Dojo Compressor - can we have an option to NOT normalize utf-8 encoded data to unicode escape sequences?

[daniel]

I'm working with some utf-8 data and the data is stored in js files, as Array values. When I compress the file using dojo compressor, the input characters become unicode escape sequences.

So the example file contains var chars = new Array('À');

and the result of the compression is var chars=new Array("u0102x80");

while I would like to be var chars=new Array("À");

Can we have some compressor option to NOT touch the original encoding and to not normalize user's data? I tried different things, including -Dfile.encoding=utf-8 or using native2ascii tool, but I can't find any reliable solution...

Thank you

[James Burke]

Moving to 0.9. This issue is in 0.4.x too, and it should not block the 0.9 beta (confirmed with Alex).

[alex]

(In [10070]) prevent shrinksafe from escaping unicode chars. Fixes #2331

[guest]

I'm sorry, but where is the fixed custom_rhino.jar file that includes the fix? The one from trunk is not fixed, the one from download section under 0.9 is not fixed, I've tried some other files too, from branches etc. Nothing seems to work... Can you point me to the fixed file, or I should build it myslef with the provided diff? (I hope not). THANKS.

[jkuhnert]

hey wait.,....I just looked at the change set - [10070]- and it looks like Alex did update the js.jar file - so is this ticket still really valid anymore?

[alex]

(In [10922]) ensure that we're UTF-8 end-to-end. Fixes #2440. Fixes #4712. Fixes #2331