[Fribidi-discuss] Activities

[Behdad Esfahbod]

Hi,

I have done a lot of work on fribidi, following has been done:

-fribidi_tables.i has been replaced with a few other files, that can
be generated from UnicodeBidi.txt and BidiMirroring.txt and some
scripts, with various memory, speed options.

-A configure option --without-charsets has been added, this allows you
to exclude the charset converter codes from library, command line tool
will use iconv to convert charsets.

-Markus Kuhn's wcwidth.c has been added to command line tools, now
command line tool breaks the lines first (according to unicode
character widths), and then reorders them, just like what UAX#9 says, 
now it breaks UTF-8 text correctly.

-Most Important: All glib material has been replaced with equivalents 
with their own namespace in fribidi_mem.[ch].

and lots of other small enhancements.


I will do the followings before releasing 0.9.1 too:

-Break fribidi_log2vis into two functions, one for finding embedding 
levels, and one for reordering each line (fribidi_log2vis will be 
still there, for one line texts).

-Another command line option, to assume two consecutive '\n's as 
paragraph separator, this will allow fribidi to be used in place of 
bidiv completely.


Any idea? feature request? api request?

Yours,
-- 
Behdad
16 Azar 1380, 2001 Dec 7

[Finger for Geek Code]