Re: Recommended CCSID for CGI outputting Arabic and JSON

[Scott Klement <web400@xxxxxxxxxxxxxxxx>]

What we do is create the data in UTF-8 on IBM i, and then base64 encode 
it.   On the browser side, we use JavaScript code to decode the base64 
back into UTF-8, and display that on the page.  This way we can still 
use the standard EBCDIC/ASCII translation for many things, but still 
fully support Unicode.

Another way to do it would be to use CgiConvMode to set things to 
'binary', meaning the Apache server won't do any translation. Then you 
can write the UTF-8 code directly without needing to base64 encode it.


On 7/16/2019 2:45 PM, Steven Scott wrote:
> Hi,
>
> I’m currently having some issues trying to figure out what CGIJobCCSID to use in Apache when I have a CGI program that will be outputting both Arabic characters, and JSON.
> So far I’ve tried 420, which appears to lack the curly braces required for JSON, and 425 which seems to be missing quite a few of the Arabic characters that are available in 420. I’ve also tried 62224, which caused Apache to throw a 500 error, and 8612 which is also missing curly braces.
>
> Hoping there might be someone on the mailing list who has encountered a requirement like this before.
>
> Thanks!
> - Steven