Re: (Con)TeX(t), Unicode and accented characters

[Hans Hagen <pragma@wxs.nl>]

Mojca Miklavec wrote:

> But when I switched to ConTeXt I came against that problem again.
> 
> In LaTeX I used
>     \v{c}\v{s}\v{z}

this also works in context

> at first, later
>     \usepackage{csz} ... "c"s"z

in this case, i assume that csz makes " active and such; if you really want that 
, we shoul dmake an enco-fcz, with definitions like:

\startlanguagespecifics[cz]

   \appendtoks \makecharacteractive " \to \everynormalcatcodes

   \installcompoundcharacter "c {\v{c}}
   \installcompoundcharacter "s {\v{s}}
   \installcompoundcharacter "z {\v{z}}

\stoplanguagespecifics

and alike; if you want utf, you should say (at the top of the file)

\enableregime[utf]

> As I didn't know how to use any other the font, I always used CMR, the 
> default, so I didn't have problems with exotic fonts either.

this should work with all fonts, since there are fallback definitions

>     % output=pdf -translate-file=cp1250cs
>     \setupbodyfont
>         [csr,ams,rm]

try to avoid code pages

> What I don't really understand: why did the Chech TUG have to design 
> *their own font*, csr, (or made changes to cmr) if accented characters 
> worked perfectly already in plain TeX?

in cmr \v{s} is actually two characters, while in csr it's one (composed) 
character (built of two characters but seen as one); therefore when you use csr 
fonts, you can get proper hyphenation (which is notthe case in cmr where the 
usage of \accent primitive spoils the game);

next year, when i can assume that the new latin modern fonts are available 
everywhere, i will drop cmr as default cum suis in favor of lsr (which has cmr, 
plr, csr, vnr, aer etc included)

> The second problem: This works under Windows when typesetting in code 
> page 1250. How can I use accented characters if text is typeset in 
> Unicode (or latin2) in Linux?

you probably need to configure you reditor to use utf

> The third problem: How do I typeset '\v{c}' in some other font? I do 
> understand that it may not function in just any font since someone has 
> to tell the computer how the accented characters are built, but as long 
> as \v{c} works, there's no reason for
>     \useencoding[utf8]
> and then continuing with unicode encoded characters not to produce the 
> desired result.

don't worry, other fonts work ok; if an encoding does not support the chars you 
need, a composed char is constructed; [font encodings have othing to do with 
input encoding but there do influence hyphenations]

if i'm right, ec, texnansi, and qx encoding all serve your purpose

Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
      tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                              | www.pragma-pod.nl
-----------------------------------------------------------------