Freemarker encoding – question marks in the place of accented characters

I am trying to print accented characters with Freemarker, but in the place of accented characters, I get only question marks. I have verified, that following statement holds:

 "UTF-8" == Environment.getCurrentEnvironment().getConfiguration().getDefaultEncoding()

I can easily see, that the accented characters are correctly held in the variable before giving it to the template.

My freemarker context can be found here: https://gist.github.com/1975239

For instance instead of:

 Jedinečný živý koncert, kde nejen, že uslyšíte, ale i uvidíte splynutí metalové kapely s padesátičlenným orchestrem včetně.

I keep getting:

 Jedine?ný ?ivý koncert, kde nejen, ?e usly?íte, ale i uvidíte splynutí metalové kapely s padesáti?lenným orchestrem v?etn?.

Thanks.

Answer

FreeMarker always treats text as UNICODE, so it doesn’t generate question marks. Since the accented letters aren’t coming from the templates (if I understand it well), it must be your output encoding that’s improper. See also: http://freemarker.org/docs/app_faq.html#faq_questionmark

BTW, getDefaultEncoding() has no role in this. That influences the decoding used when you load the templates, but you are saying that the accented characters aren’t coming from the template file, also I don’t think you can get ?-s from decoding (unless, for invalid UTF-8 byte sequences). As of the encoding of the output, FreeMarker just uses a Writer (as opposed to an OutputStream), so it can’t influence that.

Leave a Reply

Your email address will not be published. Required fields are marked *