private static string LoadAndInitIFilterAndReadAll(Stream stream, string extension)
{
// Get an IFilter for a file or file extension
IFilter filter = null;
IFilterReturnCode result = NativeMethods.LoadIFilter(extension, null, ref filter);
if (result != IFilterReturnCode.S_OK)
{
Marshal.ThrowExceptionForHR((int)result);
}
// Copy the content to global memory
byte[] buffer = new byte[stream.Length];
stream.Read(buffer, 0, buffer.Length);
IntPtr nativePtr = Marshal.AllocHGlobal(buffer.Length);
Marshal.Copy(buffer, 0, nativePtr, buffer.Length);
// Create a COM stream
IStream comStream;
NativeMethods.CreateStreamOnHGlobal(nativePtr, true, out comStream);
// Load the contents to the iFilter using IPersistStream interface
var persistStream = (IPersistStream)filter;
persistStream.Load(comStream);
// Initialize iFilter
IFILTER_INIT iflags =
IFILTER_INIT.CANON_HYPHENS |
IFILTER_INIT.CANON_PARAGRAPHS |
IFILTER_INIT.CANON_SPACES |
IFILTER_INIT.APPLY_INDEX_ATTRIBUTES |
IFILTER_INIT.HARD_LINE_BREAKS |
IFILTER_INIT.FILTER_OWNED_VALUE_OK;
IFILTER_FLAGS filterFlags;
result = filter.Init(iflags, 0, IntPtr.Zero, out filterFlags);
var text = ExtractTextFromIFilter(filter);
if (filter != null)
{
Marshal.ReleaseComObject(filter);
}
return(text);
}
public override int Read(char[] array, int offset, int count)
{
int endOfChunksCount = 0;
int charsRead = 0;
while (!_done && charsRead < count)
{
if (_charsLeftFromLastRead != null)
{
int charsToCopy = (_charsLeftFromLastRead.Length < count - charsRead)?_charsLeftFromLastRead.Length:count - charsRead;
Array.Copy(_charsLeftFromLastRead, 0, array, offset + charsRead, charsToCopy);
charsRead += charsToCopy;
if (charsToCopy < _charsLeftFromLastRead.Length)
{
char[] tmp = new char[_charsLeftFromLastRead.Length - charsToCopy];
Array.Copy(_charsLeftFromLastRead, charsToCopy, tmp, 0, tmp.Length);
_charsLeftFromLastRead = tmp;
}
else
{
_charsLeftFromLastRead = null;
}
continue;
}
;
if (!_currentChunkValid)
{
IFilterReturnCode res = _filter.GetChunk(out _currentChunk);
_currentChunkValid = (res == IFilterReturnCode.S_OK) && ((_currentChunk.flags & CHUNKSTATE.CHUNK_TEXT) != 0);
if (res == IFilterReturnCode.FILTER_E_END_OF_CHUNKS)
{
endOfChunksCount++;
}
if (endOfChunksCount > 1)
{
_done = true; //That's it. no more chuncks available
}
}
if (_currentChunkValid)
{
uint bufLength = (uint)(count - charsRead);
if (bufLength < 8192)
{
bufLength = 8192; //Read ahead
}
char[] buffer = new char[bufLength];
IFilterReturnCode res = _filter.GetText(ref bufLength, buffer);
if (res == IFilterReturnCode.S_OK || res == IFilterReturnCode.FILTER_S_LAST_TEXT)
{
int cRead = (int)bufLength;
if (cRead + charsRead > count)
{
int charsLeft = (cRead + charsRead - count);
_charsLeftFromLastRead = new char[charsLeft];
Array.Copy(buffer, cRead - charsLeft, _charsLeftFromLastRead, 0, charsLeft);
cRead -= charsLeft;
}
else
{
_charsLeftFromLastRead = null;
}
Array.Copy(buffer, 0, array, offset + charsRead, cRead);
charsRead += cRead;
}
if (res == IFilterReturnCode.FILTER_S_LAST_TEXT || res == IFilterReturnCode.FILTER_E_NO_MORE_TEXT)
{
_currentChunkValid = false;
}
}
}
return(charsRead);
}
