Received: by nummer-3.proteosys id <01C19443.4D336A1C@nummer-3.proteosys>; Thu, 3 Jan 2002 11:42:12 +0100 Return-Path: <@vm.gmd.de:LATEX-L@DHDURZ1.BITNET> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C19443.4D336A1C" x-vm-v5-data: ([nil nil nil nil nil nil nil nil nil][nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil]) X-MimeOLE: Produced By Microsoft Exchange V6.5 x-to: tex-implementors@math.ams.com, latex-l@dhdurz1 Content-class: urn:content-classes:message Subject: Loading additional hyphenation patterns Date: Tue, 5 Nov 1991 23:18:00 +0100 Message-ID: X-MS-Has-Attach: X-MS-TNEF-Correlator: From: "Don Hosek" Sender: "LaTeX-L Mailing list" To: "Rainer M. Schoepf" Reply-To: "LaTeX-L Mailing list" Status: R X-Status: X-Keywords: X-UID: 448 This is a multi-part message in MIME format. ------_=_NextPart_001_01C19443.4D336A1C Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable (The initial version of this message is going to both tex-implementors and latex-l since it started on latex-l but should really continue on tex-implementors.) A couple of weeks ago, I was flamed for my American cultural hegemony for suggesting that there was no good reason not to leave US English hyphenation patterns as pattern 0 in any format file. The arguments against this can be summarized as: (1) Why should we load English hyphenation patterns if we don't typeset English and (2) If we do typeset English, we'd rather load UK English hyphenation patterns. My direct answer to these points is that (1) you're more likely to typeset English than you think since the vast majority of accompanying documentation for TeX-related packages is written in English and (2) the argument for preference of UK patterns is more justifiable, but they aren't yet available (but _are_ under development). However, the direct answer to these points is far from my reason for proposing the scheme that I did; rather I would like to suggest a plan for multiple language definitions which has worked rather well in the multilingual environments that I use (yes, Americans are aware that there are other languages). First, the following criteria were developed for a multiple hyphenation support system: - None of the standard files should be changed. Not even to be renamed. - The scheme should work as an add-on to any format (plain, lplain, splain, amstex, lamstex, etc.) Towards this goal, the following plan seems to work pretty well. US English as defined by hyphen.tex is always present, but one doesn't _have_ to use it. First, one creates a file, say langs.tex which is best presented ab exemplo: \chardef\usenglish=3D0 \def\USEnglish{\language\usenglish \lefthyphenmin=3D2 \righthyphenmin=3D3} \newlanguage\french \language\french \input frhypha.tex % Load French hyphenation patterns \def\French{\language\french \lefthyphenmin=3D2 % I admit these values are guesses and conceivably = wrong. \righthyphenmin=3D3} \newlanguage\german \language\german \input dehypha.tex % Load German hyphenation patterns \def\German{\language\german \lefthyphenmin=3D2 \righthyphenmin=3D3} .. [ and so on ] \USEnglish % Set the default patterns A multilingual plain is then generated by initex plain \input langs \dump multilingual lplain initex lplain \input langs \dump and so forth. The choice of US English as the default patterns is strictly arbitrary, but for historical reasons, it might not be unreasonable to assume either US or UK English as the default and expect other languages to be explicitly indicated (or not--I have no strong feelings one way or the other on this issue). The definitions of \German, \USEnglish, etc. should probably be tied in with the babel package (I don't remember, does it have a plain interface or no?). The key idea here is simply the idea of having the additional languages added in at the "local modifications" stage of format generation. Certainly, it woudl seem to be better to avoid renaming hyphen.tex as most multiple-hyphenation schemes require. -dh btw, am I the only one who wishes that \patterns had stopped being restricted to initex with TeX 3.0? Perhaps the new LaTeX should be designed to have to run under IniTeX rather than VirTeX? Don Hosek dhosek@ymir.claremont.edu Quixote Digital Typography 714-621-1291 ------_=_NextPart_001_01C19443.4D336A1C Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Loading additional hyphenation patterns

(The initial version of this message is going to = both
tex-implementors and latex-l since it started on = latex-l but
should really continue on tex-implementors.)

A couple of weeks ago, I was flamed for my American = cultural
hegemony for suggesting that there was no good reason = not to
leave US English hyphenation patterns as pattern 0 in = any format
file. The arguments against this can be summarized = as:
(1) Why should we load English hyphenation patterns = if we don't
    typeset English
and
(2) If we do typeset English, we'd rather load UK = English
    hyphenation patterns.
My direct answer to these points is that
(1) you're more likely to typeset English than you = think since
the vast majority of accompanying documentation for = TeX-related
packages is written in English and (2) the argument = for
preference of UK patterns is more justifiable, but = they aren't
yet available (but _are_ under development).

However, the direct answer to these points is far from = my reason
for proposing the scheme that I did; rather I would = like to
suggest a plan for multiple language definitions = which has worked
rather well in the multilingual environments that I = use (yes,
Americans are aware that there are other = languages).

First, the following criteria were developed for a = multiple
hyphenation support system:
- None of the standard files should be changed. Not = even to be
  renamed.
- The scheme should work as an add-on to any format = (plain,
  lplain, splain, amstex, lamstex, etc.)

Towards this goal, the following plan seems to work = pretty well.
US English as defined by hyphen.tex is always = present, but one
doesn't _have_ to use it.

First, one creates a file, say langs.tex which is best = presented
ab exemplo:

\chardef\usenglish=3D0
\def\USEnglish{\language\usenglish
  \lefthyphenmin=3D2
  \righthyphenmin=3D3}
\newlanguage\french
\language\french
\input frhypha.tex % Load French hyphenation = patterns
\def\French{\language\french
  \lefthyphenmin=3D2  % I admit these = values are guesses and conceivably wrong.
  \righthyphenmin=3D3}
\newlanguage\german
\language\german
\input dehypha.tex % Load German hyphenation = patterns
\def\German{\language\german
  \lefthyphenmin=3D2
  \righthyphenmin=3D3}
.. [ and so on ]
\USEnglish % Set the default patterns

A multilingual plain is then generated by
initex plain \input langs \dump
multilingual lplain
initex lplain \input langs \dump
and so forth.

The choice of US English as the default patterns is = strictly
arbitrary, but for historical reasons, it might not = be
unreasonable to assume either US or UK English as the = default and
expect other languages to be explicitly indicated (or = not--I have
no strong feelings one way or the other on this = issue). The
definitions of \German, \USEnglish, etc. should = probably be tied
in with the babel package (I don't remember, does it = have a plain
interface or no?). The key idea here is simply the = idea of having
the additional languages added in at the "local = modifications"
stage of format generation. Certainly, it woudl seem = to be better
to avoid renaming hyphen.tex as most = multiple-hyphenation schemes
require.

-dh

btw, am I the only one who wishes that \patterns had = stopped
being restricted to initex with TeX 3.0? Perhaps the = new LaTeX
should be designed to have to run under IniTeX rather = than
VirTeX?

Don Hosek
dhosek@ymir.claremont.edu
Quixote Digital Typography
714-621-1291



------_=_NextPart_001_01C19443.4D336A1C--