/[svn]/libgig/trunk/src/RIFF.cpp

Diff of /libgig/trunk/src/RIFF.cpp

Parent Directory | Revision Log | View Patch Patch

-revision 666 by persson,
Sun Jun 19 15:18:59 2005 UTC
+revision 3921 by schoenebeck,
Mon Jun 14 09:28:04 2021 UTC
 Line 1
  /***************************************************************************
   *                                                                         *
-  *   libgig - C++ cross-platform Gigasampler format file loader library    *
+  *   libgig - C++ cross-platform Gigasampler format file access library    *
   *                                                                         *
-  *   Copyright (C) 2003-2005 by Christian Schoenebeck                      *
+  *   Copyright (C) 2003-2021 by Christian Schoenebeck                      *
   *                              <cuse@users.sourceforge.net>               *
   *                                                                         *
   *   This library is free software; you can redistribute it and/or modify  *
 Line 20
   *   Foundation, Inc., 59 Temple Place, Suite 330, Boston,                 *
   *   MA  02111-1307  USA                                                   *
   ***************************************************************************/
- #if 1
+ #include <algorithm>
+ #include <set>
+ #include <string.h>
  #include "RIFF.h"
+ #include "helper.h"
+ #if POSIX
+ # include <errno.h>
+ #endif
  namespace RIFF {
+ // *************** Internal functions **************
+ // *
+     /// Returns a human readable path of the given chunk.
+     static String __resolveChunkPath(Chunk* pCk) {
+         String sPath;
+         for (Chunk* pChunk = pCk; pChunk; pChunk = pChunk->GetParent()) {
+             if (pChunk->GetChunkID() == CHUNK_ID_LIST) {
+                 List* pList = (List*) pChunk;
+                 sPath = "->'" + pList->GetListTypeString() + "'" + sPath;
+             } else {
+                 sPath = "->'" + pChunk->GetChunkIDString() + "'" + sPath;
+             }
+         }
+         return sPath;
+     }
+     inline static bool _isValidHandle(File::Handle handle) {
+         #if defined(WIN32)
+         return handle != INVALID_HANDLE_VALUE;
+         #else
+         return handle;
+         #endif
+     }
+     inline static void _close(File::Handle handle) {
+         if (!_isValidHandle(handle)) return;
+         #if POSIX
+         close(handle);
+         #elif defined(WIN32)
+         CloseHandle(handle);
+         #else
+         fclose(handle);
+         #endif
+     }
+ // *************** progress_t ***************
+ // *
+     progress_t::progress_t() {
+         callback    = NULL;
+         custom      = NULL;
+         __range_min = 0.0f;
+         __range_max = 1.0f;
+     }
+     /**
+      * Divides this progress task into the requested amount of equal weighted
+      * sub-progress tasks and returns a vector with those subprogress tasks.
+      *
+      * @param iSubtasks - total amount sub tasks this task should be subdivided
+      * @returns subtasks
+      */
+     std::vector<progress_t> progress_t::subdivide(int iSubtasks) {
+         std::vector<progress_t> v;
+         for (int i = 0; i < iSubtasks; ++i) {
+             progress_t p;
+             __divide_progress(this, &p, iSubtasks, i);
+             v.push_back(p);
+         }
+         return v;
+     }
+     /**
+      * Divides this progress task into the requested amount of sub-progress
+      * tasks, where each one of those new sub-progress tasks is created with its
+      * requested individual weight / portion, and finally returns a vector
+      * with those new subprogress tasks.
+      *
+      * The amount of subprogresses to be created is determined by this method
+      * by calling @c vSubTaskPortions.size() .
+      *
+      * Example: consider you wanted to create 3 subprogresses where the 1st
+      * subtask should be assigned 10% of the new 3 subprogresses' overall
+      * progress, the 2nd subtask should be assigned 50% of the new 3
+      * subprogresses' overall progress, and the 3rd subtask should be assigned
+      * 40%, then you might call this method like this:
+      * @code
+      * std::vector<progress_t> subprogresses = progress.subdivide({0.1, 0.5, 0.4});
+      * @endcode
+      *
+      * @param vSubTaskPortions - amount and individual weight of subtasks to be
+      *                           created
+      * @returns subtasks
+      */
+     std::vector<progress_t> progress_t::subdivide(std::vector<float> vSubTaskPortions) {
+         float fTotal = 0.f; // usually 1.0, but we sum the portions up below to be sure
+         for (int i = 0; i < vSubTaskPortions.size(); ++i)
+             fTotal += vSubTaskPortions[i];
+         float fLow = 0.f, fHigh = 0.f;
+         std::vector<progress_t> v;
+         for (int i = 0; i < vSubTaskPortions.size(); ++i) {
+             fLow  = fHigh;
+             fHigh = vSubTaskPortions[i];
+             progress_t p;
+             __divide_progress(this, &p, fTotal, fLow, fHigh);
+             v.push_back(p);
+         }
+         return v;
+     }
  // *************** Chunk **************
  // *
-     Chunk::Chunk() {
+     Chunk::Chunk(File* pFile) {
-         #if DEBUG
+         #if DEBUG_RIFF
-         std::cout << "Chunk::Chunk()" << std::endl;
+         std::cout << "Chunk::Chunk(File* pFile)" << std::endl;
-         #endif // DEBUG
+         #endif // DEBUG_RIFF
-         ulPos      = 0;
+         chunkPos.ullPos = 0;
          pParent    = NULL;
          pChunkData = NULL;
+         ullCurrentChunkSize = 0;
+         ullNewChunkSize = 0;
+         ullChunkDataSize = 0;
+         ChunkID    = CHUNK_ID_RIFF;
+         this->pFile = pFile;
      }
-     #if POSIX
+     Chunk::Chunk(File* pFile, file_offset_t StartPos, List* Parent) {
-     Chunk::Chunk(int hFile, unsigned long StartPos, bool EndianNative, List* Parent) {
+         #if DEBUG_RIFF
-     #else
+         std::cout << "Chunk::Chunk(File*,file_offset_t,List*),StartPos=" << StartPos << std::endl;
-     Chunk::Chunk(FILE* hFile, unsigned long StartPos, bool EndianNative, List* Parent) {
+         #endif // DEBUG_RIFF
-     #endif // POSIX
+         this->pFile   = pFile;
-         #if DEBUG
+         ullStartPos   = StartPos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize);
-         std::cout << "Chunk::Chunk(FILE,ulong,bool,List*),StartPos=" << StartPos << std::endl;
-         #endif // DEBUG
-         Chunk::hFile  = hFile;
-         ulStartPos    = StartPos + CHUNK_HEADER_SIZE;
-         bEndianNative = EndianNative;
          pParent       = Parent;
-         ulPos         = 0;
+         chunkPos.ullPos = 0;
          pChunkData    = NULL;
+         ullCurrentChunkSize = 0;
+         ullNewChunkSize = 0;
+         ullChunkDataSize = 0;
          ReadHeader(StartPos);
      }
+     Chunk::Chunk(File* pFile, List* pParent, uint32_t uiChunkID, file_offset_t ullBodySize) {
+         this->pFile      = pFile;
+         ullStartPos      = 0; // arbitrary usually, since it will be updated when we write the chunk
+         this->pParent    = pParent;
+         chunkPos.ullPos  = 0;
+         pChunkData       = NULL;
+         ChunkID          = uiChunkID;
+         ullChunkDataSize = 0;
+         ullCurrentChunkSize = 0;
+         ullNewChunkSize  = ullBodySize;
+     }
      Chunk::~Chunk() {
          if (pChunkData) delete[] pChunkData;
      }
-     void Chunk::ReadHeader(unsigned long fPos) {
+     void Chunk::ReadHeader(file_offset_t filePos) {
-         #if DEBUG
+         #if DEBUG_RIFF
-         std::cout << "Chunk::Readheader(" << fPos << ") ";
+         std::cout << "Chunk::Readheader(" << filePos << ") ";
-         #endif // DEBUG
+         #endif // DEBUG_RIFF
+         ChunkID = 0;
+         ullNewChunkSize = ullCurrentChunkSize = 0;
+         const File::Handle hRead = pFile->FileHandle();
          #if POSIX
-         if (lseek(hFile, fPos, SEEK_SET) != -1) {
+         if (lseek(hRead, filePos, SEEK_SET) != -1) {
-             read(hFile, &ChunkID, 4);
+             read(hRead, &ChunkID, 4);
-             read(hFile, &ChunkSize, 4);
+             read(hRead, &ullCurrentChunkSize, pFile->FileOffsetSize);
+         #elif defined(WIN32)
+         LARGE_INTEGER liFilePos;
+         liFilePos.QuadPart = filePos;
+         if (SetFilePointerEx(hRead, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN)) {
+             DWORD dwBytesRead;
+             ReadFile(hRead, &ChunkID, 4, &dwBytesRead, NULL);
+             ReadFile(hRead, &ullCurrentChunkSize, pFile->FileOffsetSize, &dwBytesRead, NULL);
          #else
-         if (!fseek(hFile, fPos, SEEK_SET)) {
+         if (!fseeko(hRead, filePos, SEEK_SET)) {
-             fread(&ChunkID, 4, 1, hFile);
+             fread(&ChunkID, 4, 1, hRead);
-             fread(&ChunkSize, 4, 1, hFile);
+             fread(&ullCurrentChunkSize, pFile->FileOffsetSize, 1, hRead);
          #endif // POSIX
              #if WORDS_BIGENDIAN
              if (ChunkID == CHUNK_ID_RIFF) {
-                 bEndianNative = false;
+                 pFile->bEndianNative = false;
              }
              #else // little endian
              if (ChunkID == CHUNK_ID_RIFX) {
-                 bEndianNative = false;
+                 pFile->bEndianNative = false;
                  ChunkID = CHUNK_ID_RIFF;
              }
              #endif // WORDS_BIGENDIAN
-             if (!bEndianNative) {
+             if (!pFile->bEndianNative) {
                  //swapBytes_32(&ChunkID);
-                 swapBytes_32(&ChunkSize);
+                 if (pFile->FileOffsetSize == 4)
+                     swapBytes_32(&ullCurrentChunkSize);
+                 else
+                     swapBytes_64(&ullCurrentChunkSize);
              }
-             #if DEBUG
+             #if DEBUG_RIFF
              std::cout << "ckID=" << convertToString(ChunkID) << " ";
-             std::cout << "ckSize=" << ChunkSize << " ";
+             std::cout << "ckSize=" << ullCurrentChunkSize << " ";
-             std::cout << "bEndianNative=" << bEndianNative << std::endl;
+             std::cout << "bEndianNative=" << pFile->bEndianNative << std::endl;
-             #endif // DEBUG
+             #endif // DEBUG_RIFF
+             ullNewChunkSize = ullCurrentChunkSize;
          }
      }
+     void Chunk::WriteHeader(file_offset_t filePos) {
+         uint32_t uiNewChunkID = ChunkID;
+         if (ChunkID == CHUNK_ID_RIFF) {
+             #if WORDS_BIGENDIAN
+             if (pFile->bEndianNative) uiNewChunkID = CHUNK_ID_RIFX;
+             #else // little endian
+             if (!pFile->bEndianNative) uiNewChunkID = CHUNK_ID_RIFX;
+             #endif // WORDS_BIGENDIAN
+         }
+         uint64_t ullNewChunkSize = this->ullNewChunkSize;
+         if (!pFile->bEndianNative) {
+             if (pFile->FileOffsetSize == 4)
+                 swapBytes_32(&ullNewChunkSize);
+             else
+                 swapBytes_64(&ullNewChunkSize);
+         }
+         const File::Handle hWrite = pFile->FileWriteHandle();
+         #if POSIX
+         if (lseek(hWrite, filePos, SEEK_SET) != -1) {
+             write(hWrite, &uiNewChunkID, 4);
+             write(hWrite, &ullNewChunkSize, pFile->FileOffsetSize);
+         }
+         #elif defined(WIN32)
+         LARGE_INTEGER liFilePos;
+         liFilePos.QuadPart = filePos;
+         if (SetFilePointerEx(hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN)) {
+             DWORD dwBytesWritten;
+             WriteFile(hWrite, &uiNewChunkID, 4, &dwBytesWritten, NULL);
+             WriteFile(hWrite, &ullNewChunkSize, pFile->FileOffsetSize, &dwBytesWritten, NULL);
+         }
+         #else
+         if (!fseeko(hWrite, filePos, SEEK_SET)) {
+             fwrite(&uiNewChunkID, 4, 1, hWrite);
+             fwrite(&ullNewChunkSize, pFile->FileOffsetSize, 1, hWrite);
+         }
+         #endif // POSIX
+     }
      /**
       *  Returns the String representation of the chunk's ID (e.g. "RIFF",
       *  "LIST").
       */
-     String Chunk::GetChunkIDString() {
+     String Chunk::GetChunkIDString() const {
          return convertToString(ChunkID);
      }
      /**
+      * This is an internal-only method which must not be used by any application
+      * and might change at any time.
+      *
+      * Returns a reference (memory location) of the chunk's current file
+      * (read/write) position variable which depends on the current value of
+      * File::IsIOPerThread().
+      */
+     file_offset_t& Chunk::GetPosUnsafeRef() {
+         if (!pFile->IsIOPerThread()) return chunkPos.ullPos;
+         const std::thread::id tid = std::this_thread::get_id();
+         return chunkPos.byThread[tid];
+     }
+     /**
+      * Current read/write position within the chunk data body (starting with 0).
+      *
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::GetPos() const {
+         if (!pFile->IsIOPerThread()) return chunkPos.ullPos;
+         const std::thread::id tid = std::this_thread::get_id();
+         std::lock_guard<std::mutex> lock(chunkPos.mutex);
+         return chunkPos.byThread[tid];
+     }
+     /**
+      * Current, actual offset in file of current chunk data body read/write
+      * position.
+      *
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::GetFilePos() const {
+         return ullStartPos + GetPos();
+     }
+     /**
       *  Sets the position within the chunk body, thus within the data portion
       *  of the chunk (in bytes).
       *
+      *  <b>Caution:</b> the position will be reset to zero whenever
+      *  File::Save() was called.
