Comment 4 for bug 246983

The proposed workaround will work for the buglet, that's right. But in the real application, the script will be used in many other situations and will return some strings extracted from a postgreSQL DB.

So the script has to know the coding of the calling page template to determine whether to convert the result to unicode or not.

Wouldn't it be possible to use the unicedeconflictresolver for the result of the script if the tal string-expression fails with the unicode-error?

Im willing to try this but I needed a hint where to look for a proper place to catch the error and to integrate the conflictresolver.

Thanks

   Nils