ReadiumReader plugin error

[byword]

An error that occurs when the Windows username is created using non-English characters in non-English language environments. The Readium Reader plugin runs, but it fails to load supplementary files such as images and CSS.

I'm experiencing a minor issue with the Readium Reader plugin. I'm a Korean user, and when I set my Windows username in Korean, the user folder name is also created in Korean. The Readium Reader plugin has problems when any part of the full Sigil settings folder path contains Korean characters. The plugin itself runs, but it fails to load files included in the EPUB, such as images and stylesheets. Even after inserting code to convert file paths to UTF-8, the issue was not resolved. So I fixed the problem by running a server and modifying it to work with a local URL instead. Since similar issues may occur not only with Korean but also with other CJK or non-English languages, I'm sharing the modified file.

Since it temporarily runs a web server and loads the URL in a web view, you can also open the reader in an external browser when the plugin is executed:
http://127.0.0.1:51337/viewer/cloud-...=epub_content/TEMP_FOLDER/

[KevinH]

I have had no problems loading and running Sigil Korean epubs. So this is limited to the Readium Plugin only? What encoding does your Windows machine use? Is it utf-8? What filesystem encoding does python use on your Windows system. Is it utf-8?

Have you tried debugging the path in the python code to see what actual byte sequence is produced?

No other users have reported anything similar for any non-english language. And xhtml's utf-8, Qt's utf-16 and Python's utf-32 all support the full unicode character sets.

Perhaps you can modify the reader Python code to print out the exact url that is attempting to be loaded. Try modifying def acceptNavigationRequest(self, url, req_type, is_main_frame) in plugin.py to convert the url to a string and print it to a log file so we can see what is actually happenng.

[KevinH]

Based on a web search I found the following:

In Windows, the Korean file system encoding is primarily determined by the system locale and can be influenced by the chosen text encoding. For Korean characters, the default encoding is typically Code Page 949 (CP949), also known as MS949 or Windows-949, which is an extended version of KS X 1001. However, newer systems and applications often favor UTF-8 or UTF-16 for broader Unicode support.

If your Windows system is not using utf-8 or utf-16 text encoding, you might want to update it and move in that direction.