+      *
       *  @param Where  - position offset (in bytes)
       *  @param Whence - optional: defines to what <i>\a Where</i> relates to,
       *                  if omitted \a Where relates to beginning of the chunk
       *                  data
+      *  @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::SetPos(unsigned long Where, stream_whence_t Whence) {
+     file_offset_t Chunk::SetPos(file_offset_t Where, stream_whence_t Whence) {
-      #if DEBUG
+         #if DEBUG_RIFF
-      std::cout << "Chunk::SetPos(ulong)" << std::endl;
+         std::cout << "Chunk::SetPos(file_offset_t,stream_whence_t)" << std::endl;
-      #endif // DEBUG
+         #endif // DEBUG_RIFF
+         std::lock_guard<std::mutex> lock(chunkPos.mutex);
+         file_offset_t& pos = GetPosUnsafeRef();
          switch (Whence) {
              case stream_curpos:
-                 ulPos += Where;
+                 pos += Where;
                  break;
              case stream_end:
-                 ulPos = ChunkSize - 1 - Where;
+                 pos = ullCurrentChunkSize - 1 - Where;
                  break;
              case stream_backward:
-                 ulPos -= Where;
+                 pos -= Where;
                  break;
              case stream_start: default:
-                 ulPos = Where;
+                 pos = Where;
                  break;
          }
-         if (ulPos > ChunkSize) ulPos = ChunkSize;
+         if (pos > ullCurrentChunkSize) pos = ullCurrentChunkSize;
-         return ulPos;
+         return pos;
      }
      /**
-Line 141 
 namespace RIFF {
+Line 371 
 namespace RIFF {
       *  of the chunk data.
       *
       *  @returns  number of bytes left to read
+      *  @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::RemainingBytes() {
+     file_offset_t Chunk::RemainingBytes() const {
-        #if DEBUG
+         #if DEBUG_RIFF
-        std::cout << "Chunk::Remainingbytes()=" << ChunkSize - ulPos << std::endl;
+         std::cout << "Chunk::Remainingbytes()=" << ullCurrentChunkSize - ullPos << std::endl;
-        #endif // DEBUG
+         #endif // DEBUG_RIFF
-         return ChunkSize - ulPos;
+         const file_offset_t pos = GetPos();
+         return (ullCurrentChunkSize > pos) ? ullCurrentChunkSize - pos : 0;
+     }
+     /**
+      *  Returns the actual total size in bytes (including header) of this Chunk
+      *  if being stored to a file.
+      *
+      *  @param fileOffsetSize - RIFF file offset size (in bytes) assumed when
+      *                          being saved to a file
+      */
+     file_offset_t Chunk::RequiredPhysicalSize(int fileOffsetSize) {
+         return CHUNK_HEADER_SIZE(fileOffsetSize) + // RIFF chunk header
+                ullNewChunkSize + // chunks's actual data body
+                ullNewChunkSize % 2; // optional pad byte
      }
      /**
-Line 157 
 namespace RIFF {
+Line 402 
 namespace RIFF {
       *  - RIFF::stream_closed :
       *    the data stream was closed somehow, no more reading possible
       *  - RIFF::stream_end_reached :
-      *    alreaady reached the end of the chunk data, no more reading
+      *    already reached the end of the chunk data, no more reading
       *    possible without SetPos()
+      *
+      *  @see File::IsIOPerThread() for multi-threaded streaming
       */
-     stream_state_t Chunk::GetState() {
+     stream_state_t Chunk::GetState() const {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::GetState()" << std::endl;
+         std::cout << "Chunk::GetState()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
-         #if POSIX
-         if (hFile == 0)        return stream_closed;
+         const File::Handle hRead = pFile->FileHandle();
-         #else
-         if (hFile == NULL)     return stream_closed;
+         if (!_isValidHandle(hRead))
-         #endif // POSIX
+             return stream_closed;
-         if (ulPos < ChunkSize) return stream_ready;
-         else                   return stream_end_reached;
+         const file_offset_t pos = GetPos();
+         if (pos < ullCurrentChunkSize)    return stream_ready;
+         else                              return stream_end_reached;
      }
      /**
-Line 186 
 namespace RIFF {
+Line 435 
 namespace RIFF {
       *  @param WordCount  number of data words to read
       *  @param WordSize   size of each data word to read
       *  @returns          number of successfully read data words or 0 if end
-      *                    of file reached or error occured
+      *                    of file reached or error occurred
+      *  @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::Read(void* pData, unsigned long WordCount, unsigned long WordSize) {
+     file_offset_t Chunk::Read(void* pData, file_offset_t WordCount, file_offset_t WordSize) {
-        #if DEBUG
+         #if DEBUG_RIFF
-        std::cout << "Chunk::Read(void*,ulong,ulong)" << std::endl;
+         std::cout << "Chunk::Read(void*,file_offset_t,file_offset_t)" << std::endl;
-        #endif // DEBUG
+         #endif // DEBUG_RIFF
-         if (ulPos >= ChunkSize) return 0;
+         //if (ulStartPos == 0) return 0; // is only 0 if this is a new chunk, so nothing to read (yet)
-         if (ulPos + WordCount * WordSize >= ChunkSize) WordCount = (ChunkSize - ulPos) / WordSize;
+         const file_offset_t pos = GetPos();
+         if (pos >= ullCurrentChunkSize) return 0;
+         if (pos + WordCount * WordSize >= ullCurrentChunkSize)
+             WordCount = (ullCurrentChunkSize - pos) / WordSize;
+         const File::Handle hRead = pFile->FileHandle();
          #if POSIX
-         if (lseek(hFile, ulStartPos + ulPos, SEEK_SET) < 0) return 0;
+         if (lseek(hRead, ullStartPos + pos, SEEK_SET) < 0) return 0;
-         unsigned long readWords = read(hFile, pData, WordCount * WordSize);
+         ssize_t readWords = read(hRead, pData, WordCount * WordSize);
+         if (readWords < 1) {
+             #if DEBUG_RIFF
+             std::cerr << "POSIX read() failed: " << strerror(errno) << std::endl << std::flush;
+             #endif // DEBUG_RIFF
+             return 0;
+         }
+         readWords /= WordSize;
+         #elif defined(WIN32)
+         LARGE_INTEGER liFilePos;
+         liFilePos.QuadPart = ullStartPos + pos;
+         if (!SetFilePointerEx(hRead, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN))
+             return 0;
+         DWORD readWords;
+         ReadFile(hRead, pData, WordCount * WordSize, &readWords, NULL); //FIXME: does not work for reading buffers larger than 2GB (even though this should rarely be the case in practice)
          if (readWords < 1) return 0;
          readWords /= WordSize;
          #else // standard C functions
-         if (fseek(hFile, ulStartPos + ulPos, SEEK_SET)) return 0;
+         if (fseeko(hRead, ullStartPos + pos, SEEK_SET)) return 0;
-         unsigned long readWords = fread(pData, WordSize, WordCount, hFile);
+         file_offset_t readWords = fread(pData, WordSize, WordCount, hRead);
          #endif // POSIX
-         if (!bEndianNative && WordSize != 1) {
+         if (!pFile->bEndianNative && WordSize != 1) {
              switch (WordSize) {
                  case 2:
-                     for (unsigned long iWord = 0; iWord < readWords; iWord++)
+                     for (file_offset_t iWord = 0; iWord < readWords; iWord++)
                          swapBytes_16((uint16_t*) pData + iWord);
                      break;
                  case 4:
-                     for (unsigned long iWord = 0; iWord < readWords; iWord++)
+                     for (file_offset_t iWord = 0; iWord < readWords; iWord++)
                          swapBytes_32((uint32_t*) pData + iWord);
                      break;
+                 case 8:
+                     for (file_offset_t iWord = 0; iWord < readWords; iWord++)
+                         swapBytes_64((uint64_t*) pData + iWord);
+                     break;
                  default:
-                     for (unsigned long iWord = 0; iWord < readWords; iWord++)
+                     for (file_offset_t iWord = 0; iWord < readWords; iWord++)
                          swapBytes((uint8_t*) pData + iWord * WordSize, WordSize);
                      break;
              }
-Line 223 
 namespace RIFF {
+Line 497 
 namespace RIFF {
          return readWords;
      }
+     /**
+      *  Writes \a WordCount number of data words with given \a WordSize from
+      *  the buffer pointed by \a pData. Be sure to provide the correct
+      *  \a WordSize, as this will be important and taken into account for
+      *  eventual endian correction (swapping of bytes due to different
+      *  native byte order of a system). The position within the chunk will
+      *  automatically be incremented.
+      *
+      *  @param pData      source buffer (containing the data)
+      *  @param WordCount  number of data words to write
+      *  @param WordSize   size of each data word to write
+      *  @returns          number of successfully written data words
+      *  @throws RIFF::Exception  if write operation would exceed current
+      *                           chunk size or any IO error occurred
+      *  @see Resize()
+      *  @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::Write(void* pData, file_offset_t WordCount, file_offset_t WordSize) {
+         const File::HandlePair io = pFile->FileHandlePair();
+         if (io.Mode != stream_mode_read_write)
+             throw Exception("Cannot write data to chunk, file has to be opened in read+write mode first");
+         const file_offset_t pos = GetPos();
+         if (pos >= ullCurrentChunkSize || pos + WordCount * WordSize > ullCurrentChunkSize)
+             throw Exception("End of chunk reached while trying to write data");
+         if (!pFile->bEndianNative && WordSize != 1) {
+             switch (WordSize) {
+                 case 2:
+                     for (file_offset_t iWord = 0; iWord < WordCount; iWord++)
+                         swapBytes_16((uint16_t*) pData + iWord);
+                     break;
+                 case 4:
+                     for (file_offset_t iWord = 0; iWord < WordCount; iWord++)
+                         swapBytes_32((uint32_t*) pData + iWord);
+                     break;
+                 case 8:
+                     for (file_offset_t iWord = 0; iWord < WordCount; iWord++)
+                         swapBytes_64((uint64_t*) pData + iWord);
+                     break;
+                 default:
+                     for (file_offset_t iWord = 0; iWord < WordCount; iWord++)
+                         swapBytes((uint8_t*) pData + iWord * WordSize, WordSize);
+                     break;
+             }
+         }
+         #if POSIX
+         if (lseek(io.hWrite, ullStartPos + pos, SEEK_SET) < 0) {
+             throw Exception("Could not seek to position " + ToString(pos) +
+                             " in chunk (" + ToString(ullStartPos + pos) + " in file)");
+         }
+         ssize_t writtenWords = write(io.hWrite, pData, WordCount * WordSize);
+         if (writtenWords < 1) throw Exception("POSIX IO Error while trying to write chunk data");
+         writtenWords /= WordSize;
+         #elif defined(WIN32)
+         LARGE_INTEGER liFilePos;
+         liFilePos.QuadPart = ullStartPos + pos;
+         if (!SetFilePointerEx(io.hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN)) {
+             throw Exception("Could not seek to position " + ToString(pos) +
+                             " in chunk (" + ToString(ullStartPos + pos) + " in file)");
+         }
+         DWORD writtenWords;
+         WriteFile(io.hWrite, pData, WordCount * WordSize, &writtenWords, NULL); //FIXME: does not work for writing buffers larger than 2GB (even though this should rarely be the case in practice)
+         if (writtenWords < 1) throw Exception("Windows IO Error while trying to write chunk data");
+         writtenWords /= WordSize;
+         #else // standard C functions
+         if (fseeko(io.hWrite, ullStartPos + pos, SEEK_SET)) {
+             throw Exception("Could not seek to position " + ToString(pos) +
+                             " in chunk (" + ToString(ullStartPos + pos) + " in file)");
+         }
+         file_offset_t writtenWords = fwrite(pData, WordSize, WordCount, io.hWrite);
+         #endif // POSIX
+         SetPos(writtenWords * WordSize, stream_curpos);
+         return writtenWords;
+     }
      /** Just an internal wrapper for the main <i>Read()</i> method with additional Exception throwing on errors. */
-     unsigned long Chunk::ReadSceptical(void* pData, unsigned long WordCount, unsigned long WordSize) {
+     file_offset_t Chunk::ReadSceptical(void* pData, file_offset_t WordCount, file_offset_t WordSize) {
-         unsigned long readWords = Read(pData, WordCount, WordSize);
+         file_offset_t readWords = Read(pData, WordCount, WordSize);
          if (readWords != WordCount) throw RIFF::Exception("End of chunk data reached.");
          return readWords;
      }
-Line 238 
 namespace RIFF {
+Line 586 
 namespace RIFF {
       * @param pData             destination buffer
       * @param WordCount         number of 8 Bit signed integers to read
       * @returns                 number of read integers
-      * @throws RIFF::Exception  if an error occured or less than
+      * @throws RIFF::Exception  if an error occurred or less than
       *                          \a WordCount integers could be read!
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::ReadInt8(int8_t* pData, unsigned long WordCount) {
+     file_offset_t Chunk::ReadInt8(int8_t* pData, file_offset_t WordCount) {
-        #if DEBUG
+         #if DEBUG_RIFF
-        std::cout << "Chunk::ReadInt8(int8_t*,ulong)" << std::endl;
+         std::cout << "Chunk::ReadInt8(int8_t*,file_offset_t)" << std::endl;
-        #endif // DEBUG
+         #endif // DEBUG_RIFF
          return ReadSceptical(pData, WordCount, 1);
      }
      /**
+      * Writes \a WordCount number of 8 Bit signed integer words from the
+      * buffer pointed by \a pData to the chunk's body, directly to the
+      * actual "physical" file. The position within the chunk will
+      * automatically be incremented. Note: you cannot write beyond the
+      * boundaries of the chunk, to append data to the chunk call Resize()
+      * before.
+      *
+      * @param pData             source buffer (containing the data)
+      * @param WordCount         number of 8 Bit signed integers to write
+      * @returns                 number of written integers
+      * @throws RIFF::Exception  if an IO error occurred
+      * @see Resize()
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::WriteInt8(int8_t* pData, file_offset_t WordCount) {
+         return Write(pData, WordCount, 1);
+     }
+     /**
       * Reads \a WordCount number of 8 Bit unsigned integer words and copies
       * it into the buffer pointed by \a pData. The buffer has to be
       * allocated. The position within the chunk will automatically be
-Line 257 
 namespace RIFF {
+Line 625 
 namespace RIFF {
       * @param pData             destination buffer
       * @param WordCount         number of 8 Bit unsigned integers to read
       * @returns                 number of read integers
-      * @throws RIFF::Exception  if an error occured or less than
+      * @throws RIFF::Exception  if an error occurred or less than
       *                          \a WordCount integers could be read!
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::ReadUint8(uint8_t* pData, unsigned long WordCount) {
+     file_offset_t Chunk::ReadUint8(uint8_t* pData, file_offset_t WordCount) {
-        #if DEBUG
+         #if DEBUG_RIFF
-        std::cout << "Chunk::ReadUint8(uint8_t*,ulong)" << std::endl;
+         std::cout << "Chunk::ReadUint8(uint8_t*,file_offset_t)" << std::endl;
-        #endif // DEBUG
+         #endif // DEBUG_RIFF
          return ReadSceptical(pData, WordCount, 1);
      }
      /**
+      * Writes \a WordCount number of 8 Bit unsigned integer words from the
+      * buffer pointed by \a pData to the chunk's body, directly to the
+      * actual "physical" file. The position within the chunk will
+      * automatically be incremented. Note: you cannot write beyond the
+      * boundaries of the chunk, to append data to the chunk call Resize()
+      * before.
