How to unzip a Japanese ZIP file, and avoid mojibake/garbled characters

Question

I received a ZIP file from a Japanese customer.

When I try to unzip it the file and folders names are messed up:

$ unzip ~/Downloads/【新入荷ECM】資料.zip
...
 inflating: БyРVУ№Й╫ECMБzОСЧ┐/123_ГЖБ[ГXГPБ[ГX.xlsx

What is the problem, and how to avoid it?

Nicolas Raoul · Accepted Answer · 2017-07-11 08:21:16Z

16

The problem is that most ZIPs circulating in Japan have their content encoded as Shift JIS, which is not shown correctly by default on Ubuntu.

The solution is to use the -O shift-jis option in your command:

$ unzip -O shift-jis ~/Downloads/【新入荷ECM】資料.zip
...
 inflating: 【新入荷ECM】資料/123_ユースケース.xlsx

This way, the expanded files are perfectly readable in Ubuntu.

answered Jul 11, 2017 at 8:21

Nicolas Raoul

11.2k25 gold badges90 silver badges148 bronze badges

What version of unzip are you using? I am using Info-ZIP v6.0.0 on Cygwin. The option -O is not supported.
– kevinarpe
Mar 3, 2022 at 3:26
1

@kevinarpe: The option might have disappeared? In which case you might have to patch unzip using gist.github.com/hamano/573753 or use the repository suggested by Sadaharu Wakisaka.
– Nicolas Raoul
Mar 7, 2022 at 8:39

Add a comment |

Nicolas Raoul · Accepted Answer · 2017-07-12 11:50:51Z

Simple answer for this

$ sudo apt install unar
$ unar ~/Downloads/【新入荷ECM】資料.zip

unar can automatically recognize which encoding is used. It does only extract and not for compressing.

Then use 'convmv' to encode text after extract.

$ convmv -f shift_jis -t utf8 БyРVУ№Й╫ECMБzОСЧ┐/123_ГЖБ[ГXГPБ[ГX.xlsx --notest

Vice versa, if you'd like to create file(s) from utf8 into shift_jis text for windows.

$ convmv -f utf8 -t shift_jis <filename> --notest

Alternative answers, use of Ubuntu Japanese team built automatic encoding 'unzip' but you have to add repository.

2 Answers 2