Re: [Libcdio-devel] How tolerant to be towards CD-TEXT character set mislabeling ?

[Thomas Schmitt]

Hi,

Serge Pouliquen
> I checked my cds, and only one of them has cd-text with accent
> cdrskin output is :
>  27 : 8f 00 1b 00 00 ...

This one says ISO-8859-1. The right thing for continental european
texts. (Our medieval monks were so inventive about shortcuts.)


> It is correctly manage by current version of cd-info. cd-info v0.83 was
> failing to display accent.

That's 7.5 years ago. Digging in
  http://git.savannah.gnu.org/cgit/libcdio.git/log/lib/driver/cdtext.c
i see that the code with interpreting charset and converting to UTF-8
was introduced about half a year after 0.83:
  2012-03-05 "Multilanguage CD-Text from greenleon"
  
http://git.savannah.gnu.org/cgit/libcdio.git/commit/lib/driver/cdtext.c?id=abe5b8429573c9ac796c274c7abe7f76ef52dc76


> I can test some cds with ascii characters only.

Those are of interest too. If you implement the extended proposal
by changing the code in lib/driver/cdtext.c at about line 714 to:

          case CDTEXT_CHARCODE_ISO_8859_1:
            /* ... explanation comment or not ... */
            charset = (char *) "CP1252";
            break;
          case CDTEXT_CHARCODE_ASCII:
            /* ... explanation comment or not ... */
            charset = (char *) "CP1252";
            break;

then it will be appeasing to get confirmation of our theory from as
many real CDs as possible. (I.e. does real iconv with real CDs do what
i expect from character set descriptions at various places ?)

Please report any unexpected character display together with the CD's
text pack report from cdrskin.

I myself don't even have real audio CDs with CD-TEXT, but only self-made
ones with what my software deems to be correct. Cough ...


Have a nice day :)

Thomas