+      *
+      * @param pData             source buffer (containing the data)
+      * @param WordCount         number of 8 Bit unsigned integers to write
+      * @returns                 number of written integers
+      * @throws RIFF::Exception  if an IO error occurred
+      * @see Resize()
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::WriteUint8(uint8_t* pData, file_offset_t WordCount) {
+         return Write(pData, WordCount, 1);
+     }
+     /**
       * Reads \a WordCount number of 16 Bit signed integer words and copies
       * it into the buffer pointed by \a pData. The buffer has to be
       * allocated. Endian correction will automatically be done if needed.
-Line 276 
 namespace RIFF {
+Line 664 
 namespace RIFF {
       * @param pData             destination buffer
       * @param WordCount         number of 16 Bit signed integers to read
       * @returns                 number of read integers
-      * @throws RIFF::Exception  if an error occured or less than
+      * @throws RIFF::Exception  if an error occurred or less than
       *                          \a WordCount integers could be read!
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::ReadInt16(int16_t* pData, unsigned long WordCount) {
+     file_offset_t Chunk::ReadInt16(int16_t* pData, file_offset_t WordCount) {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadInt16(int16_t*,ulong)" << std::endl;
+         std::cout << "Chunk::ReadInt16(int16_t*,file_offset_t)" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          return ReadSceptical(pData, WordCount, 2);
      }
      /**
+      * Writes \a WordCount number of 16 Bit signed integer words from the
+      * buffer pointed by \a pData to the chunk's body, directly to the
+      * actual "physical" file. The position within the chunk will
+      * automatically be incremented. Note: you cannot write beyond the
+      * boundaries of the chunk, to append data to the chunk call Resize()
+      * before.
+      *
+      * @param pData             source buffer (containing the data)
+      * @param WordCount         number of 16 Bit signed integers to write
+      * @returns                 number of written integers
+      * @throws RIFF::Exception  if an IO error occurred
+      * @see Resize()
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::WriteInt16(int16_t* pData, file_offset_t WordCount) {
+         return Write(pData, WordCount, 2);
+     }
+     /**
       * Reads \a WordCount number of 16 Bit unsigned integer words and copies
       * it into the buffer pointed by \a pData. The buffer has to be
       * allocated. Endian correction will automatically be done if needed.
-Line 295 
 namespace RIFF {
+Line 703 
 namespace RIFF {
       * @param pData             destination buffer
       * @param WordCount         number of 8 Bit unsigned integers to read
       * @returns                 number of read integers
-      * @throws RIFF::Exception  if an error occured or less than
+      * @throws RIFF::Exception  if an error occurred or less than
       *                          \a WordCount integers could be read!
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::ReadUint16(uint16_t* pData, unsigned long WordCount) {
+     file_offset_t Chunk::ReadUint16(uint16_t* pData, file_offset_t WordCount) {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadUint16(uint16_t*,ulong)" << std::endl;
+         std::cout << "Chunk::ReadUint16(uint16_t*,file_offset_t)" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          return ReadSceptical(pData, WordCount, 2);
      }
      /**
+      * Writes \a WordCount number of 16 Bit unsigned integer words from the
+      * buffer pointed by \a pData to the chunk's body, directly to the
+      * actual "physical" file. The position within the chunk will
+      * automatically be incremented. Note: you cannot write beyond the
+      * boundaries of the chunk, to append data to the chunk call Resize()
+      * before.
+      *
+      * @param pData             source buffer (containing the data)
+      * @param WordCount         number of 16 Bit unsigned integers to write
+      * @returns                 number of written integers
+      * @throws RIFF::Exception  if an IO error occurred
+      * @see Resize()
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::WriteUint16(uint16_t* pData, file_offset_t WordCount) {
+         return Write(pData, WordCount, 2);
+     }
+     /**
       * Reads \a WordCount number of 32 Bit signed integer words and copies
       * it into the buffer pointed by \a pData. The buffer has to be
       * allocated. Endian correction will automatically be done if needed.
-Line 314 
 namespace RIFF {
+Line 742 
 namespace RIFF {
       * @param pData             destination buffer
       * @param WordCount         number of 32 Bit signed integers to read
       * @returns                 number of read integers
-      * @throws RIFF::Exception  if an error occured or less than
+      * @throws RIFF::Exception  if an error occurred or less than
       *                          \a WordCount integers could be read!
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::ReadInt32(int32_t* pData, unsigned long WordCount) {
+     file_offset_t Chunk::ReadInt32(int32_t* pData, file_offset_t WordCount) {
-        #if DEBUG
+         #if DEBUG_RIFF
-        std::cout << "Chunk::ReadInt32(int32_t*,ulong)" << std::endl;
+         std::cout << "Chunk::ReadInt32(int32_t*,file_offset_t)" << std::endl;
-        #endif // DEBUG
+         #endif // DEBUG_RIFF
          return ReadSceptical(pData, WordCount, 4);
      }
      /**
+      * Writes \a WordCount number of 32 Bit signed integer words from the
+      * buffer pointed by \a pData to the chunk's body, directly to the
+      * actual "physical" file. The position within the chunk will
+      * automatically be incremented. Note: you cannot write beyond the
+      * boundaries of the chunk, to append data to the chunk call Resize()
+      * before.
+      *
+      * @param pData             source buffer (containing the data)
+      * @param WordCount         number of 32 Bit signed integers to write
+      * @returns                 number of written integers
+      * @throws RIFF::Exception  if an IO error occurred
+      * @see Resize()
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::WriteInt32(int32_t* pData, file_offset_t WordCount) {
+         return Write(pData, WordCount, 4);
+     }
+     /**
       * Reads \a WordCount number of 32 Bit unsigned integer words and copies
       * it into the buffer pointed by \a pData. The buffer has to be
       * allocated. Endian correction will automatically be done if needed.
-Line 333 
 namespace RIFF {
+Line 781 
 namespace RIFF {
       * @param pData             destination buffer
       * @param WordCount         number of 32 Bit unsigned integers to read
       * @returns                 number of read integers
-      * @throws RIFF::Exception  if an error occured or less than
+      * @throws RIFF::Exception  if an error occurred or less than
       *                          \a WordCount integers could be read!
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
-     unsigned long Chunk::ReadUint32(uint32_t* pData, unsigned long WordCount) {
+     file_offset_t Chunk::ReadUint32(uint32_t* pData, file_offset_t WordCount) {
-        #if DEBUG
+         #if DEBUG_RIFF
-        std::cout << "Chunk::ReadUint32(uint32_t*,ulong)" << std::endl;
+         std::cout << "Chunk::ReadUint32(uint32_t*,file_offset_t)" << std::endl;
-        #endif // DEBUG
+         #endif // DEBUG_RIFF
          return ReadSceptical(pData, WordCount, 4);
      }
      /**
+      * Reads a null-padded string of size characters and copies it
+      * into the string \a s. The position within the chunk will
+      * automatically be incremented.
+      *
+      * @param s                 destination string
+      * @param size              number of characters to read
+      * @throws RIFF::Exception  if an error occurred or less than
+      *                          \a size characters could be read!
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     void Chunk::ReadString(String& s, int size) {
+         char* buf = new char[size];
+         ReadSceptical(buf, 1, size);
+         s.assign(buf, std::find(buf, buf + size, '\0'));
+         delete[] buf;
+     }
+     /**
+      * Writes \a WordCount number of 32 Bit unsigned integer words from the
+      * buffer pointed by \a pData to the chunk's body, directly to the
+      * actual "physical" file. The position within the chunk will
+      * automatically be incremented. Note: you cannot write beyond the
+      * boundaries of the chunk, to append data to the chunk call Resize()
+      * before.
+      *
+      * @param pData             source buffer (containing the data)
+      * @param WordCount         number of 32 Bit unsigned integers to write
+      * @returns                 number of written integers
+      * @throws RIFF::Exception  if an IO error occurred
+      * @see Resize()
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::WriteUint32(uint32_t* pData, file_offset_t WordCount) {
+         return Write(pData, WordCount, 4);
+     }
+     /**
       * Reads one 8 Bit signed integer word and increments the position within
       * the chunk.
       *
       * @returns                 read integer word
-      * @throws RIFF::Exception  if an error occured
+      * @throws RIFF::Exception  if an error occurred
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
      int8_t Chunk::ReadInt8() {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadInt8()" << std::endl;
+         std::cout << "Chunk::ReadInt8()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          int8_t word;
          ReadSceptical(&word,1,1);
          return word;
-Line 364 
 namespace RIFF {
+Line 851 
 namespace RIFF {
       * within the chunk.
       *
       * @returns                 read integer word
-      * @throws RIFF::Exception  if an error occured
+      * @throws RIFF::Exception  if an error occurred
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
      uint8_t Chunk::ReadUint8() {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadUint8()" << std::endl;
+         std::cout << "Chunk::ReadUint8()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          uint8_t word;
          ReadSceptical(&word,1,1);
          return word;
-Line 381 
 namespace RIFF {
+Line 869 
 namespace RIFF {
       * needed.
       *
       * @returns                 read integer word
-      * @throws RIFF::Exception  if an error occured
+      * @throws RIFF::Exception  if an error occurred
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
      int16_t Chunk::ReadInt16() {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadInt16()" << std::endl;
+         std::cout << "Chunk::ReadInt16()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          int16_t word;
          ReadSceptical(&word,1,2);
          return word;
-Line 398 
 namespace RIFF {
+Line 887 
 namespace RIFF {
       * needed.
       *
       * @returns                 read integer word
-      * @throws RIFF::Exception  if an error occured
+      * @throws RIFF::Exception  if an error occurred
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
      uint16_t Chunk::ReadUint16() {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadUint16()" << std::endl;
+         std::cout << "Chunk::ReadUint16()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          uint16_t word;
          ReadSceptical(&word,1,2);
          return word;
-Line 415 
 namespace RIFF {
+Line 905 
 namespace RIFF {
       * needed.
       *
       * @returns                 read integer word
-      * @throws RIFF::Exception  if an error occured
+      * @throws RIFF::Exception  if an error occurred
+      * @see File::IsIOPerThread() for multi-threaded streaming
       */
      int32_t Chunk::ReadInt32() {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadInt32()" << std::endl;
+         std::cout << "Chunk::ReadInt32()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          int32_t word;
          ReadSceptical(&word,1,4);
          return word;
-Line 432 
 namespace RIFF {
+Line 923 
 namespace RIFF {
       * needed.
       *
       * @returns                 read integer word
-      * @throws RIFF::Exception  if an error occured
+      * @throws RIFF::Exception  if an error occurred
+      * @see File::IsIOPerThread() for multi-threaded streamings
       */
      uint32_t Chunk::ReadUint32() {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "Chunk::ReadUint32()" << std::endl;
+         std::cout << "Chunk::ReadUint32()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          uint32_t word;
          ReadSceptical(&word,1,4);
          return word;
      }
+     /** @brief Load chunk body into RAM.
+      *
+      * Loads the whole chunk body into memory. You can modify the data in
+      * RAM and save the data by calling File::Save() afterwards.
+      *
+      * <b>Caution:</b> the buffer pointer will be invalidated once
+      * File::Save() was called. You have to call LoadChunkData() again to
+      * get a new, valid pointer whenever File::Save() was called.
+      *
+      * You can call LoadChunkData() again if you previously scheduled to
+      * enlarge this chunk with a Resize() call. In that case the buffer will
+      * be enlarged to the new, scheduled chunk size and you can already
+      * place the new chunk data to the buffer and finally call File::Save()
+      * to enlarge the chunk physically and write the new data in one rush.
+      * This approach is definitely recommended if you have to enlarge and
+      * write new data to a lot of chunks.
+      *
+      * @returns a pointer to the data in RAM on success, NULL otherwise
+      * @throws Exception if data buffer could not be enlarged
+      * @see ReleaseChunkData()
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
      void* Chunk::LoadChunkData() {
-         if (!pChunkData) {
+         if (!pChunkData && pFile->Filename != "" /*&& ulStartPos != 0*/) {
+             File::Handle hRead = pFile->FileHandle();
              #if POSIX
-             if (lseek(hFile, ulStartPos, SEEK_SET) == -1) return NULL;
+             if (lseek(hRead, ullStartPos, SEEK_SET) == -1) return NULL;
-             pChunkData = new uint8_t[GetSize()];
+             #elif defined(WIN32)
-             if (!pChunkData) return NULL;
+             LARGE_INTEGER liFilePos;
-             unsigned long readWords = read(hFile, pChunkData, GetSize());
+             liFilePos.QuadPart = ullStartPos;
+             if (!SetFilePointerEx(hRead, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN)) return NULL;
              #else
-             if (fseek(hFile, ulStartPos, SEEK_SET)) return NULL;
+             if (fseeko(hRead, ullStartPos, SEEK_SET)) return NULL;
-             pChunkData = new uint8_t[GetSize()];
+             #endif // POSIX
+             file_offset_t ullBufferSize = (ullCurrentChunkSize > ullNewChunkSize) ? ullCurrentChunkSize : ullNewChunkSize;
+             pChunkData = new uint8_t[ullBufferSize];
              if (!pChunkData) return NULL;
-             unsigned long readWords = fread(pChunkData, 1, GetSize(), hFile);
+             memset(pChunkData, 0, ullBufferSize);
+             #if POSIX
+             file_offset_t readWords = read(hRead, pChunkData, GetSize());
+             #elif defined(WIN32)
+             DWORD readWords;
+             ReadFile(hRead, pChunkData, GetSize(), &readWords, NULL); //FIXME: won't load chunks larger than 2GB !
+             #else
+             file_offset_t readWords = fread(pChunkData, 1, GetSize(), hRead);
              #endif // POSIX
              if (readWords != GetSize()) {
                  delete[] pChunkData;
                  return (pChunkData = NULL);
              }
+             ullChunkDataSize = ullBufferSize;
+         } else if (ullNewChunkSize > ullChunkDataSize) {
+             uint8_t* pNewBuffer = new uint8_t[ullNewChunkSize];
+             if (!pNewBuffer) throw Exception("Could not enlarge chunk data buffer to " + ToString(ullNewChunkSize) + " bytes");
+             memset(pNewBuffer, 0 , ullNewChunkSize);
+             if (pChunkData) {
+                 memcpy(pNewBuffer, pChunkData, ullChunkDataSize);
+                 delete[] pChunkData;
+             }
+             pChunkData       = pNewBuffer;
+             ullChunkDataSize = ullNewChunkSize;
          }
          return pChunkData;
      }
+     /** @brief Free loaded chunk body from RAM.
