7zin.c

一个7ZIP的解压源码。比较详细。里面含有四种语言的实现代码。

      {
        UInt32 c2;
        if (value >= 0xDC00)
          return SZE_ARCHIVE_ERROR;
        if (pos + 2 > sd->Size)
          return SZE_ARCHIVE_ERROR;
        c2 = (UInt32)(sd->Data[pos] | (((UInt32)sd->Data[pos + 1]) << 8));
        pos += 2;
        if (c2 < 0xDC00 || c2 >= 0xE000)
          return SZE_ARCHIVE_ERROR;
        value = ((value - 0xD800) << 10) | (c2 - 0xDC00);
      }
      for (numAdds = 1; numAdds < 5; numAdds++)
        if (value < (((UInt32)1) << (numAdds * 5 + 6)))
          break;
      len += numAdds;
    }

    MY_ALLOC(char, file->Name, (size_t)len, allocFunc);

    len = 0;
    while(2 <= sd->Size)
    {
      int numAdds;
      UInt32 value = (UInt32)(sd->Data[0] | (((UInt32)sd->Data[1]) << 8));
      SzSkeepDataSize(sd, 2);
      if (value < 0x80)
      {
        file->Name[len++] = (char)value;
        if (value == 0)
          break;
        continue;
      }
      if (value >= 0xD800 && value < 0xE000)
      {
        UInt32 c2 = (UInt32)(sd->Data[0] | (((UInt32)sd->Data[1]) << 8));
        SzSkeepDataSize(sd, 2);
        value = ((value - 0xD800) << 10) | (c2 - 0xDC00);
      }
      for (numAdds = 1; numAdds < 5; numAdds++)
        if (value < (((UInt32)1) << (numAdds * 5 + 6)))
          break;
      file->Name[len++] = (char)(kUtf8Limits[numAdds - 1] + (value >> (6 * numAdds)));
      do
      {
        numAdds--;
        file->Name[len++] = (char)(0x80 + ((value >> (6 * numAdds)) & 0x3F));
      }
      while(numAdds > 0);

      len += numAdds;
    }
  }
  return SZ_OK;
}

SZ_RESULT SzReadHeader2(
    CSzData *sd, 
    CArchiveDatabaseEx *db,   /* allocMain */
    CFileSize **unPackSizes,  /* allocTemp */
    Byte **digestsDefined,    /* allocTemp */
    UInt32 **digests,         /* allocTemp */
    Byte **emptyStreamVector, /* allocTemp */
    Byte **emptyFileVector,   /* allocTemp */
    Byte **lwtVector,         /* allocTemp */
    ISzAlloc *allocMain, 
    ISzAlloc *allocTemp)
{
  UInt64 type;
  UInt32 numUnPackStreams = 0;
  UInt32 numFiles = 0;
  CFileItem *files = 0;
  UInt32 numEmptyStreams = 0;
  UInt32 i;

  RINOK(SzReadID(sd, &type));

  if (type == k7zIdArchiveProperties)
  {
    RINOK(SzReadArchiveProperties(sd));
    RINOK(SzReadID(sd, &type));
  }
 
 
  if (type == k7zIdMainStreamsInfo)
  {
    RINOK(SzReadStreamsInfo(sd,
        &db->ArchiveInfo.DataStartPosition,
        &db->Database, 
        &numUnPackStreams,
        unPackSizes,
        digestsDefined,
        digests, allocMain->Alloc, allocTemp));
    db->ArchiveInfo.DataStartPosition += db->ArchiveInfo.StartPositionAfterHeader;
    RINOK(SzReadID(sd, &type));
  }

  if (type == k7zIdEnd)
    return SZ_OK;
  if (type != k7zIdFilesInfo)
    return SZE_ARCHIVE_ERROR;
  
  RINOK(SzReadNumber32(sd, &numFiles));
  db->Database.NumFiles = numFiles;

  MY_ALLOC(CFileItem, files, (size_t)numFiles, allocMain->Alloc);

  db->Database.Files = files;
  for(i = 0; i < numFiles; i++)
    SzFileInit(files + i);

  for (;;)
  {
    UInt64 type;
    UInt64 size;
    RINOK(SzReadID(sd, &type));
    if (type == k7zIdEnd)
      break;
    RINOK(SzReadNumber(sd, &size));

    if ((UInt64)(int)type != type)
    {
      RINOK(SzSkeepDataSize(sd, size));
    }
    else
    switch((int)type)
    {
      case k7zIdName:
      {
        RINOK(SzReadSwitch(sd));
        RINOK(SzReadFileNames(sd, numFiles, files, allocMain->Alloc))
        break;
      }
      case k7zIdEmptyStream:
      {
        RINOK(SzReadBoolVector(sd, numFiles, emptyStreamVector, allocTemp->Alloc));
        numEmptyStreams = 0;
        for (i = 0; i < numFiles; i++)
          if ((*emptyStreamVector)[i])
            numEmptyStreams++;
        break;
      }
      case k7zIdEmptyFile:
      {
        RINOK(SzReadBoolVector(sd, numEmptyStreams, emptyFileVector, allocTemp->Alloc));
        break;
      }
      case k7zIdLastWriteTime:
      {
        RINOK(SzReadBoolVector2(sd, numFiles, lwtVector, allocTemp->Alloc));
        RINOK(SzReadSwitch(sd));
        for (i = 0; i < numFiles; i++)
        {
          CFileItem *f = &files[i];
          Byte defined = (*lwtVector)[i];
          f->IsLastWriteTimeDefined = defined;
          f->LastWriteTime.Low = f->LastWriteTime.High = 0;
          if (defined)
          {
            RINOK(SzReadUInt32(sd, &f->LastWriteTime.Low));
            RINOK(SzReadUInt32(sd, &f->LastWriteTime.High));
          }
        }
        break;
      }
      default:
      {
        RINOK(SzSkeepDataSize(sd, size));
      }
    }
  }

