Re: [Libcdio-devel] CD-TEXT tests

[Rocky Bernstein]

On Mon, Apr 25, 2011 at 3:43 PM, Leon Merten Lohse <address@hidden>wrote:

> Hello,
>
> Rocky mentioned the lack of tests for the new CD-TEXT code several times. I
> don't quite understand what to do. BIN/CUE images won't store the
> subchannels and I am not sure about NRG.


If NRG handles this but the current image reader doesn't handle it, we need
to  improve the current image reader.


> Even if NRGs would store this information it still depends on the NRG
> reading functions and might behave differently.
>

Yes, but we would still be closer to getting a specification as to what
behavior we want. If "specification" is too lofty sounding, maybe
"examples" is better. We'd like to be able to say that if the input has
this, you should see that. Right now all of this is muddy.



>
> I tested everything (except double byte characters) on my local machine
> with several CDs.
>

I believe that and believe you did a very thorough job. But things change,
so how would we know whether
things have regressed? Relying on users to complain isn't very satisfying.


>
> I found out that PEAK Studio LE supports writing Kanji (double byte)
> CD-TEXT and they offer a 14-day trial. I am going to try that.
>

What image formats to they support? A scan of the website doesn't show
anything.


>
> If you see a way to create an image that includes the lead-in subchannels,
> please let me know.
>

Yes, I realize this is a tough one. And I don't have a solution. Perhaps
others have ideas. However another approach is
to write a scaffold or mock for the subchannel reading and test against
that. This is how the CD Paranoia code is tested.


>
> Regards
> Leon
>
>