Print

Print


On Thu, 19 Jun 1997, Vladimir Volovich wrote:

> I noticed two problems with using the `canonical' approach (with \DeclareInputText)
> in the LCY encoding:
> first, when I do not redefine \lccode and \uccode values, and use \inputenc,
> then all seems Ok, because russian letters become active characters after
> the line \usepackage[...]{inputenc}. But if I do not use inputenc, then
> russian letters are not active, -- and I get incorrect results in \MakeUppercase
> (as well as \uppercase).

Well, this is an unfurtunate consequence of this approach. If you have,
say, \CYRA for the Cyrillic A declared with \DeclareTextSymbol, then you
must either use \CYRA in your text, or you define an input encoding which
maps this macro to an active character (via \DeclareInputText).

There are no other choices.

But tell me a reason why you want to omit inputenc?

> Second, I ran a LaTeX file which contained
> \setbox0=\hbox{...}                 % Here ... mean all (33) *uppercase* russian letters
> \showbox0
> \setbox0=\hbox{\MakeUppercase{...}} % Here ... mean all (33) *lowercase* russian letters
> \showbox0
>
> I expected to get the same results for these two \showbox commands.
> However, all kernings between letters disappeared in the second box. :-(
> I've got in a transcript file:

This info isn't enough to really check it. You must make sure that only
expansion of macros take place---not a single other token may occur,
otherwise kerning disappears.

Send a log file (for only _two_ letters with kerning :-).


    Werner