And just to be clear; The font renders fine in the SWF its just when you extract the underlying text using a textsnapshot that you get the incorrect character.
Erik On 21 August 2010 15:32, Erik Engström <erik.engstr...@gmail.com> wrote: > Thanks for your reply Matthias, > I might be able to help if you can point me to where in PDF2SWF you are > performing the mapping of unicode characters? I could give it a try and send > you the fix for it if I succeed :-) > > Cheers, > Erik > > > On 21 August 2010 14:36, Matthias Kramm <matth...@scribd.com> wrote: > >> On Sat, Aug 21, 2010 at 09:24:17AM +1200, Erik Engstr?m < >> erik.engstr...@gmail.com> wrote: >> > PDF2SWF seems to have an issue with converting/embedding text in some >> > actionscript 9 converted documents. I have only seen it happening to >> > the character "f" oddly enough but it could possibly happen in other >> > scenarios as well. The effect can be seen when using GetText in >> > actionscript or searchText on textsnapshots. It basically cannot find >> > strings with the character "f" as they have been incorrectly >> > converted. This has been tested in the latest snapshot build and 0.9.1 >> > with the same result. >> >> It's a known problem, that unfortunately is difficult to fix- >> some fonts encode two versions of e.g. the "f" character. In order to be >> able to have a unique Unicode index for each (something required by >> Flash), you have to remap one to some other Unicode. >> >> I'm going to add code at some point that always remaps the version of the >> character that occurs less often, though. >> >> Matthias >> >> >> >
