Re: [Libcdio-devel] Ready for testing: libcdio branch trackno-higher-one with .cue files

[Rocky Bernstein]

Feel free to change the log level of cd-info messages as you think
appropriate. If down the line it turns out to cause problems, we'll find
out and adjust accordingly then.
ᐧ

On Sun, Dec 30, 2018 at 12:47 PM Thomas Schmitt <address@hidden> wrote:

> Hi,
>
> > Merged in now. Thanks.
>
> I will next try to get my test tested ... "make check" ?
>
>
> > Would a --versose or --log-level option on the  cd-info command work?
>
> A more vorbose documentation of its option "-d" would suffice.
> Finally debugging level 4 made the INFO message visible.
>
>
> But in what situation does it make sense to reject the .CUE file and
> telling the reason as INFO, whereas WARN is used when the .CUE file was
> accepted and a tolerable problem gets reported ?
>
> There is a difference between the two message outputs
>
>   INFO: Track number out of range 1 to 99, got 00
>
> which is followed by premature program end, and
>
>   ++ WARN: Track number out of sequence. Expected 2, got 5
>
> after which the program goes on, giving the track the number 2.
>
> Both come from cdio_log(log_level) called in the same function
> parse_cuefile(). But they get emitted with log_level 2 and 3, respectively.
>
> This is because the checking for .CUE file uses a different log_level
> than the reading of the .CUE file. The warning gets emitted only if
> the .CUE file data shall be remembered. In this case the first function
> parameter "cd" is not NULL.
> parse_cuefile() makes a difference in log levels, depending on "cd":
>
>   cdio_log_level_t log_level = (NULL == cd) ? CDIO_LOG_INFO :
> CDIO_LOG_WARN;
>
> cdio_is_cuefile() calls for checking
>
>       if (parse_cuefile(NULL, psz_cue_name))
>
> _init_bincue() calls for reading
>
>       if ( !parse_cuefile(p_env, p_env->psz_cue_name) ) return false;
>
> At least for the purpose of track number sanity, this difference in
> log_level is inappropriate.
>
>
> Have a nice day :)
>
> Thomas
>
>
>