+      *
+      * Frees loaded chunk body data from memory (RAM). You should call
+      * File::Save() before calling this method if you modified the data to
+      * make the changes persistent.
+      */
      void Chunk::ReleaseChunkData() {
          if (pChunkData) {
              delete[] pChunkData;
-Line 471 
 namespace RIFF {
+Line 1013 
 namespace RIFF {
          }
      }
+     /** @brief Resize chunk.
+      *
+      * Resizes this chunk's body, that is the actual size of data possible
+      * to be written to this chunk. This call will return immediately and
+      * just schedule the resize operation. You should call File::Save() to
+      * actually perform the resize operation(s) "physically" to the file.
+      * As this can take a while on large files, it is recommended to call
+      * Resize() first on all chunks which have to be resized and finally to
+      * call File::Save() to perform all those resize operations in one rush.
+      *
+      * <b>Caution:</b> You cannot directly write to enlarged chunks before
+      * calling File::Save() as this might exceed the current chunk's body
+      * boundary!
+      *
+      * @param NewSize - new chunk body size in bytes (must be greater than zero)
+      * @throws RIFF::Exception  if \a NewSize is less than 1 or unrealistic large
+      * @see File::Save()
+      */
+     void Chunk::Resize(file_offset_t NewSize) {
+         if (NewSize == 0)
+             throw Exception("There is at least one empty chunk (zero size): " + __resolveChunkPath(this));
+         if ((NewSize >> 48) != 0)
+             throw Exception("Unrealistic high chunk size detected: " + __resolveChunkPath(this));
+         if (ullNewChunkSize == NewSize) return;
+         ullNewChunkSize = NewSize;
+     }
+     /** @brief Write chunk persistently e.g. to disk.
+      *
+      * Stores the chunk persistently to its actual "physical" file.
+      *
+      * @param ullWritePos - position within the "physical" file where this
+      *                     chunk should be written to
+      * @param ullCurrentDataOffset - offset of current (old) data within
+      *                              the file
+      * @param pProgress - optional: callback function for progress notification
+      * @returns new write position in the "physical" file, that is
+      *          \a ullWritePos incremented by this chunk's new size
+      *          (including its header size of course)
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     file_offset_t Chunk::WriteChunk(file_offset_t ullWritePos, file_offset_t ullCurrentDataOffset, progress_t* pProgress) {
+         const file_offset_t ullOriginalPos = ullWritePos;
+         ullWritePos += CHUNK_HEADER_SIZE(pFile->FileOffsetSize);
+         const File::HandlePair io = pFile->FileHandlePair();
+         if (io.Mode != stream_mode_read_write)
+             throw Exception("Cannot write list chunk, file has to be opened in read+write mode");
+         // if the whole chunk body was loaded into RAM
+         if (pChunkData) {
+             // make sure chunk data buffer in RAM is at least as large as the new chunk size
+             LoadChunkData();
+             // write chunk data from RAM persistently to the file
+             #if POSIX
+             lseek(io.hWrite, ullWritePos, SEEK_SET);
+             if (write(io.hWrite, pChunkData, ullNewChunkSize) != ullNewChunkSize) {
+                 throw Exception("Writing Chunk data (from RAM) failed");
+             }
+             #elif defined(WIN32)
+             LARGE_INTEGER liFilePos;
+             liFilePos.QuadPart = ullWritePos;
+             SetFilePointerEx(io.hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+             DWORD dwBytesWritten;
+             WriteFile(io.hWrite, pChunkData, ullNewChunkSize, &dwBytesWritten, NULL); //FIXME: won't save chunks larger than 2GB !
+             if (dwBytesWritten != ullNewChunkSize) {
+                 throw Exception("Writing Chunk data (from RAM) failed");
+             }
+             #else
+             fseeko(io.hWrite, ullWritePos, SEEK_SET);
+             if (fwrite(pChunkData, 1, ullNewChunkSize, io.hWrite) != ullNewChunkSize) {
+                 throw Exception("Writing Chunk data (from RAM) failed");
+             }
+             #endif // POSIX
+         } else {
+             // move chunk data from the end of the file to the appropriate position
+             int8_t* pCopyBuffer = new int8_t[4096];
+             file_offset_t ullToMove = (ullNewChunkSize < ullCurrentChunkSize) ? ullNewChunkSize : ullCurrentChunkSize;
+             #if defined(WIN32)
+             DWORD iBytesMoved = 1; // we have to pass it via pointer to the Windows API, thus the correct size must be ensured
+             #else
+             int iBytesMoved = 1;
+             #endif
+             for (file_offset_t ullOffset = 0; ullToMove > 0 && iBytesMoved > 0; ullOffset += iBytesMoved, ullToMove -= iBytesMoved) {
+                 iBytesMoved = (ullToMove < 4096) ? int(ullToMove) : 4096;
+                 #if POSIX
+                 lseek(io.hRead, ullStartPos + ullCurrentDataOffset + ullOffset, SEEK_SET);
+                 iBytesMoved = (int) read(io.hRead, pCopyBuffer, (size_t) iBytesMoved);
+                 lseek(io.hWrite, ullWritePos + ullOffset, SEEK_SET);
+                 iBytesMoved = (int) write(io.hWrite, pCopyBuffer, (size_t) iBytesMoved);
+                 #elif defined(WIN32)
+                 LARGE_INTEGER liFilePos;
+                 liFilePos.QuadPart = ullStartPos + ullCurrentDataOffset + ullOffset;
+                 SetFilePointerEx(io.hRead, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+                 ReadFile(io.hRead, pCopyBuffer, iBytesMoved, &iBytesMoved, NULL);
+                 liFilePos.QuadPart = ullWritePos + ullOffset;
+                 SetFilePointerEx(io.hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+                 WriteFile(io.hWrite, pCopyBuffer, iBytesMoved, &iBytesMoved, NULL);
+                 #else
+                 fseeko(io.hRead, ullStartPos + ullCurrentDataOffset + ullOffset, SEEK_SET);
+                 iBytesMoved = fread(pCopyBuffer, 1, iBytesMoved, io.hRead);
+                 fseeko(io.hWrite, ullWritePos + ullOffset, SEEK_SET);
+                 iBytesMoved = fwrite(pCopyBuffer, 1, iBytesMoved, io.hWrite);
+                 #endif
+             }
+             delete[] pCopyBuffer;
+             if (iBytesMoved < 0) throw Exception("Writing Chunk data (from file) failed");
+         }
+         // update this chunk's header
+         ullCurrentChunkSize = ullNewChunkSize;
+         WriteHeader(ullOriginalPos);
+         if (pProgress)
+             __notify_progress(pProgress, 1.0); // notify done
+         // update chunk's position pointers
+         ullStartPos = ullOriginalPos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize);
+         Chunk::__resetPos();
+         // add pad byte if needed
+         if ((ullStartPos + ullNewChunkSize) % 2 != 0) {
+             const char cPadByte = 0;
+             #if POSIX
+             lseek(io.hWrite, ullStartPos + ullNewChunkSize, SEEK_SET);
+             write(io.hWrite, &cPadByte, 1);
+             #elif defined(WIN32)
+             LARGE_INTEGER liFilePos;
+             liFilePos.QuadPart = ullStartPos + ullNewChunkSize;
+             SetFilePointerEx(io.hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+             DWORD dwBytesWritten;
+             WriteFile(io.hWrite, &cPadByte, 1, &dwBytesWritten, NULL);
+             #else
+             fseeko(io.hWrite, ullStartPos + ullNewChunkSize, SEEK_SET);
+             fwrite(&cPadByte, 1, 1, io.hWrite);
+             #endif
+             return ullStartPos + ullNewChunkSize + 1;
+         }
+         return ullStartPos + ullNewChunkSize;
+     }
+     void Chunk::__resetPos() {
+         std::lock_guard<std::mutex> lock(chunkPos.mutex);
+         chunkPos.ullPos = 0;
+         chunkPos.byThread.clear();
+     }
  // *************** List ***************
  // *
-     List::List() : Chunk() {
+     List::List(File* pFile) : Chunk(pFile) {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "List::List()" << std::endl;
+         std::cout << "List::List(File* pFile)" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          pSubChunks    = NULL;
          pSubChunksMap = NULL;
      }
-     #if POSIX
+     List::List(File* pFile, file_offset_t StartPos, List* Parent)
-     List::List(int hFile, unsigned long StartPos, bool EndianNative, List* Parent)
+       : Chunk(pFile, StartPos, Parent) {
-     #else
+         #if DEBUG_RIFF
-     List::List(FILE* hFile, unsigned long StartPos, bool EndianNative, List* Parent)
+         std::cout << "List::List(File*,file_offset_t,List*)" << std::endl;
-     #endif // POSIX
+         #endif // DEBUG_RIFF
-       : Chunk(hFile, StartPos, EndianNative, Parent) {
-         #if DEBUG
-         std::cout << "List::List(FILE*,ulong,bool,List*)" << std::endl;
-         #endif // DEBUG
          pSubChunks    = NULL;
          pSubChunksMap = NULL;
          ReadHeader(StartPos);
-         ulStartPos    = StartPos + LIST_HEADER_SIZE;
+         ullStartPos = StartPos + LIST_HEADER_SIZE(pFile->FileOffsetSize);
+     }
+     List::List(File* pFile, List* pParent, uint32_t uiListID)
+       : Chunk(pFile, pParent, CHUNK_ID_LIST, 0) {
+         pSubChunks    = NULL;
+         pSubChunksMap = NULL;
+         ListType      = uiListID;
      }
      List::~List() {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "List::~List()" << std::endl;
+         std::cout << "List::~List()" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
+         DeleteChunkList();
+     }
+     void List::DeleteChunkList() {
          if (pSubChunks) {
              ChunkList::iterator iter = pSubChunks->begin();
              ChunkList::iterator end  = pSubChunks->end();
-Line 511 
 namespace RIFF {
+Line 1209 
 namespace RIFF {
                  iter++;
              }
              delete pSubChunks;
+             pSubChunks = NULL;
+         }
+         if (pSubChunksMap) {
+             delete pSubChunksMap;
+             pSubChunksMap = NULL;
          }
-         if (pSubChunksMap) delete pSubChunksMap;
+     }
+     /**
+      *  Returns subchunk at supplied @a pos position within this chunk list.
+      *  If supplied @a pos is out of bounds then @c NULL is returned. The
+      *  returned subchunk can either by an ordinary data chunk or a list chunk.
+      *
+      *  @param pos - position of sought subchunk within this list
+      *  @returns pointer to the subchunk or NULL if supplied position is not
+      *           within the valid range of this list
+      */
+     Chunk* List::GetSubChunkAt(size_t pos) {
+         if (!pSubChunks) LoadSubChunks();
+         if (pos >= pSubChunks->size()) return NULL;
+         return (*pSubChunks)[pos];
      }
      /**
-Line 527 
 namespace RIFF {
+Line 1244 
 namespace RIFF {
       *                   that ID
       */
      Chunk* List::GetSubChunk(uint32_t ChunkID) {
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "List::GetSubChunk(uint32_t)" << std::endl;
+         std::cout << "List::GetSubChunk(uint32_t)" << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
          if (!pSubChunksMap) LoadSubChunks();
          return (*pSubChunksMap)[ChunkID];
      }
      /**
+      *  Returns sublist chunk with list type <i>\a ListType</i> at supplied
+      *  @a pos position among all subchunks of type <i>\a ListType</i> within
+      *  this chunk list. If supplied @a pos is out of bounds then @c NULL is
+      *  returned.
+      *
+      *  @param pos - position of sought sublist within this list
+      *  @returns pointer to the sublist or NULL if if supplied position is not
+      *           within valid range
+      */
+     List* List::GetSubListAt(size_t pos) {
+         if (!pSubChunks) LoadSubChunks();
+         if (pos >= pSubChunks->size()) return NULL;
+         for (size_t iCk = 0, iLst = 0; iCk < pSubChunks->size(); ++iCk) {
+             Chunk* pChunk = (*pSubChunks)[iCk];
+             if (pChunk->GetChunkID() != CHUNK_ID_LIST) continue;
+             if (iLst == pos) return (List*) pChunk;
+             ++iLst;
+         }
+         return NULL;
+     }
+     /**
       *  Returns sublist chunk with list type <i>\a ListType</i> within this
       *  chunk list. Use this method if you expect only one sublist chunk of
-      *  that type in the list. It there are more than one, it's undetermined
+      *  that type in the list. If there are more than one, it's undetermined
       *  which one of them will be returned! If there are no sublists with
       *  that desired list type, NULL will be returned.
       *
-Line 546 
 namespace RIFF {
+Line 1285 
 namespace RIFF {
       *                    that type
       */
      List* List::GetSubList(uint32_t ListType) {
-         #if DEBUG
+         #if DEBUG_RIFF
          std::cout << "List::GetSubList(uint32_t)" << std::endl;
-         #endif // DEBUG
+         #endif // DEBUG_RIFF
          if (!pSubChunks) LoadSubChunks();
          ChunkList::iterator iter = pSubChunks->begin();
          ChunkList::iterator end  = pSubChunks->end();
-Line 563 
 namespace RIFF {
+Line 1302 
 namespace RIFF {
      }
      /**
-      *  Returns the first subchunk within the list. You have to call this
+      *  Returns the first subchunk within the list (which may be an ordinary
+      *  chunk as well as a list chunk). You have to call this
       *  method before you can call GetNextSubChunk(). Recall it when you want
       *  to start from the beginning of the list again.
       *
       *  @returns  pointer to the first subchunk within the list, NULL
       *            otherwise
+      *  @deprecated  This method is not reentrant-safe, use GetSubChunkAt()
+      *               instead.
       */
      Chunk* List::GetFirstSubChunk() {
-         #if DEBUG
+         #if DEBUG_RIFF
          std::cout << "List::GetFirstSubChunk()" << std::endl;
-         #endif // DEBUG
+         #endif // DEBUG_RIFF
          if (!pSubChunks) LoadSubChunks();
          ChunksIterator = pSubChunks->begin();
          return (ChunksIterator != pSubChunks->end()) ? *ChunksIterator : NULL;
      }
      /**
-      *  Returns the next subchunk within the list. You have to call
+      *  Returns the next subchunk within the list (which may be an ordinary
+      *  chunk as well as a list chunk). You have to call
       *  GetFirstSubChunk() before you can use this method!
