Leon posted the following to libcdio-dev and it relates that the email was rejected. I don't know why, but here's the email in full.
- - - Hi, finally found some time to work on CD-TEXT implementation again. <http://savannah.gnu.org/patch/?7532> reads DISCID field reads and converts GENRE field ... a single cd out of about 20 that had CD-TEXT had the GENRE field set. A fairly new one though. (Pink Floyd - Dark Side of the Moon remastered) reads BLOCKSIZE and sets charset according to that info BLOCKSIZE also has some nice statistics, that could be used for debugging. Documentation is very thin on the internet. Best sources were wikipedia and cdrtools sourcecode. Looks like originally the supported character sets were supposed to be ASCII, ISO-8859-1/Windows-1252, Japanese, Korean and Chinese. Shift-JIS was used for Japanese characters. I don't think Korean and Chinese was ever supported by anything. Now. I did not even find a single cd authoring tool, that supports specifying the encoding. Looking through my cd collection I found that most of them were ASCII, a few had some special characters though and were ISO-8859-1. How did I tell? One of the new features I implemented is reading of the PACKSIZE information and those have a byte to specify the charset. I believe that some old CDs with Shift-JIS charset exist in Japan. @Rocky: Since I did not find a tool to burn dbc, I cannot provide an image. With those patches we should have a pretty complete implementation of the CD-TEXT standard how it is used. I believe now, that the TOC and TOC2 fields were never used. I really don't think it is necessary to extract block 1-7 anymore. Same reason. For the future: implement CRC checks of cd-text data. Is there crc code in cdio at all? add public methods to access blocksize data and a direct method to access charset. I hope you can use it somehow. Remark: Does logging work? Gave me segfaults when I tried. Regards Leon
