+------------------------------------------------------------+
    | How to include hyphenation-patterns for different laguages |
    |           (including accented char's) in LaTeX.            |
    +------------------------------------------------------------+

                Claudio Fleiner (fleinerc@cfruni52.bitnet)
                             May 11, 1992


Thanks to Rainer Schoepf, Frank Mittelbach, (NFSS and DC-Fontsyle)
          Norbert Schwarz (DC-Fonts, ghyphen3.tex)
          INRS-Telecommuniactions (fhyphen.tex)

I used the following and it works quit nice. The dc-fonts are exactly the
same as Computer-Modern, but they include the accented characters in
positions above 128.

WARNING: - It works ONLY with LaTeX, NFSS and a special font
         - It should be relativly easy to use it in TeX,
         - you may perhaps be able to use it without NFSS (with some
           work)
         - You may use it with other fonts, as long as they have
           accented characters in the same postions as dc-fonts,
           otherwise you have the recode the hyphenation-patterns

First, get all this files:

        NFSS:     at ftp.uni-stuttgart.de [129.69.1.12]
                  in soft/tex/latex-styles-supported/nfss
        DC-Fonts: at ftp.uni-stuttgart.de [129.69.1.12]
                  in soft/tex/fonts/metafont/dc
        DC-Style: at ftp.uni-stuttgart.de [129.69.1.12]
                  in soft/tex/latex-styles-supported/nfss-dc

        Hyphenation-Patterns:
            german: included in dc-font
            french: use the INRS-Patterns and convert them for the
                    dc-fonts (you may get them from me
                    (fleinerc@cfruni52.bitnet)
            italian, spanish:
                    look in the directory soft/tex/hyphenation
                    at ftp.uni-stuttgard.de

Now:
        install the dc-fonts and create all the tfm-files
        (look at the documentation)

        rename hyphen.tex to uhyphen.tex
        recreate hyphen.tex with something like:

        \def\english{\language0}
        \def\german{\language1}
        \def\french{\language2}
        \english\input uhyphen.tex
        \german\input ghyphen3.tex
        \french\input fhyphen.tex
        \english

        and put the hyphenation-patterns where TeX can find them.
        Perhaps you may also want to change \pretolerance, \tolerance,
        \hbadness, \doublehyphendemerits, \finalhyphendemerits,
        \adjdemerits. I use
                \def\german{\language1
                   \pretolerance 100 \tolerance 2500  \hbadness 1500
                   \doublehyphendemerits 50000 \finalhyphendemerits 25000
                   \adjdemerits 50000
                }
        and
                   \pretolerance 100 \tolerance 200  \hbadness 1000
                   \doublehyphendemerits 10000 \finalhyphendemerits 5000
                   \adjdemerits 10000
        for french and english (default values).

        WARNING: this works only, if dclfont.sty is loaded BEFORE
                 this file, as otherwise the \catcode's, \uccode and
                 \lccode are not defined for accented chars.

        I have created a dc.tex-style, that does this and is loaded before
        the hyphenation-patterns, if you are interested, tell me.

        install nfss for LaTex, but:

        instead of fontdef.ori use fontdef.dc from the dc-fonts
        instead of preload.ori use preload.dc       "
        instead of oldlfont.sty or newlfont.sty use
                   dclfont.sty

Now you should have a working LaTeX, that does hyphenation in
3 languages. To switch, simply say \german or \french or ... .
If your keyboard does not use ISO-Chars but, for example, IBM-ASCII,
you can use conversiontables in LaTeX as follows:
(to convert IBM-ASCII to ISO, for use with DC-Fonts)

%
% Conversion from IBM-ASCII --> DC-Fonts
% (WARNING: this table includes 8-BIT Characters, that may not pass
%           gateways !!)
\catcode`
\catcode`=13 \def{^^e9} \catcode`=13 \def{^^e2}
\catcode`=13 \def{^^e4} \catcode`=13 \def{^^e0}
.
.
.
.
[The table in full can be retrieved from noa.huji.ac.il
in /tex/misc      -- it is called there   conversion_table .
It is not included here as it blocks www's transfer.]

this table is not complete, but works well with all accented characters.
It may be stored in a file like ibm.sty and included as a style option
in \documentstyle[ibm]{book}.