       *
       *  @returns  pointer to the next subchunk within the list or NULL if
       *            end of list is reached
+      *  @deprecated  This method is not reentrant-safe, use GetSubChunkAt()
+      *               instead.
       */
      Chunk* List::GetNextSubChunk() {
-         #if DEBUG
+         #if DEBUG_RIFF
          std::cout << "List::GetNextSubChunk()" << std::endl;
-         #endif // DEBUG
+         #endif // DEBUG_RIFF
          if (!pSubChunks) return NULL;
          ChunksIterator++;
          return (ChunksIterator != pSubChunks->end()) ? *ChunksIterator : NULL;
-Line 603 
 namespace RIFF {
+Line 1348 
 namespace RIFF {
       *
       *  @returns  pointer to the first sublist within the list, NULL
       *            otherwise
+      *  @deprecated  This method is not reentrant-safe, use GetSubListAt()
+      *               instead.
       */
      List* List::GetFirstSubList() {
-         #if DEBUG
+         #if DEBUG_RIFF
          std::cout << "List::GetFirstSubList()" << std::endl;
-         #endif // DEBUG
+         #endif // DEBUG_RIFF
          if (!pSubChunks) LoadSubChunks();
          ListIterator            = pSubChunks->begin();
          ChunkList::iterator end = pSubChunks->end();
-Line 625 
 namespace RIFF {
+Line 1372 
 namespace RIFF {
       *
       *  @returns  pointer to the next sublist within the list, NULL if
       *            end of list is reached
+      *  @deprecated  This method is not reentrant-safe, use GetSubListAt()
+      *               instead.
       */
      List* List::GetNextSubList() {
-         #if DEBUG
+         #if DEBUG_RIFF
          std::cout << "List::GetNextSubList()" << std::endl;
-         #endif // DEBUG
+         #endif // DEBUG_RIFF
          if (!pSubChunks) return NULL;
          if (ListIterator == pSubChunks->end()) return NULL;
          ListIterator++;
-Line 642 
 namespace RIFF {
+Line 1391 
 namespace RIFF {
      }
      /**
-      *  Returns number of subchunks within the list.
+      *  Returns number of subchunks within the list (including list chunks).
       */
-     unsigned int List::CountSubChunks() {
+     size_t List::CountSubChunks() {
          if (!pSubChunks) LoadSubChunks();
          return pSubChunks->size();
      }
-Line 653 
 namespace RIFF {
+Line 1402 
 namespace RIFF {
       *  Returns number of subchunks within the list with chunk ID
       *  <i>\a ChunkId</i>.
       */
-     unsigned int List::CountSubChunks(uint32_t ChunkID) {
+     size_t List::CountSubChunks(uint32_t ChunkID) {
-         unsigned int result = 0;
+         size_t result = 0;
          if (!pSubChunks) LoadSubChunks();
          ChunkList::iterator iter = pSubChunks->begin();
          ChunkList::iterator end  = pSubChunks->end();
-Line 670 
 namespace RIFF {
+Line 1419 
 namespace RIFF {
      /**
       *  Returns number of sublists within the list.
       */
-     unsigned int List::CountSubLists() {
+     size_t List::CountSubLists() {
          return CountSubChunks(CHUNK_ID_LIST);
      }
-Line 678 
 namespace RIFF {
+Line 1427 
 namespace RIFF {
       *  Returns number of sublists within the list with list type
       *  <i>\a ListType</i>
       */
-     unsigned int List::CountSubLists(uint32_t ListType) {
+     size_t List::CountSubLists(uint32_t ListType) {
-         unsigned int result = 0;
+         size_t result = 0;
          if (!pSubChunks) LoadSubChunks();
          ChunkList::iterator iter = pSubChunks->begin();
          ChunkList::iterator end  = pSubChunks->end();
-Line 693 
 namespace RIFF {
+Line 1442 
 namespace RIFF {
          return result;
      }
-     void List::ReadHeader(unsigned long fPos) {
+     /** @brief Creates a new sub chunk.
-       #if DEBUG
+      *
-       std::cout << "List::Readheader(ulong) ";
+      * Creates and adds a new sub chunk to this list chunk. Note that the
-       #endif // DEBUG
+      * chunk's body size given by \a ullBodySize must be greater than zero.
-         Chunk::ReadHeader(fPos);
+      * You have to call File::Save() to make this change persistent to the
-         ChunkSize -= 4;
+      * actual file and <b>before</b> performing any data write operations
+      * on the new chunk!
+      *
+      * @param uiChunkID  - chunk ID of the new chunk
+      * @param ullBodySize - size of the new chunk's body, that is its actual
+      *                      data size (without header)
+      * @throws RIFF::Exception if \a ullBodySize equals zero
+      */
+     Chunk* List::AddSubChunk(uint32_t uiChunkID, file_offset_t ullBodySize) {
+         if (ullBodySize == 0) throw Exception("Chunk body size must be at least 1 byte");
+         if (!pSubChunks) LoadSubChunks();
+         Chunk* pNewChunk = new Chunk(pFile, this, uiChunkID, 0);
+         pSubChunks->push_back(pNewChunk);
+         (*pSubChunksMap)[uiChunkID] = pNewChunk;
+         pNewChunk->Resize(ullBodySize);
+         ullNewChunkSize += CHUNK_HEADER_SIZE(pFile->FileOffsetSize);
+         return pNewChunk;
+     }
+     /** @brief Moves a sub chunk witin this list.
+      *
+      * Moves a sub chunk from one position in this list to another
+      * position in the same list. The pSrc chunk is placed before the
+      * pDst chunk.
+      *
+      * @param pSrc - sub chunk to be moved
+      * @param pDst - the position to move to. pSrc will be placed
+      *               before pDst. If pDst is 0, pSrc will be placed
+      *               last in list.
+      */
+     void List::MoveSubChunk(Chunk* pSrc, Chunk* pDst) {
+         if (!pSubChunks) LoadSubChunks();
+         for (size_t i = 0; i < pSubChunks->size(); ++i) {
+             if ((*pSubChunks)[i] == pSrc) {
+                 pSubChunks->erase(pSubChunks->begin() + i);
+                 ChunkList::iterator iter =
+                     find(pSubChunks->begin(), pSubChunks->end(), pDst);
+                 pSubChunks->insert(iter, pSrc);
+                 return;
+             }
+         }
+     }
+     /** @brief Moves a sub chunk from this list to another list.
+      *
+      * Moves a sub chunk from this list list to the end of another
+      * list.
+      *
+      * @param pSrc - sub chunk to be moved
+      * @param pDst - destination list where the chunk shall be moved to
+      */
+     void List::MoveSubChunk(Chunk* pSrc, List* pNewParent) {
+         if (pNewParent == this || !pNewParent) return;
+         if (!pSubChunks) LoadSubChunks();
+         if (!pNewParent->pSubChunks) pNewParent->LoadSubChunks();
+         ChunkList::iterator iter =
+             find(pSubChunks->begin(), pSubChunks->end(), pSrc);
+         if (iter == pSubChunks->end()) return;
+         pSubChunks->erase(iter);
+         pNewParent->pSubChunks->push_back(pSrc);
+         // update chunk id map of this List
+         if ((*pSubChunksMap)[pSrc->GetChunkID()] == pSrc) {
+             pSubChunksMap->erase(pSrc->GetChunkID());
+             // try to find another chunk of the same chunk ID
+             ChunkList::iterator iter = pSubChunks->begin();
+             ChunkList::iterator end  = pSubChunks->end();
+             for (; iter != end; ++iter) {
+                 if ((*iter)->GetChunkID() == pSrc->GetChunkID()) {
+                     (*pSubChunksMap)[pSrc->GetChunkID()] = *iter;
+                     break; // we're done, stop search
+                 }
+             }
+         }
+         // update chunk id map of other list
+         if (!(*pNewParent->pSubChunksMap)[pSrc->GetChunkID()])
+             (*pNewParent->pSubChunksMap)[pSrc->GetChunkID()] = pSrc;
+     }
+     /** @brief Creates a new list sub chunk.
+      *
+      * Creates and adds a new list sub chunk to this list chunk. Note that
+      * you have to add sub chunks / sub list chunks to the new created chunk
+      * <b>before</b> trying to make this change persisten to the actual
+      * file with File::Save()!
+      *
+      * @param uiListType - list ID of the new list chunk
+      */
+     List* List::AddSubList(uint32_t uiListType) {
+         if (!pSubChunks) LoadSubChunks();
+         List* pNewListChunk = new List(pFile, this, uiListType);
+         pSubChunks->push_back(pNewListChunk);
+         (*pSubChunksMap)[CHUNK_ID_LIST] = pNewListChunk;
+         ullNewChunkSize += LIST_HEADER_SIZE(pFile->FileOffsetSize);
+         return pNewListChunk;
+     }
+     /** @brief Removes a sub chunk.
+      *
+      * Removes the sub chunk given by \a pSubChunk from this list and frees
+      * it completely from RAM. The given chunk can either be a normal sub
+      * chunk or a list sub chunk. In case the given chunk is a list chunk,
+      * all its subchunks (if any) will be removed recursively as well. You
+      * should call File::Save() to make this change persistent at any time.
+      *
+      * @param pSubChunk - sub chunk or sub list chunk to be removed
+      */
+     void List::DeleteSubChunk(Chunk* pSubChunk) {
+         if (!pSubChunks) LoadSubChunks();
+         ChunkList::iterator iter =
+             find(pSubChunks->begin(), pSubChunks->end(), pSubChunk);
+         if (iter == pSubChunks->end()) return;
+         pSubChunks->erase(iter);
+         if ((*pSubChunksMap)[pSubChunk->GetChunkID()] == pSubChunk) {
+             pSubChunksMap->erase(pSubChunk->GetChunkID());
+             // try to find another chunk of the same chunk ID
+             ChunkList::iterator iter = pSubChunks->begin();
+             ChunkList::iterator end  = pSubChunks->end();
+             for (; iter != end; ++iter) {
+                 if ((*iter)->GetChunkID() == pSubChunk->GetChunkID()) {
+                     (*pSubChunksMap)[pSubChunk->GetChunkID()] = *iter;
+                     break; // we're done, stop search
+                 }
+             }
+         }
+         delete pSubChunk;
+     }
+     /**
+      *  Returns the actual total size in bytes (including List chunk header and
+      *  all subchunks) of this List Chunk if being stored to a file.
+      *
+      *  @param fileOffsetSize - RIFF file offset size (in bytes) assumed when
+      *                          being saved to a file
+      */
+     file_offset_t List::RequiredPhysicalSize(int fileOffsetSize) {
+         if (!pSubChunks) LoadSubChunks();
+         file_offset_t size = LIST_HEADER_SIZE(fileOffsetSize);
+         ChunkList::iterator iter = pSubChunks->begin();
+         ChunkList::iterator end  = pSubChunks->end();
+         for (; iter != end; ++iter)
+             size += (*iter)->RequiredPhysicalSize(fileOffsetSize);
+         return size;
+     }
+     void List::ReadHeader(file_offset_t filePos) {
+         #if DEBUG_RIFF
+         std::cout << "List::Readheader(file_offset_t) ";
+         #endif // DEBUG_RIFF
+         Chunk::ReadHeader(filePos);
+         if (ullCurrentChunkSize < 4) return;
+         ullNewChunkSize = ullCurrentChunkSize -= 4;
+         const File::Handle hRead = pFile->FileHandle();
          #if POSIX
-         lseek(hFile, fPos + CHUNK_HEADER_SIZE, SEEK_SET);
+         lseek(hRead, filePos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize), SEEK_SET);
-         read(hFile, &ListType, 4);
+         read(hRead, &ListType, 4);
+         #elif defined(WIN32)
+         LARGE_INTEGER liFilePos;
+         liFilePos.QuadPart = filePos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize);
+         SetFilePointerEx(hRead, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+         DWORD dwBytesRead;
+         ReadFile(hRead, &ListType, 4, &dwBytesRead, NULL);
          #else
-         fseek(hFile, fPos + CHUNK_HEADER_SIZE, SEEK_SET);
+         fseeko(hRead, filePos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize), SEEK_SET);
-         fread(&ListType, 4, 1, hFile);
+         fread(&ListType, 4, 1, hRead);
          #endif // POSIX
-       #if DEBUG
+         #if DEBUG_RIFF
-       std::cout << "listType=" << convertToString(ListType) << std::endl;
+         std::cout << "listType=" << convertToString(ListType) << std::endl;
-       #endif // DEBUG
+         #endif // DEBUG_RIFF
-         if (!bEndianNative) {
+         if (!pFile->bEndianNative) {
              //swapBytes_32(&ListType);
          }
      }
-     void List::LoadSubChunks() {
+     void List::WriteHeader(file_offset_t filePos) {
-        #if DEBUG
+         // the four list type bytes officially belong the chunk's body in the RIFF format
-        std::cout << "List::LoadSubChunks()";
+         ullNewChunkSize += 4;
-        #endif // DEBUG
+         Chunk::WriteHeader(filePos);
+         ullNewChunkSize -= 4; // just revert the +4 incrementation
+         const File::Handle hWrite = pFile->FileWriteHandle();
+         #if POSIX
+         lseek(hWrite, filePos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize), SEEK_SET);
+         write(hWrite, &ListType, 4);
+         #elif defined(WIN32)
+         LARGE_INTEGER liFilePos;
+         liFilePos.QuadPart = filePos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize);
+         SetFilePointerEx(hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+         DWORD dwBytesWritten;
+         WriteFile(hWrite, &ListType, 4, &dwBytesWritten, NULL);
+         #else
+         fseeko(hWrite, filePos + CHUNK_HEADER_SIZE(pFile->FileOffsetSize), SEEK_SET);
+         fwrite(&ListType, 4, 1, hWrite);
+         #endif // POSIX
+     }
+     void List::LoadSubChunks(progress_t* pProgress) {
+         #if DEBUG_RIFF
+         std::cout << "List::LoadSubChunks()";
+         #endif // DEBUG_RIFF
          if (!pSubChunks) {
              pSubChunks    = new ChunkList();
              pSubChunksMap = new ChunkMap();
-             unsigned long uiOriginalPos = GetPos();
+             const File::Handle hRead = pFile->FileHandle();
+             if (!_isValidHandle(hRead)) return;
+             const file_offset_t ullOriginalPos = GetPos();
              SetPos(0); // jump to beginning of list chunk body
-             while (RemainingBytes() >= CHUNK_HEADER_SIZE) {
+             while (RemainingBytes() >= CHUNK_HEADER_SIZE(pFile->FileOffsetSize)) {
                  Chunk* ck;
                  uint32_t ckid;
-                 Read(&ckid, 4, 1);
+                 // return value check is required here to prevent a potential
-        #if DEBUG
+                 // garbage data use of 'ckid' below in case Read() failed
-        std::cout << " ckid=" << convertToString(ckid) << std::endl;
+                 if (Read(&ckid, 4, 1) != 4)
-        #endif // DEBUG
+                     throw Exception("LoadSubChunks(): Failed reading RIFF chunk ID");
+                 #if DEBUG_RIFF
+                 std::cout << " ckid=" << convertToString(ckid) << std::endl;
+                 #endif // DEBUG_RIFF
+                 const file_offset_t pos = GetPos();
                  if (ckid == CHUNK_ID_LIST) {
-                     ck = new RIFF::List(hFile, ulStartPos + ulPos - 4, bEndianNative, this);
+                     ck = new RIFF::List(pFile, ullStartPos + pos - 4, this);
-                     SetPos(ck->GetSize() + LIST_HEADER_SIZE - 4, RIFF::stream_curpos);
+                     SetPos(ck->GetSize() + LIST_HEADER_SIZE(pFile->FileOffsetSize) - 4, RIFF::stream_curpos);
                  }
                  else { // simple chunk
-                     ck = new RIFF::Chunk(hFile, ulStartPos + ulPos - 4, bEndianNative, this);
+                     ck = new RIFF::Chunk(pFile, ullStartPos + pos - 4, this);
-                     SetPos(ck->GetSize() + CHUNK_HEADER_SIZE - 4, RIFF::stream_curpos);
+                     SetPos(ck->GetSize() + CHUNK_HEADER_SIZE(pFile->FileOffsetSize) - 4, RIFF::stream_curpos);
                  }
                  pSubChunks->push_back(ck);
                  (*pSubChunksMap)[ckid] = ck;
                  if (GetPos() % 2 != 0) SetPos(1, RIFF::stream_curpos); // jump over pad byte
              }
-             SetPos(uiOriginalPos); // restore position before this call
+             SetPos(ullOriginalPos); // restore position before this call
+         }
+         if (pProgress)
+             __notify_progress(pProgress, 1.0); // notify done
+     }
+     void List::LoadSubChunksRecursively(progress_t* pProgress) {
+         const int n = (int) CountSubLists();
+         int i = 0;
+         for (List* pList = GetSubListAt(i); pList; pList = GetSubListAt(++i)) {
+             if (pProgress) {
+                 // divide local progress into subprogress
+                 progress_t subprogress;
+                 __divide_progress(pProgress, &subprogress, n, i);
+                 // do the actual work
+                 pList->LoadSubChunksRecursively(&subprogress);
+             } else
+                 pList->LoadSubChunksRecursively(NULL);
+         }
+         if (pProgress)
+             __notify_progress(pProgress, 1.0); // notify done
+     }
+     /** @brief Write list chunk persistently e.g. to disk.