  {
    UInt32 emptyFileIndex = 0;
    UInt32 sizeIndex = 0;
    for(i = 0; i < numFiles; i++)
    {
      CFileItem *file = files + i;
      file->IsAnti = 0;
      if (*emptyStreamVector == 0)
        file->HasStream = 1;
      else
        file->HasStream = (Byte)((*emptyStreamVector)[i] ? 0 : 1);
      if(file->HasStream)
      {
        file->IsDirectory = 0;
        file->Size = (*unPackSizes)[sizeIndex];
        file->FileCRC = (*digests)[sizeIndex];
        file->IsFileCRCDefined = (Byte)(*digestsDefined)[sizeIndex];
        sizeIndex++;
      }
      else
      {
        if (*emptyFileVector == 0)
          file->IsDirectory = 1;
        else
          file->IsDirectory = (Byte)((*emptyFileVector)[emptyFileIndex] ? 0 : 1);
        emptyFileIndex++;
        file->Size = 0;
        file->IsFileCRCDefined = 0;
      }
    }
  }
  return SzArDbExFill(db, allocMain->Alloc);
}

SZ_RESULT SzReadHeader(
    CSzData *sd, 
    CArchiveDatabaseEx *db, 
    ISzAlloc *allocMain, 
    ISzAlloc *allocTemp)
{
  CFileSize *unPackSizes = 0;
  Byte *digestsDefined = 0;
  UInt32 *digests = 0;
  Byte *emptyStreamVector = 0;
  Byte *emptyFileVector = 0;
  Byte *lwtVector = 0;
  SZ_RESULT res = SzReadHeader2(sd, db, 
      &unPackSizes, &digestsDefined, &digests,
      &emptyStreamVector, &emptyFileVector, &lwtVector, 
      allocMain, allocTemp);
  allocTemp->Free(unPackSizes);
  allocTemp->Free(digestsDefined);
  allocTemp->Free(digests);
  allocTemp->Free(emptyStreamVector);
  allocTemp->Free(emptyFileVector);
  allocTemp->Free(lwtVector);
  return res;
} 

SZ_RESULT SzReadAndDecodePackedStreams2(
    ISzInStream *inStream, 
    CSzData *sd,
    CSzByteBuffer *outBuffer,
    CFileSize baseOffset, 
    CArchiveDatabase *db,
    CFileSize **unPackSizes,
    Byte **digestsDefined,
    UInt32 **digests,
    #ifndef _LZMA_IN_CB
    Byte **inBuffer,
    #endif
    ISzAlloc *allocTemp)
{

  UInt32 numUnPackStreams = 0;
  CFileSize dataStartPos;
  CFolder *folder;
  #ifndef _LZMA_IN_CB
  CFileSize packSize = 0;
  UInt32 i = 0;
  #endif
  CFileSize unPackSize;
  SZ_RESULT res;

  RINOK(SzReadStreamsInfo(sd, &dataStartPos, db,
      &numUnPackStreams,  unPackSizes, digestsDefined, digests, 
      allocTemp->Alloc, allocTemp));
  
  dataStartPos += baseOffset;
  if (db->NumFolders != 1)
    return SZE_ARCHIVE_ERROR;

  folder = db->Folders;
  unPackSize = SzFolderGetUnPackSize(folder);
  
  RINOK(inStream->Seek(inStream, dataStartPos));

  #ifndef _LZMA_IN_CB
  for (i = 0; i < db->NumPackStreams; i++)
    packSize += db->PackSizes[i];

  MY_ALLOC(Byte, *inBuffer, (size_t)packSize, allocTemp->Alloc);

  RINOK(SafeReadDirect(inStream, *inBuffer, (size_t)packSize));
  #endif

  if (!SzByteBufferCreate(outBuffer, (size_t)unPackSize, allocTemp->Alloc))
    return SZE_OUTOFMEMORY;
  
  res = SzDecode(db->PackSizes, folder, 
          #ifdef _LZMA_IN_CB
          inStream, dataStartPos, 
          #else
          *inBuffer, 
          #endif
          outBuffer->Items, (size_t)unPackSize, allocTemp);
  RINOK(res)
  if (folder->UnPackCRCDefined)
    if (CrcCalc(outBuffer->Items, (size_t)unPackSize) != folder->UnPackCRC)
      return SZE_FAIL;
  return SZ_OK;
}

