Attachment UTF-8 Character Set

Question

I have a Apex Code which generates a CSV file and saves it as an Attachment under a record. The issue is that when the file is opened using Excel, few characters do not show properly. Eg: ' as well as Arabic characters. However, while opening the file, if I set the character set as UTF-8, the characters show up properly. I wanted to know if there is a way to set the character set as UTF-8 of the Attachment when creating it?

Current code I use to save the attachment:

Attachment attc = new Attachment(ParentId = exportFile.Id);
attc.Body = Blob.valueOf(excelFileData);
attc.Name = 'My Attachment.csv';
insert attc;

Solved this: I was earlier giving contenttype as 'text/csv' The correct contenttype should have been

attc.ContentType = 'text/csv; charset=UTF-8';

Answer 1

1. SFDC Apex operates in UTF-8 so any strings you create will be in UTF-8
2. If your SFDC Apex code is reading data from SObjects, those strings will also be in UTF-8
3. If your APEX code is reading data from external sources, then APEX will think the data is UTF-8 so you need to be sure the source system created its strings in UTF-8
4. I believe, by default, Excel CSV will not display most UTF-8 characters as UTF-8. I've noticed this when taking SFDC reports of UTF-8 fields and exporting as UTF-8 CSV vs UTF-8 Excel .xls (the latter will show UTF-8 by default). You may see different behavior opening the csv in Google Docs or Open Office.
5. The best way to debug this is to take your CSV file and open in Notepad, Notepad++ or a browser with encoding set to UTF-8. This way you can verify that you are creating the CSV attachment correctly. If yes, then the issue is more with native Microsoft Excel behavior.