+      *
+      * Stores the list chunk persistently to its actual "physical" file. All
+      * subchunks (including sub list chunks) will be stored recursively as
+      * well.
+      *
+      * @param ullWritePos - position within the "physical" file where this
+      *                     list chunk should be written to
+      * @param ullCurrentDataOffset - offset of current (old) data within
+      *                              the file
+      * @param pProgress - optional: callback function for progress notification
+      * @returns new write position in the "physical" file, that is
+      *          \a ullWritePos incremented by this list chunk's new size
+      *          (including its header size of course)
+      */
+     file_offset_t List::WriteChunk(file_offset_t ullWritePos, file_offset_t ullCurrentDataOffset, progress_t* pProgress) {
+         const file_offset_t ullOriginalPos = ullWritePos;
+         ullWritePos += LIST_HEADER_SIZE(pFile->FileOffsetSize);
+         if (pFile->GetMode() != stream_mode_read_write)
+             throw Exception("Cannot write list chunk, file has to be opened in read+write mode");
+         // write all subchunks (including sub list chunks) recursively
+         if (pSubChunks) {
+             size_t i = 0;
+             const size_t n = pSubChunks->size();
+             for (ChunkList::iterator iter = pSubChunks->begin(), end = pSubChunks->end(); iter != end; ++iter, ++i) {
+                 if (pProgress) {
+                     // divide local progress into subprogress for loading current Instrument
+                     progress_t subprogress;
+                     __divide_progress(pProgress, &subprogress, n, i);
+                     // do the actual work
+                     ullWritePos = (*iter)->WriteChunk(ullWritePos, ullCurrentDataOffset, &subprogress);
+                 } else
+                     ullWritePos = (*iter)->WriteChunk(ullWritePos, ullCurrentDataOffset, NULL);
+             }
+         }
+         // update this list chunk's header
+         ullCurrentChunkSize = ullNewChunkSize = ullWritePos - ullOriginalPos - LIST_HEADER_SIZE(pFile->FileOffsetSize);
+         WriteHeader(ullOriginalPos);
+         // offset of this list chunk in new written file may have changed
+         ullStartPos = ullOriginalPos + LIST_HEADER_SIZE(pFile->FileOffsetSize);
+         if (pProgress)
+             __notify_progress(pProgress, 1.0); // notify done
+         return ullWritePos;
+     }
+     void List::__resetPos() {
+         Chunk::__resetPos();
+         if (pSubChunks) {
+             for (ChunkList::iterator iter = pSubChunks->begin(), end = pSubChunks->end(); iter != end; ++iter) {
+                 (*iter)->__resetPos();
+             }
          }
      }
      /**
       *  Returns string representation of the lists's id
       */
-     String List::GetListTypeString() {
+     String List::GetListTypeString() const {
          return convertToString(ListType);
      }
-Line 758 
 namespace RIFF {
+Line 1776 
 namespace RIFF {
  // *************** File ***************
  // *
-     File::File(const String& path) : List(), Filename(path) {
+     /** @brief Create new RIFF file.
-       #if DEBUG
+      *
-       std::cout << "File::File("<<path<<")" << std::endl;
+      * Use this constructor if you want to create a new RIFF file completely
-       #endif // DEBUG
+      * "from scratch". Note: there must be no empty chunks or empty list
+      * chunks when trying to make the new RIFF file persistent with Save()!
+      *
+      * Note: by default, the RIFF file will be saved in native endian
+      * format; that is, as a RIFF file on little-endian machines and
+      * as a RIFX file on big-endian. To change this behaviour, call
+      * SetByteOrder() before calling Save().
+      *
+      * @param FileType - four-byte identifier of the RIFF file type
+      * @see AddSubChunk(), AddSubList(), SetByteOrder()
+      */
+     File::File(uint32_t FileType)
+         : List(this), bIsNewFile(true), Layout(layout_standard),
+           FileOffsetPreference(offset_size_auto)
+     {
+         io.isPerThread = false;
+         #if defined(WIN32)
+         io.hRead = io.hWrite = INVALID_HANDLE_VALUE;
+         #else
+         io.hRead = io.hWrite = 0;
+         #endif
+         io.Mode = stream_mode_closed;
+         bEndianNative = true;
+         ListType = FileType;
+         FileOffsetSize = 4;
+         ullStartPos = RIFF_HEADER_SIZE(FileOffsetSize);
+     }
+     /** @brief Load existing RIFF file.
+      *
+      * Loads an existing RIFF file with all its chunks.
+      *
+      * @param path - path and file name of the RIFF file to open
+      * @throws RIFF::Exception if error occurred while trying to load the
+      *                         given RIFF file
+      */
+     File::File(const String& path)
+         : List(this), Filename(path), bIsNewFile(false), Layout(layout_standard),
+           FileOffsetPreference(offset_size_auto)
+     {
+         #if DEBUG_RIFF
+         std::cout << "File::File("<<path<<")" << std::endl;
+         #endif // DEBUG_RIFF
          bEndianNative = true;
+         FileOffsetSize = 4;
+         try {
+             __openExistingFile(path);
+             if (ChunkID != CHUNK_ID_RIFF && ChunkID != CHUNK_ID_RIFX) {
+                 throw RIFF::Exception("Not a RIFF file");
+             }
+         }
+         catch (...) {
+             Cleanup();
+             throw;
+         }
+     }
+     /** @brief Load existing RIFF-like file.
+      *
+      * Loads an existing file, which is not a "real" RIFF file, but similar to
+      * an ordinary RIFF file.
+      *
+      * A "real" RIFF file contains at top level a List chunk either with chunk
+      * ID "RIFF" or "RIFX". The simple constructor above expects this to be
+      * case, and if it finds the toplevel List chunk to have another chunk ID
+      * than one of those two expected ones, it would throw an Exception and
+      * would refuse to load the file accordingly.
+      *
+      * Since there are however a lot of file formats which use the same simple
+      * principles of the RIFF format, with another toplevel List chunk ID
+      * though, you can use this alternative constructor here to be able to load
+      * and handle those files in the same way as you would do with "real" RIFF
+      * files.
+      *
+      * @param path - path and file name of the RIFF-alike file to be opened
+      * @param FileType - expected toplevel List chunk ID (this is the very
+      *                   first chunk found in the file)
+      * @param Endian - whether the file uses little endian or big endian layout
+      * @param layout - general file structure type
+      * @param fileOffsetSize - (optional) preference how to deal with large files
+      * @throws RIFF::Exception if error occurred while trying to load the
+      *                         given RIFF-alike file
+      */
+     File::File(const String& path, uint32_t FileType, endian_t Endian, layout_t layout, offset_size_t fileOffsetSize)
+         : List(this), Filename(path), bIsNewFile(false), Layout(layout),
+           FileOffsetPreference(fileOffsetSize)
+     {
+         SetByteOrder(Endian);
+         if (fileOffsetSize < offset_size_auto || fileOffsetSize > offset_size_64bit)
+             throw Exception("Invalid RIFF::offset_size_t");
+         FileOffsetSize = 4;
+         try {
+             __openExistingFile(path, &FileType);
+         }
+         catch (...) {
+             Cleanup();
+             throw;
+         }
+     }
+     /**
+      * Opens an already existing RIFF file or RIFF-alike file. This method
+      * shall only be called once (in a File class constructor).
+      *
+      * @param path - path and file name of the RIFF file or RIFF-alike file to
+      *               be opened
+      * @param FileType - (optional) expected chunk ID of first chunk in file
+      * @throws RIFF::Exception if error occurred while trying to load the
+      *                         given RIFF file or RIFF-alike file
+      */
+     void File::__openExistingFile(const String& path, uint32_t* FileType) {
+         io.isPerThread = false;
          #if POSIX
-         hFile = open(path.c_str(), O_RDONLY | O_NONBLOCK);
+         io.hRead = io.hWrite = open(path.c_str(), O_RDONLY | O_NONBLOCK);
-         if (hFile <= 0) {
+         if (io.hRead == -1) {
-             hFile = 0;
+             io.hRead = io.hWrite = 0;
+             String sError = strerror(errno);
+             throw RIFF::Exception("Can't open \"" + path + "\": " + sError);
+         }
+         #elif defined(WIN32)
+         io.hRead = io.hWrite = CreateFile(
+                                      path.c_str(), GENERIC_READ,
+                                      FILE_SHARE_READ | FILE_SHARE_WRITE,
+                                      NULL, OPEN_EXISTING,
+                                      FILE_ATTRIBUTE_NORMAL |
+                                      FILE_FLAG_RANDOM_ACCESS, NULL
+                                  );
+         if (io.hRead == INVALID_HANDLE_VALUE) {
+             io.hRead = io.hWrite = INVALID_HANDLE_VALUE;
              throw RIFF::Exception("Can't open \"" + path + "\"");
          }
          #else
-         hFile = fopen(path.c_str(), "rb");
+         io.hRead = io.hWrite = fopen(path.c_str(), "rb");
-         if (!hFile) throw RIFF::Exception("Can't open \"" + path + "\"");
+         if (!io.hRead) throw RIFF::Exception("Can't open \"" + path + "\"");
          #endif // POSIX
-         ulStartPos = RIFF_HEADER_SIZE;
+         io.Mode = stream_mode_read;
-         ReadHeader(0);
-         if (ChunkID != CHUNK_ID_RIFF) {
+         // determine RIFF file offset size to be used (in RIFF chunk headers)
-             throw RIFF::Exception("Not a RIFF file");
+         // according to the current file offset preference
+         FileOffsetSize = FileOffsetSizeFor(GetCurrentFileSize());
+         switch (Layout) {
+             case layout_standard: // this is a normal RIFF file
+                 ullStartPos = RIFF_HEADER_SIZE(FileOffsetSize);
+                 ReadHeader(0);
+                 if (FileType && ChunkID != *FileType)
+                     throw RIFF::Exception("Invalid file container ID");
+                 break;
+             case layout_flat: // non-standard RIFF-alike file
+                 ullStartPos = 0;
+                 ullNewChunkSize = ullCurrentChunkSize = GetCurrentFileSize();
+                 if (FileType) {
+                     uint32_t ckid;
+                     if (Read(&ckid, 4, 1) != 4) {
+                         throw RIFF::Exception("Invalid file header ID (premature end of header)");
+                     } else if (ckid != *FileType) {
+                         String s = " (expected '" + convertToString(*FileType) + "' but got '" + convertToString(ckid) + "')";
+                         throw RIFF::Exception("Invalid file header ID" + s);
+                     }
+                     SetPos(0); // reset to first byte of file
+                 }
+                 LoadSubChunks();
+                 break;
          }
      }
-     File::~File() {
+     String File::GetFileName() const {
-        #if DEBUG
+         return Filename;
-        std::cout << "File::~File()" << std::endl;
+     }
-        #endif // DEBUG
-         #if POSIX
+     void File::SetFileName(const String& path) {
-         if (hFile) close(hFile);
+         Filename = path;
+     }
+     /**
+      * This is an internal-only method which must not be used by any application
+      * and might change at any time.
+      *
+      * Resolves and returns a reference (memory location) of the RIFF file's
+      * internal and OS dependent file I/O handles which are intended to be used
+      * by the calling thread.
+      */
+     File::HandlePair& File::FileHandlePairUnsafeRef() {
+         if (io.byThread.empty()) return io;
+         const std::thread::id tid = std::this_thread::get_id();
+         const auto it = io.byThread.find(tid);
+         return (it != io.byThread.end()) ?
+             it->second :
+             io.byThread[tid] = {
+                 #if defined(WIN32)
+                 .hRead  = INVALID_HANDLE_VALUE,
+                 .hWrite = INVALID_HANDLE_VALUE,
+                 #else
+                 .hRead  = 0,
+                 .hWrite = 0,
+                 #endif
+                 .Mode = stream_mode_closed
+             };
+     }
+     /**
+      * Returns the OS dependent file I/O read and write handles intended to be
+      * used by the calling thread.
