This topic was archived here:

https://web.archive.org/web/20230000000000fw_/https://www.zenhax.com/viewtopic.php?t=14391

gillyf, posted Mon Nov 09, 2020 2:30 am (60092)

---

The Secret of Monkey Island & Monkey Island 2: LeChuck's Revenge for FM Towns were released with both English & Japanese versions of the games on the disc, and I'm trying to get Japanese text dumps out of them. There's an old translation tool floating around called SCUMMtr that can export/import the full text of SCUMM games, but it obviously wasn't designed with Japanese text in mind. I successfully dumped the English texts from the FM Towns copy of each game with no problems, but because the tool only exports *.TXT files with a limited number of formatting options, all of my attempts to dump the Japanese script are garbled text.

One of the options in SCUMMtr is to export hexadecimal char codes, which seems like the most likely candidate to convert into readable text, but it looks like there's still a bunch of garbage characters in there that I can't figure out what to do with (e.g. {, n, I).

Here's the first 10 lines of the SOMI text dump in English, with matching Japanese text transcribed from the game as reference:

Code:

jungle : ?????
rowboat and oars : ???????
jungle : ?????
jungle : ?????
jungle : ?????
It's a gross old bottle.\255\003I don't want it. : ?????????????\255\003?????????
bottle : ?????
note : ??
Herman Toothrot : ????????????
There's a message in it! : ?????????!

The corresponding Japanese hexadecimal codes exported by SCUMMtr:

Code:

\x83W\x83\x83\x83\x93\x83O\x83\x8B
\x83{\x81[\x83g\x82\xC6\x83I\x81[\x83\x8B
\x83W\x83\x83\x83\x93\x83O\x83\x8B
\x83W\x83\x83\x83\x93\x83O\x83\x8B
\x83W\x83\x83\x83\x93\x83O\x83\x8B
\x94\x96\x89\x98\x82\xEA\x82\xBD\x8C\xC3\x82\xA2\x83K\x83\x89\x83X\x82\xD1\x82\xF1\x82\xBE\x81B\xFF\x03\x82\xB1\x82\xF1\x82\xC8\x95\xA8\x82\xA2\x82\xE7\x82\xC8\x82\xA2\x81B
\x83K\x83\x89\x83X\x82\xD1\x82\xF1
\x8E\xE8\x8E\x86
\x83n\x81[\x83}\x83\x93\x81E\x83g\x83D\x81[\x83X\x83\x8D\x83b\x83g
\x92\x86\x82\xC9\x8E\xE8\x8E\x86\x82\xAA\x93\xFC\x82\xC1\x82\xC4\x82\xE9\x81I

I've attached the exported hex code dumps from both games, and a sample file comparing the first 10 lines of the SOMI text dump using SCUMMtr's other export options, if they can help. MI2's other dumps are pretty much identical to SOMI's, except for some reason it kept generating UTF-16 text files instead of ANSI when that option was used.


Is it possible to get a properly encoded Japanese text dump out of this, or is SCUMMtr just not the right tool for the job?

gillyf, posted Mon Mar 22, 2021 11:08 pm (62976)

---

Solved.

After wasting an embarrassing amount of time researching character encoding, trying to manually deconstruct the mojibake & hex I got out of scummtr, and failing to reencode it into readable Japanese, the answer was to simply open the original garbled *.TXTs with Writer in Shift-JIS.

Full English & Japanese script dumps for both SOMI & MI2 are attached, if they're of use to anyone. The Japanese files use [#] instead of new lines to separate text entries, otherwise the structure of the en & jp files should be identical.

dwa, posted Tue Feb 22, 2022 10:49 pm (70078)

---

Hi gillyf!

I just wanted to let you know that the ScummTR project is now open-source and a new release has just been made (thanks to its original author allowing a new, open-source release):

https://github.com/dwatteau/scummtr/releases

Among other things, it adds a new `-r` option, which just dumps the text in its original encoding (SHIFT-JIS in your case). The `-b` option worked too, but it's quite buggy with some other games, while the `-r` option does leave most of the SCUMM opcodes untouched, so that should make it easier to use with most text editors (although SHIFT-JIS does this backslash to ? conversion).

I hope this will make the tool a bit more convenient for you or any other user having similar needs. ScummTR is still quite Western-European-centric at the moment, though:

Cheers.