actions/itgmania212121
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Yuck: fix MP3 sync compatibility with DWI for certain mildly corrupt MP3s.

Browse Source 
This used to work; I just noticed that it broke at some point, though I'm
not sure when.  Most of these changes is just reordering code.
This commit is contained in:
Glenn Maynard
2004-10-05 18:44:07 +00:00
parent 15020e2dbb
commit 66931a2e22
1 changed files with 41 additions and 43 deletions
Download Patch File Download Diff File Expand all files Collapse all files
stepmania/src/RageSoundReader_MP3.cpp
+41 -43
View File
@@ -391,7 +391,42 @@ int RageSoundReader_MP3::do_mad_frame_decode( bool headers_only )
ret=mad_header_decode( &mad->Frame.header,&mad->Stream );
else
ret=mad_frame_decode( &mad->Frame,&mad->Stream );
if( mad->Stream.error == MAD_ERROR_BADDATAPTR && mad->first_frame )
if( ret == -1 && (mad->Stream.error == MAD_ERROR_BUFLEN || mad->Stream.error == MAD_ERROR_BUFPTR) )
{
if( bytes_read > 25000 )
{
/* We've read this much without actually getting a frame; error. */
SetError( "Can't find data" );
return -1;
}
int ret = fill_buffer();
if( ret <= 0 )
return ret;
bytes_read += ret;
continue;
}
if(mad->Stream.error == MAD_ERROR_LOSTSYNC)
{
/* This might be an ID3V2 tag. */
const int tagsize = id3_tag_query(mad->Stream.this_frame,
mad->Stream.bufend - mad->Stream.this_frame);
if( tagsize )
{
mad_stream_skip(&mad->Stream, tagsize);
/* Don't count the tagsize against the max-read-per-call figure. */
bytes_read -= tagsize;
continue;
}
}
if( mad->Stream.error == MAD_ERROR_BADDATAPTR || mad->Stream.error == MAD_ERROR_LOSTSYNC )
{
/*
* Something's corrupt. One cause of this is cutting an MP3 in the middle
@@ -399,11 +434,8 @@ int RageSoundReader_MP3::do_mad_frame_decode( bool headers_only )
* frames that have been removed. The frame is valid--we can get a header from
* it, we just can't synth useful data.
*
* BASS pretends the bad frame is silent. Let's emulate that if we're at the
* beginning, since that's probably a cut. The only time I've seen this in
* the middle of a stream is a corruption, which will probably lose sync anyway.
* BASS pretends the bad frames are silent. Emulate that, for compatibility.
*/
ret = 0; /* pretend success */
}
@@ -439,23 +471,6 @@ int RageSoundReader_MP3::do_mad_frame_decode( bool headers_only )
return 1;
}
if(mad->Stream.error == MAD_ERROR_LOSTSYNC)
{
/* This might be an ID3V2 tag. */
const int tagsize = id3_tag_query(mad->Stream.this_frame,
mad->Stream.bufend - mad->Stream.this_frame);
if( tagsize )
{
mad_stream_skip(&mad->Stream, tagsize);
/* Don't count the tagsize against the max-read-per-call figure. */
bytes_read -= tagsize;
continue;
}
}
if( mad->Stream.error == MAD_ERROR_BADCRC )
{
/* XXX untested */
@@ -465,29 +480,12 @@ int RageSoundReader_MP3::do_mad_frame_decode( bool headers_only )
continue;
}
if( mad->Stream.error == MAD_ERROR_BUFLEN || mad->Stream.error == MAD_ERROR_BUFPTR )
if( !MAD_RECOVERABLE(mad->Stream.error) )
{
if( bytes_read > 25000 )
{
/* We've read this much without actually getting a frame; error. */
SetError( "Can't find data" );
return -1;
}
int ret = fill_buffer();
if( ret <= 0 )
return ret;
bytes_read += ret;
continue;
/* We've received an unrecoverable error. */
SetError( mad_stream_errorstr(&mad->Stream) );
return -1;
}
/* If it's some other recoverable error, just keep going. */
if( MAD_RECOVERABLE(mad->Stream.error) )
continue;
SetError( mad_stream_errorstr(&mad->Stream) );
return -1;
}
}