+      *
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     File::HandlePair File::FileHandlePair() const {
+         std::lock_guard<std::mutex> lock(io.mutex);
+         if (io.byThread.empty()) return io;
+         const std::thread::id tid = std::this_thread::get_id();
+         const auto it = io.byThread.find(tid);
+         return (it != io.byThread.end()) ?
+             it->second :
+             io.byThread[tid] = {
+                 #if defined(WIN32)
+                 .hRead  = INVALID_HANDLE_VALUE,
+                 .hWrite = INVALID_HANDLE_VALUE,
+                 #else
+                 .hRead  = 0,
+                 .hWrite = 0,
+                 #endif
+                 .Mode = stream_mode_closed
+             };
+      }
+     /**
+      * Returns the OS dependent file I/O read handle intended to be used by the
+      * calling thread.
+      *
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     File::Handle File::FileHandle() const {
+         return FileHandlePair().hRead;
+     }
+     /**
+      * Returns the OS dependent file I/O write handle intended to be used by the
+      * calling thread.
+      *
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     File::Handle File::FileWriteHandle() const {
+         return FileHandlePair().hWrite;
+     }
+     /**
+      * Returns the file I/O mode currently being available for the calling
+      * thread for this RIFF file (either ro, rw or closed).
+      *
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     stream_mode_t File::GetMode() const {
+         return FileHandlePair().Mode;
+     }
+     layout_t File::GetLayout() const {
+         return Layout;
+     }
+     /** @brief Change file access mode.
+      *
+      * Changes files access mode either to read-only mode or to read/write
+      * mode.
+      *
+      * @param NewMode - new file access mode
+      * @returns true if mode was changed, false if current mode already
+      *          equals new mode
+      * @throws RIFF::Exception if file could not be opened in requested file
+      *         access mode or if passed access mode is unknown
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     bool File::SetMode(stream_mode_t NewMode) {
+         bool bResetPos = false;
+         bool res = SetModeInternal(NewMode, &bResetPos);
+         // resetting position must be handled outside above's call to avoid any
+         // potential dead lock, as SetModeInternal() acquires a lock and
+         // __resetPos() acquires a lock by itself (not a theoretical issue!)
+         if (bResetPos)
+             __resetPos(); // reset read/write position of ALL 'Chunk' objects
+         return res;
+     }
+     bool File::SetModeInternal(stream_mode_t NewMode, bool* pResetPos) {
+         std::lock_guard<std::mutex> lock(io.mutex);
+         HandlePair& io = FileHandlePairUnsafeRef();
+         if (NewMode != io.Mode) {
+             switch (NewMode) {
+                 case stream_mode_read:
+                     if (_isValidHandle(io.hRead)) _close(io.hRead);
+                     #if POSIX
+                     io.hRead = io.hWrite = open(Filename.c_str(), O_RDONLY | O_NONBLOCK);
+                     if (io.hRead == -1) {
+                         io.hRead = io.hWrite = 0;
+                         String sError = strerror(errno);
+                         throw Exception("Could not (re)open file \"" + Filename + "\" in read mode: " + sError);
+                     }
+                     #elif defined(WIN32)
+                     io.hRead = io.hWrite = CreateFile(
+                                                  Filename.c_str(), GENERIC_READ,
+                                                  FILE_SHARE_READ | FILE_SHARE_WRITE,
+                                                  NULL, OPEN_EXISTING,
+                                                  FILE_ATTRIBUTE_NORMAL |
+                                                  FILE_FLAG_RANDOM_ACCESS,
+                                                  NULL
+                                              );
+                     if (io.hRead == INVALID_HANDLE_VALUE) {
+                         io.hRead = io.hWrite = INVALID_HANDLE_VALUE;
+                         throw Exception("Could not (re)open file \"" + Filename + "\" in read mode");
+                     }
+                     #else
+                     io.hRead = io.hWrite = fopen(Filename.c_str(), "rb");
+                     if (!io.hRead) throw Exception("Could not (re)open file \"" + Filename + "\" in read mode");
+                     #endif
+                     *pResetPos = true;
+                     break;
+                 case stream_mode_read_write:
+                     if (_isValidHandle(io.hRead)) _close(io.hRead);
+                     #if POSIX
+                     io.hRead = io.hWrite = open(Filename.c_str(), O_RDWR | O_NONBLOCK);
+                     if (io.hRead == -1) {
+                         io.hRead = io.hWrite = open(Filename.c_str(), O_RDONLY | O_NONBLOCK);
+                         String sError = strerror(errno);
+                         throw Exception("Could not open file \"" + Filename + "\" in read+write mode: " + sError);
+                     }
+                     #elif defined(WIN32)
+                     io.hRead = io.hWrite = CreateFile(
+                                                  Filename.c_str(),
+                                                  GENERIC_READ | GENERIC_WRITE,
+                                                  FILE_SHARE_READ,
+                                                  NULL, OPEN_ALWAYS,
+                                                  FILE_ATTRIBUTE_NORMAL |
+                                                  FILE_FLAG_RANDOM_ACCESS,
+                                                  NULL
+                                              );
+                     if (io.hRead == INVALID_HANDLE_VALUE) {
+                         io.hRead = io.hWrite = CreateFile(
+                                                      Filename.c_str(), GENERIC_READ,
+                                                      FILE_SHARE_READ | FILE_SHARE_WRITE,
+                                                      NULL, OPEN_EXISTING,
+                                                      FILE_ATTRIBUTE_NORMAL |
+                                                      FILE_FLAG_RANDOM_ACCESS,
+                                                      NULL
+                                                  );
+                         throw Exception("Could not (re)open file \"" + Filename + "\" in read+write mode");
+                     }
+                     #else
+                     io.hRead = io.hWrite = fopen(Filename.c_str(), "r+b");
+                     if (!io.hRead) {
+                         io.hRead = io.hWrite = fopen(Filename.c_str(), "rb");
+                         throw Exception("Could not open file \"" + Filename + "\" in read+write mode");
+                     }
+                     #endif
+                     *pResetPos = true;
+                     break;
+                 case stream_mode_closed:
+                     if (_isValidHandle(io.hRead)) _close(io.hRead);
+                     if (_isValidHandle(io.hWrite)) _close(io.hWrite);
+                     #if POSIX
+                     io.hRead = io.hWrite = 0;
+                     #elif defined(WIN32)
+                     io.hRead = io.hWrite = INVALID_HANDLE_VALUE;
+                     #else
+                     io.hRead = io.hWrite = NULL;
+                     #endif
+                     break;
+                 default:
+                     throw Exception("Unknown file access mode");
+             }
+             io.Mode = NewMode;
+             return true;
+         }
+         return false;
+     }
+     /** @brief Set the byte order to be used when saving.
+      *
+      * Set the byte order to be used in the file. A value of
+      * endian_little will create a RIFF file, endian_big a RIFX file
+      * and endian_native will create a RIFF file on little-endian
+      * machines and RIFX on big-endian machines.
+      *
+      * @param Endian - endianess to use when file is saved.
+      */
+     void File::SetByteOrder(endian_t Endian) {
+         #if WORDS_BIGENDIAN
+         bEndianNative = Endian != endian_little;
          #else
-         if (hFile) fclose(hFile);
+         bEndianNative = Endian != endian_big;
-         #endif // POSIX
+         #endif
      }
-     unsigned long File::GetFileSize() {
+     /** @brief Save changes to same file.
+      *
+      * Make all changes of all chunks persistent by writing them to the
+      * actual (same) file.
+      *
+      * @param pProgress - optional: callback function for progress notification
+      * @throws RIFF::Exception if there is an empty chunk or empty list
+      *                         chunk or any kind of IO error occurred
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     void File::Save(progress_t* pProgress) {
+         //TODO: implementation for the case where first chunk is not a global container (List chunk) is not implemented yet (i.e. Korg files)
+         if (Layout == layout_flat)
+             throw Exception("Saving a RIFF file with layout_flat is not implemented yet");
+         // make sure the RIFF tree is built (from the original file)
+         if (pProgress) {
+             // divide progress into subprogress
+             progress_t subprogress;
+             __divide_progress(pProgress, &subprogress, 3.f, 0.f); // arbitrarily subdivided into 1/3 of total progress
+             // do the actual work
+             LoadSubChunksRecursively(&subprogress);
+             // notify subprogress done
+             __notify_progress(&subprogress, 1.f);
+         } else
+             LoadSubChunksRecursively(NULL);
+         // reopen file in write mode
+         SetMode(stream_mode_read_write);
+         // get the current file size as it is now still physically stored on disk
+         const file_offset_t workingFileSize = GetCurrentFileSize();
+         // get the overall file size required to save this file
+         const file_offset_t newFileSize = GetRequiredFileSize(FileOffsetPreference);
+         // determine whether this file will yield in a large file (>=4GB) and
+         // the RIFF file offset size to be used accordingly for all chunks
+         FileOffsetSize = FileOffsetSizeFor(newFileSize);
+         const HandlePair io = FileHandlePair();
+         const Handle hRead  = io.hRead;
+         const Handle hWrite = io.hWrite;
+         // to be able to save the whole file without loading everything into
+         // RAM and without having to store the data in a temporary file, we
+         // enlarge the file with the overall positive file size change,
+         // then move current data towards the end of the file by the calculated
+         // positive file size difference and finally update / rewrite the file
+         // by copying the old data back to the right position at the beginning
+         // of the file
+         // if there are positive size changes...
+         file_offset_t positiveSizeDiff = 0;
+         if (newFileSize > workingFileSize) {
+             positiveSizeDiff = newFileSize - workingFileSize;
+             // divide progress into subprogress
+             progress_t subprogress;
+             if (pProgress)
+                 __divide_progress(pProgress, &subprogress, 3.f, 1.f); // arbitrarily subdivided into 1/3 of total progress
+             // ... we enlarge this file first ...
+             ResizeFile(newFileSize);
+             // ... and move current data by the same amount towards end of file.
+             int8_t* pCopyBuffer = new int8_t[4096];
+             #if defined(WIN32)
+             DWORD iBytesMoved = 1; // we have to pass it via pointer to the Windows API, thus the correct size must be ensured
+             #else
+             ssize_t iBytesMoved = 1;
+             #endif
+             for (file_offset_t ullPos = workingFileSize, iNotif = 0; iBytesMoved > 0; ++iNotif) {
+                 iBytesMoved = (ullPos < 4096) ? ullPos : 4096;
+                 ullPos -= iBytesMoved;
+                 #if POSIX
+                 lseek(hRead, ullPos, SEEK_SET);
+                 iBytesMoved = read(hRead, pCopyBuffer, iBytesMoved);
+                 lseek(hWrite, ullPos + positiveSizeDiff, SEEK_SET);
+                 iBytesMoved = write(hWrite, pCopyBuffer, iBytesMoved);
+                 #elif defined(WIN32)
+                 LARGE_INTEGER liFilePos;
+                 liFilePos.QuadPart = ullPos;
+                 SetFilePointerEx(hRead, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+                 ReadFile(hRead, pCopyBuffer, iBytesMoved, &iBytesMoved, NULL);
+                 liFilePos.QuadPart = ullPos + positiveSizeDiff;
+                 SetFilePointerEx(hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN);
+                 WriteFile(hWrite, pCopyBuffer, iBytesMoved, &iBytesMoved, NULL);
+                 #else
+                 fseeko(hRead, ullPos, SEEK_SET);
+                 iBytesMoved = fread(pCopyBuffer, 1, iBytesMoved, hRead);
+                 fseeko(hWrite, ullPos + positiveSizeDiff, SEEK_SET);
+                 iBytesMoved = fwrite(pCopyBuffer, 1, iBytesMoved, hWrite);
+                 #endif
+                 if (pProgress && !(iNotif % 8) && iBytesMoved > 0)
+                     __notify_progress(&subprogress, float(workingFileSize - ullPos) / float(workingFileSize));
+             }
+             delete[] pCopyBuffer;
+             if (iBytesMoved < 0) throw Exception("Could not modify file while trying to enlarge it");
+             if (pProgress)
+                 __notify_progress(&subprogress, 1.f); // notify subprogress done
+         }
+         // rebuild / rewrite complete RIFF tree ...
+         // divide progress into subprogress
+         progress_t subprogress;
+         if (pProgress)
+             __divide_progress(pProgress, &subprogress, 3.f, 2.f); // arbitrarily subdivided into 1/3 of total progress
+         // do the actual work
+         const file_offset_t finalSize = WriteChunk(0, positiveSizeDiff, pProgress ? &subprogress : NULL);
+         const file_offset_t finalActualSize = __GetFileSize(hWrite);
+         // notify subprogress done
+         if (pProgress)
+             __notify_progress(&subprogress, 1.f);
+         // resize file to the final size
+         if (finalSize < finalActualSize) ResizeFile(finalSize);
+         if (pProgress)
+             __notify_progress(pProgress, 1.0); // notify done
+     }
+     /** @brief Save changes to another file.
+      *
+      * Make all changes of all chunks persistent by writing them to another
+      * file. <b>Caution:</b> this method is optimized for writing to
+      * <b>another</b> file, do not use it to save the changes to the same
+      * file! Use File::Save() in that case instead! Ignoring this might
+      * result in a corrupted file, especially in case chunks were resized!
+      *
+      * After calling this method, this File object will be associated with
+      * the new file (given by \a path) afterwards.
