On Sun, Nov 1, 2009 at 15:31, Vyatcheslav Yatskovsky wrote:

...

1. you can do \def\myacute#1{\buildtextaccent\textacute{#1}} Апроксим{\myacute а}ція

2. of course you can also define an ugly shortcut: \catcode`'=\active \let\'=\myacute Апроксим\'ація

3. though in reality I would be most grateful if \'а would work out of the box (at least it mkiv) and generate accent over whatever letter comes next. If glyph doesn't exist, it could just place one over a.

Thank you, Mojca. (1) works fine.

Before asking, I tried (3) option, but failed. Why it doesn't work? This is the classic way.

Not sure, but if that letter exists in unicode and not in your font ... that might be the most probably cause. If not, read further. In mkii it has never worked since Hans reimplemented the mechanism (it was different from Knuth's) to enable choosing the right glyph - if font encoding has some glyph, it comes from the font, else it gets composed from accent and letter. Maybe I'm wrong, but from what I remember you only get those characters that indeed exist in Unicode and the whole set of allowed combinations is defined somewhere in source (see enco-def.mkii). I don't know what exactly needs to be done in mkiv. In mkii it will never get changed; I don't remember too many complaints, but the truth is that (all) the characters that people need are usually covered. If any were missing, they were added to the list. 4.) I indeed forgot the fourth option: you can also build a slightly larger file with all letter-accent combinations that you possibly need. I would assume that mkiv just builds all the accent combinations (to support "\v c" constructs) from Unicode specification. If you need some special character with a special accent like those definitions in enco-def.mkii ... then you need to ask someone else for mkiv since I stayed behind of a time for quite a giant leap. (Hans, why cannot I put any accent on \schwa any more?) Mojca