuf::ProcessAdditionOrRemoval(TableUpdateV4& aTableUpdate,
                                                 const ThreatEntrySetList& aUpdate,
                                                 bool aIsAddition)
{
  nsresult ret = NS_OK;

  for (int i = 0; i < aUpdate.size(); i++) {
    auto update = aUpdate.Get(i);
    if (!update.has_compression_type()) {
      NS_WARNING(nsPrintfCString("%s with no compression type.",
                                  aIsAddition ? "Addition" : "Removal").get());
      continue;
    }

    switch (update.compression_type()) {
    case COMPRESSION_TYPE_UNSPECIFIED:
      NS_WARNING("Unspecified compression type.");
      break;

    case RAW:
      ret = (aIsAddition ? ProcessRawAddition(aTableUpdate, update)
                         : ProcessRawRemoval(aTableUpdate, update));
      break;

    case RICE:
      ret = (aIsAddition ? ProcessEncodedAddition(aTableUpdate, update)
                         : ProcessEncodedRemoval(aTableUpdate, update));
      break;
    }
  }

  return ret;
}

nsresult
ProtocolParserProtobuf::ProcessRawAddition(TableUpdateV4& aTableUpdate,
                                           const ThreatEntrySet& aAddition)
{
  if (!aAddition.has_raw_hashes()) {
    PARSER_LOG(("* No raw addition."));
    return NS_OK;
  }

  auto rawHashes = aAddition.raw_hashes();
  if (!rawHashes.has_prefix_size()) {
    NS_WARNING("Raw hash has no prefix size");
    return NS_OK;
  }

  uint32_t prefixSize = rawHashes.prefix_size();
  MOZ_ASSERT(prefixSize >= PREFIX_SIZE && prefixSize <= COMPLETE_SIZE);

  nsCString prefixes;
  if (!prefixes.Assign(rawHashes.raw_hashes().c_str(),
                       rawHashes.raw_hashes().size(), mozilla::fallible)) {
    return NS_ERROR_OUT_OF_MEMORY;
  }
  MOZ_ASSERT(prefixes.Length() % prefixSize == 0,
             "PrefixString length must be a multiple of the prefix size.");

  if (LOG_ENABLED()) {
    PARSER_LOG((" Raw addition (%d-byte prefixes)", prefixSize));
    PARSER_LOG(("  - # of prefixes: %u", prefixes.Length() / prefixSize));
    if (4 == prefixSize) {
      uint32_t* fixedLengthPrefixes = (uint32_t*)prefixes.get();
      PARSER_LOG(("  - Memory address: 0x%p", fixedLengthPrefixes));
    }
  }

  aTableUpdate.NewPrefixes(prefixSize, prefixes);
  return NS_OK;
}

nsresult
ProtocolParserProtobuf::ProcessRawRemoval(TableUpdateV4& aTableUpdate,
                                          const ThreatEntrySet& aRemoval)
{
  if (!aRemoval.has_raw_indices()) {
    NS_WARNING("A removal has no indices.");
    return NS_OK;
  }

  // indices is an array of int32.
  auto indices = aRemoval.raw_indices().indices();
  PARSER_LOG(("* Raw removal"));
  PARSER_LOG(("  - # of removal: %d", indices.size()));

  nsresult rv = aTableUpdate.NewRemovalIndices((const uint32_t*)indices.data(),
                                               indices.size());
  if (NS_FAILED(rv)) {
    PARSER_LOG(("Failed to create new removal indices."));
    return rv;
  }

  return NS_OK;
}

static nsresult
DoRiceDeltaDecode(const RiceDeltaEncoding& aEncoding,
                  nsTArray<uint32_t>& aDecoded)
{
  if (!aEncoding.has_first_value()) {
    PARSER_LOG(("The encoding info is incomplete."));
    return NS_ERROR_UC_PARSER_MISSING_PARAM;
  }
  if (aEncoding.num_entries() > 0 &&
      (!aEncoding.has_rice_parameter() || !aEncoding.has_encoded_data())) {
    PARSER_LOG(("Rice parameter or encoded data is missing."));
    return NS_ERROR_UC_PARSER_MISSING_PARAM;
  }

  PARSER_LOG(("* Encoding info:"));
  PARSER_LOG(("  - First value: %" PRId64, aEncoding.first_value()));
  PARSER_LOG(("  - Num of entries: %d", aEncoding.num_entries()));
  PARSER_LOG(("  - Rice parameter: %d", aEncoding.rice_parameter()));

