Forum Discussion

Parvez_AL's avatar
Parvez_AL
Advisor
8 months ago

Invalid characters for Analytics downloaded topics for other languages

All downloaded CSV content topics for other languages has invalid characters in Community Analytics. Can this be fixed? As per support and engineering this is an issue with MS Excel as it works fine with Google sheets. I am surprised to know that Khoros coding is not compatible with MS Excel.

Curious to know how local language teams use these exports? Do they all follow that tedious process of converting the MS excel files manually?

  • luk's avatar
    luk
    Boss
    8 months ago

    I would expect Excel to not deal properly with UTF-8 data in CSV, so what you could try instead of directly opening the .csv with Excel is to use the import function within the "Data" tab where Excel allows you to specify the character-encoding. Select import from text and choose your csv file then select UTF-8 encoding. Not sure that helps, but that's what I would try (not using Excel for years 😉).

  • Parvez_AL's avatar
    Parvez_AL
    Advisor
    8 months ago

    luk , Thanks. I use that workaround and it works fine. However, our bloggers are not technical and find it very difficult to even open a single report using this workaround. And to be able to do it for 12 other languages by different teams everytime they need a report is not practical. It is surely a coding issue that needs to be modified from Khoros 's end.

    • luk's avatar
      luk
      Boss
      7 months ago
      It is surely a coding issue that needs to be modified from Khoros 's end.

      Well, I don't think it's that easy, afaik Excel by default opens csv files with ISO-8859-1 encoding, but the world has moved on for many years and UTF-8 is de-facto standard. Imagine you want to use your csv data in let's say python or whatever language to do programmatic analysis but Khoros gives you ISO-8859-1 encoded data, it would be a nightmare. So it's not that Khoros couldn't encode the exported csv that way, the question is more: Would it be better?

      PS: As I mentioned above, I'm not using excel at all for several years, so take the above with a grain of salt.

      EDIT: A quick google points me to this https://answers.microsoft.com/en-us/msoffice/forum/all/excel-read-csv-set-utf-8-as-default-for-all-csv/62eb4068-fc70-4f9b-9bd7-c904713beaf0 so I would assume something like changing a registry value to switch the default encoding to UTF-8 when opening csv files would still work (also not using windows 😉) and might solve your problem (if you can then change your registry, might be a problem on a managed device)?