Jump to content

OBD:Text encoding: Difference between revisions

m
(disambiguated/corrected all I could)
Line 890: Line 890:
The Chinese (Windows) version of Oni is unique in that no game content was actually localized except for text. Because of the relative simplicity of the task, the Chinese team did not build a new set of game data files, and merely modified the original .dat and .raw from the US version. [[OBD:WMDD|WMDD]], [[OBD:WMM_|WMM_]] and [[OBD:IGSt|IGSt]] instances were patched inside each level's .dat, whereas the two [[OBD:SUBT|SUBT]] files were patched in level0_Final.raw. In the case of an IGSt, text is stored in a fixed-size array (384 bytes), which has more than enough space for any translated text. WMDD and WMM_ also have fixed-size arrays (256 and 64 bytes, respectively) with at least some spare space. SUBT files, however, have a much more compact storage.
The Chinese (Windows) version of Oni is unique in that no game content was actually localized except for text. Because of the relative simplicity of the task, the Chinese team did not build a new set of game data files, and merely modified the original .dat and .raw from the US version. [[OBD:WMDD|WMDD]], [[OBD:WMM_|WMM_]] and [[OBD:IGSt|IGSt]] instances were patched inside each level's .dat, whereas the two [[OBD:SUBT|SUBT]] files were patched in level0_Final.raw. In the case of an IGSt, text is stored in a fixed-size array (384 bytes), which has more than enough space for any translated text. WMDD and WMM_ also have fixed-size arrays (256 and 64 bytes, respectively) with at least some spare space. SUBT files, however, have a much more compact storage.


The text strings of a SUBT file (stores in level0_Final.raw and indexed from the .dat part of the SUBT) are typically packed right next to each other, separated only by a single null char. Chinese text typically uses fewer glyphs, but each glyph is taking up two bytes instead of one, including punctuation and the trailing null. Thus for short sentences or interjections it is possible for a Chinese translation to completely fill up the space used by the original string and even extend into the next entry.
The text strings of a SUBT file (stored in level0_Final.raw and indexed from the .dat part of the SUBT) are typically packed right next to each other, separated only by a single null char. Chinese text typically uses fewer glyphs, but each glyph is taking up two bytes instead of one, including punctuation and the trailing null. Thus for short sentences or interjections it is possible for a Chinese translation to completely fill up the space used by the original string and even extend into the next entry.


None of the Chinese translations in SUBTmessages or SUBTsubtitles are actually longer than the original English text, and it is only the extra null byte that intrudes on the next entry's handle on several occasions. The affected handle essentially becomes a null string, and the corresponding subtitle is never found and displayed.
None of the Chinese translations in SUBTmessages or SUBTsubtitles are actually longer than the original English text, and it is only the extra null byte that intrudes on the next entry's handle on several occasions. The affected handle essentially becomes a null string, and the corresponding subtitle is never found and displayed.