[XeTeX] How do mapping files affect hyphenation?

[Zdenek Wagner]

2012/2/24 Ulrike Fischer <news3 at nililand.de>:
> Am Fri, 24 Feb 2012 14:25:37 +0000 schrieb Arthur Reutenauer:
>
>>  I agree with that, that's the most sensible behaviour.  What Ulrike
>> needs are hyphenation patterns suited to the particular transliteration
>> scheme she's using.
>
> I "need" nothing ;-). I even can't read russian (and I hope the text
> doesn't say something indecent). But "translittered" hyphenation
> patters could be interesting for the OP on tex.stackexchange.
>
> I personally was only curious when I saw in the answer of egreg the
> hyphenation remark, did some research and now I'm curious about the
> discrepency between the actual behaviour and the statements of
> Jonathan in the old thread. Did I misunderstood the statements or
> did the behaviour of xetex change in between?
>
I think that Jonathan has different kind of mapping in mind. First
macros are expanded (active characters will be expanded at the very
same time). When the expanded material goes to the horizontal list,
TECkit map is applied. TeX still sees characters f and i, not a fi
ligature. When searching feasible points for line breaks, TeX measures
the widths using the actual font, thus ligatures and kernings are
applied according to the information from tfm or from GSUB and GPOS
tables. TeX measures widths of certain pieces that were built before
the ligatures were formed and kerning inserted. Of course, you can
(heoreticaly) build "fi" ligature by applying a rule in a TECkit map.
It will then affect hyphenation.
>
> --
> Ulrike Fischer
>
>
>
> --------------------------------------------------
> Subscriptions, Archive, and List information, etc.:
>  http://tug.org/mailman/listinfo/xetex



-- 
Zdeněk Wagner
http://hroch486.icpf.cas.cz/wagner/
http://icebearsoft.euweb.cz