  // Set up the input buffer. Note that the bits should be read
  // from LSB to MSB so that we in-place reverse the bits before
  // feeding to the decoder.
  auto encoded = const_cast<RiceDeltaEncoding&>(aEncoding).mutable_encoded_data();
  RiceDeltaDecoder decoder((uint8_t*)encoded->c_str(), encoded->size());

  // Setup the output buffer. The "first value" is included in
  // the output buffer.
  if (!aDecoded.SetLength(aEncoding.num_entries() + 1, mozilla::fallible)) {
    NS_WARNING("Not enough memory to decode the RiceDelta input.");
    return NS_ERROR_OUT_OF_MEMORY;
  }

  // Decode!
  bool rv = decoder.Decode(aEncoding.rice_parameter(),
                           aEncoding.first_value(), // first value.
                           aEncoding.num_entries(), // # of entries (first value not included).
                           &aDecoded[0]);

  NS_ENSURE_TRUE(rv, NS_ERROR_UC_PARSER_DECODE_FAILURE);

  return NS_OK;
}

nsresult
ProtocolParserProtobuf::ProcessEncodedAddition(TableUpdateV4& aTableUpdate,
                                               const ThreatEntrySet& aAddition)
{
  if (!aAddition.has_rice_hashes()) {
    PARSER_LOG(("* No rice encoded addition."));
    return NS_OK;
  }

  nsTArray<uint32_t> decoded;
  nsresult rv = DoRiceDeltaDecode(aAddition.rice_hashes(), decoded);
  if (NS_FAILED(rv)) {
    PARSER_LOG(("Failed to parse encoded prefixes."));
    return rv;
  }

  //  Say we have the following raw prefixes
  //                              BE            LE
  //   00 00 00 01                 1      16777216
  //   00 00 02 00               512        131072
  //   00 03 00 00            196608           768
  //   04 00 00 00          67108864             4
  //
  // which can be treated as uint32 (big-endian) sorted in increasing order:
  //
  // [1, 512, 196608, 67108864]
  //
  // According to https://developers.google.com/safe-browsing/v4/compression,
  // the following should be done prior to compression:
  //
  // 1) re-interpret in little-endian ==> [16777216, 131072, 768, 4]
  // 2) sort in increasing order       ==> [4, 768, 131072, 16777216]
  //
  // In order to get the original byte stream from |decoded|
  // ([4, 768, 131072, 16777216] in this case), we have to:
  //
  // 1) sort in big-endian order      ==> [16777216, 131072, 768, 4]
  // 2) copy each uint32 in little-endian to the result string
  //

  // The 4-byte prefixes have to be re-sorted in Big-endian increasing order.
  struct CompareBigEndian
  {
    bool Equals(const uint32_t& aA, const uint32_t& aB) const
    {
      return aA == aB;
    }

    bool LessThan(const uint32_t& aA, const uint32_t& aB) const
    {
      return NativeEndian::swapToBigEndian(aA) <
             NativeEndian::swapToBigEndian(aB);
    }
  };
  decoded.Sort(CompareBigEndian());

  // The encoded prefixes are always 4 bytes.
  nsCString prefixes;
  if (!prefixes.SetCapacity(decoded.Length() * 4, mozilla::fallible)) {
    return NS_ERROR_OUT_OF_MEMORY;
  }
  for (size_t i = 0; i < decoded.Length(); i++) {
    // Note that the third argument is the number of elements we want
    // to copy (and swap) but not the number of bytes we want to copy.
    char p[4];
    NativeEndian::copyAndSwapToLittleEndian(p, &decoded[i], 1);
    prefixes.Append(p, 4);
  }

  aTableUpdate.NewPrefixes(4, prefixes);
  return NS_OK;
}

nsresult
ProtocolParserProtobuf::ProcessEncodedRemoval(TableUpdateV4& aTableUpdate,
                                              const ThreatEntrySet& aRemoval)
{
  if (!aRemoval.has_rice_indices()) {
    PARSER_LOG(("* No rice encoded removal."));
    return NS_OK;
  }

  nsTArray<uint32_t> decoded;
  nsresult rv = DoRiceDeltaDecode(aRemoval.rice_indices(), decoded);
  if (NS_FAILED(rv)) {
    PARSER_LOG(("Failed to decode encoded removal indices."));
    return rv;
  }

  // The encoded prefixes are always 4 bytes.
  rv = aTableUpdate.NewRemovalIndices(&decoded[0], decoded.Length());
  if (NS_FAILED(rv)) {
    PARSER_LOG(("Failed to create new removal indices."));
    return rv;
  }

  return NS_OK;
}

} // namespace safebrowsing
} // namespace mozilla