How to decode mojibake in old Macintosh text files?

Question

I hope this is an OK place to ask this question.

The Internet Archive has a Macintosh floppy image containing presets for an old E-mu synthesizer module. The page is here Proteus Preset Libraries

Using HFS-utils I can list it and extract the files. The binary files for the synthesizer are OK. It's the text files that look like gibberish.

In LibreOffice it's gibberish:

�@�@�@�@�@E-mu PROTEUS Family Factory Presets Library�@

���̃f�B�X�N�ɂ͂��ׂĂ� E-mu PROTEUS �t�@�~���[�̏o�׎��̃v���Z�b�g���[�߂��Ă��܂��B
�t�@�C���`���̓X�^���_�[�h MIDI �t�@�C���E�t�H�[�}�b�g�� Opcode

Running a Macintosh emulator, some filenames are gibberish, as is the "text" file.

I'm running macOS Catalina 10.15.

My question: what can I do to see the full text of those files?

Thank you!

Answer 1

(Summarizing the various comments.)

Filenames and text files on HFS disks are usually encoded with an Apple-specific character set, often Mac OS Roman. This particular disk uses a mix of that and Shift-JIS, which was a common way to encode Japanese text in the 1990s. There are two steps to decoding a file: (1) extracting it from HFS, (2) converting the Shift-JIS encoding to readable form. (And possibly (3) translating it to your preferred language.)

You found a solution for step 1. Step 2 can be accomplished with the iconv utility:

iconv -f sjis ç≈èâÇ…Ç®ì«Ç›Ç≠ÇæÇ≥Ç¢ÅB

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! 　PROTEUS ファミリーをたくさんお持ちの方へ 　! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! E-mu PROTEUS ファミリーの MIDI エクスクルーシブ・メッセージは特定の機種だけに対応する形式もありますが、プリセットや各種のマップやテーブルなどでは、ほとんど機種が共通で対応できるようになっています。このことにより、多数のライブラリアンのモ各コンピューター/各ソフトウェアで共通に使用できる標準ファイル・フォーマットです。このディスクは Apple 社の Macintosh のフォーマットのディスクに書かれていますが、 [...]

You can send the output to Google Translate to get:

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! PROTEUS ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! E-mu PROTEUS family's MIDI キシッシブ・messages are available in a format that is only compatible with specific devices, but in presets, maps, tables, etc., they are compatible with most devices. It is a standard file format that can be used in common with many computers/software. This disk is written in Apple Macintosh format disk.

Translating the filenames is a little harder. You can convert the characters back to Mac OS Roman encoding, and then feed that into iconv. "ç≈èâÇ…Ç®ì«Ç›Ç≠ÇæÇ≥Ç¢ÅB" is 82 a8 93 c7 82 dd 82 ad 82 be 82 b3 82 a2 81 42, which decodes to "お読みください。", which Google Translate interprets as "Please read."