On 10/12/10 19:16, Duncan Gibson wrote:

> Ignoring input, which character sets do you need FLTK to be able to display?
> 1. ASCII only
> 2. ASCII / pure UTF-8

Yes.
As I see it, (1) is strictly a subset of (2) anyway, so that's a given.

ANY other encoding has to be handled by the user - there is no credible 
way we can process the incoming data as we read it and ascertain the 
correct transcoding to get it into the utf8 that we actually need.

That said, we do already provide helper functions to do a lot of the 
actual work when the user wants to the conversion - indeed, in many 
cases we provide two functions, as when we merged inthe utf8 pacthes we 
got both the helper functions that Oksid had written in his 1.1.6-patch 
and also the versions I imported from fltk-2 that Bill had written...
Note that include the utf16 and various MS (16-bit) wide-char options.

> 3. ISO-8859-1 / CP1252
> 4. ISO-8859-* / MacRoman / Mac*
> 5. UTF-16 / wide characters
> 6. Other (leave a comment)

Apropos of nothing in particular... I noticed in the recent patches that 
in the logs someone (maybe Manolo, I wasn't paying attention, sorry...) 
about some of the text_editor changes and had mentioned hooks to handle 
various conversions - that must be relevant here? (Though I didn't look 
at the code referenced, so don't know what it does.)

Also, there was some passing mention of utf16 being a fixed-size encoding.
It isn't.
We can't make that assumption. Many character encodings in utf16 require 
surrogate pairs, so it is not a fixed size encoding.

-- 
Ian


_______________________________________________
fltk-dev mailing list
[email protected]
http://lists.easysw.com/mailman/listinfo/fltk-dev

Reply via email to