[paranoia-dev] Large hack needed to get correct CDDB IDs

[Rik Hemsley]

Hi,

We currently have this function in our code, in order to work around
a problem with libcdda_interface. The details are explained in the
large comment. Any chance of fixing this for the next release please ?

int start_of_first_data_as_in_toc;
int hack_track;

/* Mega hack.  This function comes from libcdda_interface, and is called
 * by it.  We need to override it, so we implement it ourself in the
 * hope, that shared lib semantics make the calls in libcdda_interface
 * to FixupTOC end up here, instead of it's own copy.  This usually
 * works. You don't want to know the reason for this.
 */
int FixupTOC(cdrom_drive *d, int tracks)
{
  int j;

  for (j = 0; j < tracks; j++)
  {
    if (d->disc_toc[j].dwStartSector < 0)
      d->disc_toc[j].dwStartSector = 0;

    if (j < tracks-1
      && d->disc_toc[j].dwStartSector > d->disc_toc[j+1].dwStartSector)
      d->disc_toc[j].dwStartSector = 0;
  }

  long last = d->disc_toc[0].dwStartSector;

  for (j = 1; j < tracks; j++)
  {
    if (d->disc_toc[j].dwStartSector < last)
      d->disc_toc[j].dwStartSector = last;
  }

  start_of_first_data_as_in_toc = -1;
  hack_track = -1;

  if (d->ioctl_fd != -1)
  {
    struct cdrom_multisession ms_str;

    ms_str.addr_format = CDROM_LBA;

    if (ioctl(d->ioctl_fd, CDROMMULTISESSION, &ms_str) == -1)
      return -1;

    if (ms_str.addr.lba > 100)
    {
      for (j = tracks-1; j >= 0; j--)
      {
        if (j > 0 && !IS_AUDIO(d,j) && IS_AUDIO(d,j-1))
        {
          if (d->disc_toc[j].dwStartSector > ms_str.addr.lba - 11400)
          {
            /* The next two code lines are the purpose of duplicating
             * this function, all others are an exact copy of paranoias
             * FixupTOC().  The gory details: CD-Extra consist of N
             * audio-tracks in the first session and one data-track in
             * the next session.  This means, the first sector of the
             * data track is not right behind the last sector of the
             * last audio track, so all length calculation for that last
             * audio track would be wrong.  For this the start sector of
             * the data track is adjusted (we don't need the real start
             * sector, as we don't rip that track anyway), so that the
             * last audio track end in the first session.  All well and
             * good so far.  BUT: The CDDB disc-id is based on the real
             * TOC entries so this adjustment would result in a wrong
             * Disc-ID.  We can only solve this conflict, when we save
             * the old (toc-based) start sector of the data track.  Of
             * course the correct solution would be, to only adjust the
             * _length_ of the last audio track, not the start of the
             * next track, but the internal structures of cdparanoia are
             * as they are, so the length is only implicitely given.
             * Bloody sh*.  */

            start_of_first_data_as_in_toc = d->disc_toc[j].dwStartSector;
            hack_track = j + 1;

            d->disc_toc[j].dwStartSector = ms_str.addr.lba - 11400;
          }
          break;
        }
      }
      return 1;
    }
  }
  return 0;
}

--- >8 ----
List archives:  http://www.xiph.org/archives/
Paranoia homepage: http://www.xiph.org/paranoia/
To unsubscribe from this list, send mail to 'paranoia-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.