SZ_RESULT SzReadAndDecodePackedStreams(
    ISzInStream *inStream, 
    CSzData *sd,
    CSzByteBuffer *outBuffer,
    CFileSize baseOffset, 
    ISzAlloc *allocTemp)
{
  CArchiveDatabase db;
  CFileSize *unPackSizes = 0;
  Byte *digestsDefined = 0;
  UInt32 *digests = 0;
  #ifndef _LZMA_IN_CB
  Byte *inBuffer = 0;
  #endif
  SZ_RESULT res;
  SzArchiveDatabaseInit(&db);
  res = SzReadAndDecodePackedStreams2(inStream, sd, outBuffer, baseOffset, 
    &db, &unPackSizes, &digestsDefined, &digests, 
    #ifndef _LZMA_IN_CB
    &inBuffer,
    #endif
    allocTemp);
  SzArchiveDatabaseFree(&db, allocTemp->Free);
  allocTemp->Free(unPackSizes);
  allocTemp->Free(digestsDefined);
  allocTemp->Free(digests);
  #ifndef _LZMA_IN_CB
  allocTemp->Free(inBuffer);
  #endif
  return res;
}

SZ_RESULT SzArchiveOpen2(
    ISzInStream *inStream, 
    CArchiveDatabaseEx *db,
    ISzAlloc *allocMain, 
    ISzAlloc *allocTemp)
{
  Byte signature[k7zSignatureSize];
  Byte version;
  UInt32 crcFromArchive;
  UInt64 nextHeaderOffset;
  UInt64 nextHeaderSize;
  UInt32 nextHeaderCRC;
  UInt32 crc = 0;
  CFileSize pos = 0;
  CSzByteBuffer buffer;
  CSzData sd;
  SZ_RESULT res;

  RINOK(SafeReadDirect(inStream, signature, k7zSignatureSize));

  if (!TestSignatureCandidate(signature))
    return SZE_ARCHIVE_ERROR;

  /*
  db.Clear();
  db.ArchiveInfo.StartPosition = _arhiveBeginStreamPosition;
  */
  RINOK(SafeReadDirectByte(inStream, &version));
  if (version != k7zMajorVersion)
    return SZE_ARCHIVE_ERROR;
  RINOK(SafeReadDirectByte(inStream, &version));

  RINOK(SafeReadDirectUInt32(inStream, &crcFromArchive, &crc));

  crc = CRC_INIT_VAL;
  RINOK(SafeReadDirectUInt64(inStream, &nextHeaderOffset, &crc));
  RINOK(SafeReadDirectUInt64(inStream, &nextHeaderSize, &crc));
  RINOK(SafeReadDirectUInt32(inStream, &nextHeaderCRC, &crc));

  pos = k7zStartHeaderSize;
  db->ArchiveInfo.StartPositionAfterHeader = pos;
  
  if (CRC_GET_DIGEST(crc) != crcFromArchive)
    return SZE_ARCHIVE_ERROR;

  if (nextHeaderSize == 0)
    return SZ_OK;

  RINOK(inStream->Seek(inStream, (CFileSize)(pos + nextHeaderOffset)));

  if (!SzByteBufferCreate(&buffer, (size_t)nextHeaderSize, allocTemp->Alloc))
    return SZE_OUTOFMEMORY;

  res = SafeReadDirect(inStream, buffer.Items, (size_t)nextHeaderSize);
  if (res == SZ_OK)
  {
    res = SZE_ARCHIVE_ERROR;
    if (CrcCalc(buffer.Items, (UInt32)nextHeaderSize) == nextHeaderCRC)
    {
      for (;;)
      {
        UInt64 type;
        sd.Data = buffer.Items;
        sd.Size = buffer.Capacity;
        res = SzReadID(&sd, &type);
        if (res != SZ_OK)
          break;
        if (type == k7zIdHeader)
        {
          res = SzReadHeader(&sd, db, allocMain, allocTemp);
          break;
        }
        if (type != k7zIdEncodedHeader)
        {
          res = SZE_ARCHIVE_ERROR;
          break;
        }
        {
          CSzByteBuffer outBuffer;
          res = SzReadAndDecodePackedStreams(inStream, &sd, &outBuffer, 
              db->ArchiveInfo.StartPositionAfterHeader, 
              allocTemp);
          if (res != SZ_OK)
          {
            SzByteBufferFree(&outBuffer, allocTemp->Free);
            break;
          }
          SzByteBufferFree(&buffer, allocTemp->Free);
          buffer.Items = outBuffer.Items;
          buffer.Capacity = outBuffer.Capacity;
        }
      }
    }
  }
  SzByteBufferFree(&buffer, allocTemp->Free);
  return res;
}

SZ_RESULT SzArchiveOpen(
    ISzInStream *inStream, 
    CArchiveDatabaseEx *db,
    ISzAlloc *allocMain, 
    ISzAlloc *allocTemp)
{
  SZ_RESULT res = SzArchiveOpen2(inStream, db, allocMain, allocTemp);
  if (res != SZ_OK)
    SzArDbExFree(db, allocMain->Free);
  return res;
}