Closed Onurtag closed 4 years ago
I must admit, I didn't consider what encoding io.open defaulted to when i used it. it's used 6 places in the code, so I need update all of them for consistency.
I'm in favor of using UTF-8, I just want to verify that it doesn't break interactions with existing files before i patch it.
It's working fine in python 3 to change all uses of open
to have encoding="UTF-8"
, but python 2 is being difficult.
I'm looking into it a bit more, but otherwise I might just drop the python 2 compatibility and remove the compatibility code.
Alright, I ended up ripping out the python 2 compatibility.
The new release has encoding set as in your example: https://github.com/Pectojin/duplicati-client/releases/tag/0.5.2_beta
So both importing and exporting with your example name works correctly for me on mac, windows, and linux.
I also updated the yaml output to use UTF-8, so it even looks correct in the terminal when listing backups now (well, except on Windows where CMD seems to have no concept of how to render UTF-8, but regular text renders as usual)
Hello. While trying to import a json file as a backup, I received the "Failed to load file as JSON" error. The problem can be reproducing by importing a json backup file with any unicode text (backup name or source folder) such as "テスト".
I tested this small change and it fixes the problem for my use case. I didn't check the full source code so there might be other places that need a similar fix.
https://github.com/Onurtag/duplicati-client/commit/f22f409f717b5bf38150f2fe719d15a095667563
Thanks for developing duplicati-client. I am using it with my small python script. Stay safe.