After trying to get some reports from a website using OJS 3.1.1-2, with the Arabic UI, non of the reports reflected the Arabic characters correctly.
It seems that OJS reports don’t support non-English characters yet! or this is a bad choice of the database character set?
Glad to hear back from you.
Hi @vormia. If you are downloading the .csv files and opening it with Excel, by default you will see the rubbish characters. It looks Excel doesn’t understand the characters in the file and applies its default character encoding. You can try the following to resolve it:
- Open Excel
- Import the .csv file using Data–>Get External Data → Import Text File and choosing the .csv file
- In the import wizard change the File_Origin to “Unicode UTF-8” (or choose correct language character identifier)
- Change the Delimiter to comma
- Complete the next steps and Finish.
If you are in Mac, you can use the Numbers app without any issue.
Thank a lot. That procedure helped indeed.
Even though, a user would have expected to get data as it was entered.
I appreciate your help.
Have a nice day.
Hi @vormia. That’s right, a user shouldn’t go through all those to get the encoding right. So I created an issue and submitted a PR and is now merged in to the repository. See: Fix report generator character encoding problem in RTL languages · Issue #4219 · pkp/pkp-lib · GitHub. I tested it both in Arabic and Persian UI.
Thanks a lot.