+      *
+      * @param path - path and file name where everything should be written to
+      * @param pProgress - optional: callback function for progress notification
+      * @see File::IsIOPerThread() for multi-threaded streaming
+      */
+     void File::Save(const String& path, progress_t* pProgress) {
+         //TODO: we should make a check here if somebody tries to write to the same file and automatically call the other Save() method in that case
+         //TODO: implementation for the case where first chunk is not a global container (List chunk) is not implemented yet (i.e. Korg files)
+         if (Layout == layout_flat)
+             throw Exception("Saving a RIFF file with layout_flat is not implemented yet");
+         // make sure the RIFF tree is built (from the original file)
+         if (pProgress) {
+             // divide progress into subprogress
+             progress_t subprogress;
+             __divide_progress(pProgress, &subprogress, 2.f, 0.f); // arbitrarily subdivided into 1/2 of total progress
+             // do the actual work
+             LoadSubChunksRecursively(&subprogress);
+             // notify subprogress done
+             __notify_progress(&subprogress, 1.f);
+         } else
+             LoadSubChunksRecursively(NULL);
+         if (!bIsNewFile) SetMode(stream_mode_read);
+         {
+             std::lock_guard<std::mutex> lock(io.mutex);
+             HandlePair& io = FileHandlePairUnsafeRef();
+             // open the other (new) file for writing and truncate it to zero size
+             #if POSIX
+             io.hWrite = open(path.c_str(), O_RDWR | O_CREAT, S_IRUSR | S_IWUSR | S_IRGRP);
+             if (io.hWrite == -1) {
+                 io.hWrite = io.hRead;
+                 String sError = strerror(errno);
+                 throw Exception("Could not open file \"" + path + "\" for writing: " + sError);
+             }
+             #elif defined(WIN32)
+             io.hWrite = CreateFile(
+                 path.c_str(), GENERIC_WRITE, FILE_SHARE_READ,
+                 NULL, OPEN_ALWAYS, FILE_ATTRIBUTE_NORMAL |
+                 FILE_FLAG_RANDOM_ACCESS, NULL
+             );
+             if (io.hWrite == INVALID_HANDLE_VALUE) {
+                 io.hWrite = io.hRead;
+                 throw Exception("Could not open file \"" + path + "\" for writing");
+             }
+             #else
+             io.hWrite = fopen(path.c_str(), "w+b");
+             if (!io.hWrite) {
+                 io.hWrite = io.hRead;
+                 throw Exception("Could not open file \"" + path + "\" for writing");
+             }
+             #endif // POSIX
+             io.Mode = stream_mode_read_write;
+         }
+         // get the overall file size required to save this file
+         const file_offset_t newFileSize = GetRequiredFileSize(FileOffsetPreference);
+         // determine whether this file will yield in a large file (>=4GB) and
+         // the RIFF file offset size to be used accordingly for all chunks
+         FileOffsetSize = FileOffsetSizeFor(newFileSize);
+         // write complete RIFF tree to the other (new) file
+         file_offset_t ullTotalSize;
+         if (pProgress) {
+             // divide progress into subprogress
+             progress_t subprogress;
+             __divide_progress(pProgress, &subprogress, 2.f, 1.f); // arbitrarily subdivided into 1/2 of total progress
+             // do the actual work
+             ullTotalSize = WriteChunk(0, 0, &subprogress);
+             // notify subprogress done
+             __notify_progress(&subprogress, 1.f);
+         } else
+             ullTotalSize = WriteChunk(0, 0, NULL);
+         const file_offset_t ullActualSize = __GetFileSize(FileWriteHandle());
+         // resize file to the final size (if the file was originally larger)
+         if (ullActualSize > ullTotalSize) ResizeFile(ullTotalSize);
+         {
+             std::lock_guard<std::mutex> lock(io.mutex);
+             HandlePair& io = FileHandlePairUnsafeRef();
+             if (_isValidHandle(io.hWrite)) _close(io.hWrite);
+             io.hWrite = io.hRead;
+             // associate new file with this File object from now on
+             Filename = path;
+             bIsNewFile = false;
+             io.Mode = (stream_mode_t) -1; // Just set it to an undefined mode ...
+         }
+         SetMode(stream_mode_read_write); // ... so SetMode() has to reopen the file handles.
+         if (pProgress)
+             __notify_progress(pProgress, 1.0); // notify done
+     }
+     void File::ResizeFile(file_offset_t ullNewSize) {
+         const Handle hWrite = FileWriteHandle();
          #if POSIX
-         struct stat filestat;
+         if (ftruncate(hWrite, ullNewSize) < 0)
-         fstat(hFile, &filestat);
+             throw Exception("Could not resize file \"" + Filename + "\"");
-         long size = filestat.st_size;
+         #elif defined(WIN32)
-         #else // standard C functions
+         LARGE_INTEGER liFilePos;
-         long curpos = ftell(hFile);
+         liFilePos.QuadPart = ullNewSize;
-         fseek(hFile, 0, SEEK_END);
+         if (
-         long size = ftell(hFile);
+             !SetFilePointerEx(hWrite, liFilePos, NULL/*new pos pointer*/, FILE_BEGIN) ||
-         fseek(hFile, curpos, SEEK_SET);
+             !SetEndOfFile(hWrite)
-         #endif // POSIX
+         ) throw Exception("Could not resize file \"" + Filename + "\"");
+         #else
+         # error Sorry, this version of libgig only supports POSIX and Windows systems yet.
+         # error Reason: portable implementation of RIFF::File::ResizeFile() is missing (yet)!
+         #endif
+     }
+     File::~File() {
+         #if DEBUG_RIFF
+         std::cout << "File::~File()" << std::endl;
+         #endif // DEBUG_RIFF
+         Cleanup();
+     }
+     /**
+      * Returns @c true if this file has been created new from scratch and
+      * has not been stored to disk yet.
+      */
+     bool File::IsNew() const {
+         return bIsNewFile;
+     }
+     void File::Cleanup() {
+         if (IsIOPerThread()) {
+             for (auto it = io.byThread.begin(); it != io.byThread.end(); ++it) {
+                 _close(it->second.hRead);
+             }
+         } else {
+             _close(io.hRead);
+         }
+         DeleteChunkList();
+         pFile = NULL;
+     }
+     /**
+      * Returns the current size of this file (in bytes) as it is currently
+      * yet stored on disk. If this file does not yet exist on disk (i.e. when
+      * this RIFF File has just been created from scratch and Save() has not
+      * been called yet) then this method returns 0.
+      */
+     file_offset_t File::GetCurrentFileSize() const {
+         file_offset_t size = 0;
+         const Handle hRead = FileHandle();
+         try {
+             size = __GetFileSize(hRead);
+         } catch (...) {
+             size = 0;
+         }
          return size;
      }
+     /**
+      * Returns the required size (in bytes) for this RIFF File to be saved to
+      * disk. The precise size of the final file on disk depends on the RIFF
+      * file offset size actually used internally in all headers of the RIFF
+      * chunks. By default libgig handles the required file offset size
+      * automatically for you; that means it is using 32 bit offsets for files
+      * smaller than 4 GB and 64 bit offsets for files equal or larger than
+      * 4 GB. You may however also override this default behavior by passing the
+      * respective option to the RIFF File constructor to force one particular
+      * offset size. In the latter case this method will return the file size
+      * for the requested forced file offset size that will be used when calling
+      * Save() later on.
+      *
+      * You may also use the overridden method below to get the file size for
+      * an arbitrary other file offset size instead.
+      *
+      * @see offset_size_t
+      * @see GetFileOffsetSize()
+      */
+     file_offset_t File::GetRequiredFileSize() {
+         return GetRequiredFileSize(FileOffsetPreference);
+     }
+     /**
+      * Returns the rquired size (in bytes) for this RIFF file to be saved to
+      * disk, assuming the passed @a fileOffsestSize would be used for the
+      * Save() operation.
+      *
+      * This overridden method essentialy behaves like the above method, with
+      * the difference that you must provide a specific RIFF @a fileOffsetSize
+      * for calculating the theoretical final file size.
+      *
+      * @see GetFileOffsetSize()
+      */
+     file_offset_t File::GetRequiredFileSize(offset_size_t fileOffsetSize) {
+         switch (fileOffsetSize) {
+             case offset_size_auto: {
+                 file_offset_t fileSize = GetRequiredFileSize(offset_size_32bit);
+                 if (fileSize >> 32)
+                     return GetRequiredFileSize(offset_size_64bit);
+                 else
+                     return fileSize;
+             }
+             case offset_size_32bit: break;
+             case offset_size_64bit: break;
+             default: throw Exception("Internal error: Invalid RIFF::offset_size_t");
+         }
+         return RequiredPhysicalSize(FileOffsetSize);
+     }
+     int File::FileOffsetSizeFor(file_offset_t fileSize) const {
+         switch (FileOffsetPreference) {
+             case offset_size_auto:
+                 return (fileSize >> 32) ? 8 : 4;
+             case offset_size_32bit:
+                 return 4;
+             case offset_size_64bit:
+                 return 8;
+             default:
+                 throw Exception("Internal error: Invalid RIFF::offset_size_t");
+         }
+     }
+     /**
+      * Returns the current size (in bytes) of file offsets stored in the
+      * headers of all chunks of this file.
+      *
+      * Most RIFF files are using 32 bit file offsets internally, which limits
+      * them to a maximum file size of less than 4 GB though. In contrast to the
+      * common standard, this RIFF File class implementation supports handling of
+      * RIFF files equal or larger than 4 GB. In such cases 64 bit file offsets
+      * have to be used in all headers of all RIFF Chunks when being stored to a
+      * physical file. libgig by default automatically selects the correct file
+      * offset size for you. You may however also force one particular file
+      * offset size by supplying the respective option to the RIFF::File
+      * constructor.
+      *
+      * This method can be used to check which RIFF file offset size is currently
+      * being used for this RIFF File.
+      *
+      * @returns current RIFF file offset size used (in bytes)
+      * @see offset_size_t
+      */
+     int File::GetFileOffsetSize() const {
+         return FileOffsetSize;
+     }
+     /**
+      * Returns the required size (in bytes) of file offsets stored in the
+      * headers of all chunks of this file if the current RIFF tree would be
+      * saved to disk by calling Save().
+      *
+      * See GetFileOffsetSize() for mor details about RIFF file offsets.
+      *
+      * @returns RIFF file offset size required (in bytes) if being saved
+      * @see offset_size_t
+      */
+     int File::GetRequiredFileOffsetSize() {
+         return FileOffsetSizeFor(GetCurrentFileSize());
+     }
+     /** @brief Whether file streams are independent for each thread.
+      *
+      * All file I/O operations like reading from a RIFF chunk body (e.g. by
+      * calling Chunk::Read(), Chunk::ReadInt8()), writing to a RIFF chunk body
+      * (e.g. by calling Chunk::Write(), Chunk::WriteInt8()) or saving the
+      * current RIFF tree structure to some file (e.g. by calling Save())
+      * operate on a file I/O stream state, i.e. there is a "current" file
+      * read/write position and reading/writing by a certain amount of bytes
+      * automatically advances that "current" file position.
+      *
+      * By default there is only one stream state for a RIFF::File object, which
+      * is not an issue as long as only one thread is using the RIFF::File
+      * object at a time (which might also be the case in a collaborative /
+      * coroutine multi-threaded scenario).
+      *
+      * If however a RIFF::File object is read/written @b simultaniously by
+      * multiple threads this can lead to undefined behaviour as the individual
+      * threads would concurrently alter the file stream position. For such a
+      * concurrent multithreaded file I/O scenario @c SetIOPerThread(true) might
+      * be enabled which causes each thread to automatically use its own file
+      * stream state.
+      *
+      * @returns true if each thread has its own file stream state
+      *          (default: false)
+      * @see SetIOPerThread()
+      */
+     bool File::IsIOPerThread() const {
+         //NOTE: Not caring about atomicity here at all, for three reasons:
+         // 1. SetIOPerThread() is assumed to be called only once for the entire
+         //    life time of a RIFF::File, usually very early at its lifetime, and
+         //    hence a change to isPerThread should already safely be propagated
+         //    before any other thread would actually read this boolean flag.
+         // 2. This method is called very frequently, and therefore any
+         //    synchronization techique would hurt runtime efficiency.
+         // 3. Using even a mutex lock here might easily cause a deadlock due to
+         //    other locks been taken in this .cpp file, i.e. at a higher call
+         //    stack level (and this is the main reason why I removed it here).
+         return io.isPerThread;
+     }
+     /** @brief Enable/disable file streams being independent for each thread.
+      *
+      * By enabling this feature (default off) each thread will automatically use
+      * its own file I/O stream state for allowing simultanious multi-threaded
+      * file read/write operations.
+      *
+      * @b NOTE: After having enabled this feature, the individual threads must
+      * at least once check GetState() and if their file I/O stream is yet closed
+      * they must call SetMode() (i.e. once) respectively to open their own file
+      * handles before being able to use any of the Read() or Write() methods.
+      *
+      * @param enable - @c true: one independent stream state per thread,
+      *                 @c false: only one stream in total shared by @b all threads
+      * @see IsIOPerThread() for more details about this feature
+      */
+     void File::SetIOPerThread(bool enable) {
+         std::lock_guard<std::mutex> lock(io.mutex);
+         if (!io.byThread.empty() == enable) return;
+         io.isPerThread = enable;
+         if (enable) {
+             const std::thread::id tid = std::this_thread::get_id();
+             io.byThread[tid] = io;
+         } else {
+             // retain an arbitrary handle pair, close all other handle pairs
+             for (auto it = io.byThread.begin(); it != io.byThread.end(); ++it) {
+                 if (it == io.byThread.begin()) {
+                     io.hRead  = it->second.hRead;
+                     io.hWrite = it->second.hWrite;
+                 } else {
+                     _close(it->second.hRead);
+                     _close(it->second.hWrite);
+                 }
+             }
+             io.byThread.clear();
+         }
+     }
+     #if POSIX
+     file_offset_t File::__GetFileSize(int hFile) const {
+         struct stat filestat;
+         if (fstat(hFile, &filestat) == -1)
+             throw Exception("POSIX FS error: could not determine file size");
+         return filestat.st_size;
+     }
+     #elif defined(WIN32)
+     file_offset_t File::__GetFileSize(HANDLE hFile) const {
+         LARGE_INTEGER size;
+         if (!GetFileSizeEx(hFile, &size))
+             throw Exception("Windows FS error: could not determine file size");
+         return size.QuadPart;
+     }
+     #else // standard C functions
+     file_offset_t File::__GetFileSize(FILE* hFile) const {
+         off_t curpos = ftello(hFile);
+         if (fseeko(hFile, 0, SEEK_END) == -1)
+             throw Exception("FS error: could not determine file size");
+         off_t size = ftello(hFile);
+         fseeko(hFile, curpos, SEEK_SET);
+         return size;
+     }
+     #endif
  // *************** Exception ***************
  // *
+     Exception::Exception() {
+     }
+     Exception::Exception(String format, ...) {
+         va_list arg;
+         va_start(arg, format);
+         Message = assemble(format, arg);
+         va_end(arg);
+     }
+     Exception::Exception(String format, va_list arg) {
+         Message = assemble(format, arg);
+     }
      void Exception::PrintMessage() {
          std::cout << "RIFF::Exception: " << Message << std::endl;
      }
+     String Exception::assemble(String format, va_list arg) {
+         char* buf = NULL;
+         vasprintf(&buf, format.c_str(), arg);
+         String s = buf;
+         free(buf);
+         return s;
+     }
  // *************** functions ***************
  // *
-Line 836 
 namespace RIFF {
+Line 2722 
 namespace RIFF {
      }
  } // namespace RIFF
- #endif

 Legend:



Removed from v.666
 


changed lines


 
Added in v.3921
 Legend:



Removed from v.666
 


changed lines


 
Added in v.3921
-Removed from v.666
+Added in v.3921

	ViewVC Help
Powered by ViewVC