Man Linux: Main Page and Category List

NAME

       wml::std::lang - Multi-Lingual Support

SYNOPSIS

        #use wml::std::lang

        <lang:new id=xx [short]>

        <lang:area>
        (xx) ... (yy) ...
        </lang:area>

        <lang:set-wildcard ...>

        <lang:star: ...*..>
        <lang:star:href: index.*.html|index.html>
        <lang:star:slice: index.*.html>

        <lang:xx>...</lang:xx>
        <lang:xx: ...>

        <xx>...</xx>
        <xx: ...>

        <lang:current>
        <lang:list>

DESCRIPTION

       This include file provides high-level multi-lingual support via Slices.
       Its purpose is to define the slices ‘‘"LANG_XX"’’ according to the
       multi-lingual selection tags.

       The general intend of this slice-based approach is to use the defined
       slices in Pass 9 (Slice) via WMLs -o option.  A typical shebang-line
       example for the use with a webserver’s content negotiation feature is:

         #!wml -o (ALL-LANG_*)+LANG_EN:index.html.en \
               -o (ALL-LANG_*)+LANG_DE:index.html.de

       Since WML 1.7.0, the "<lang:star:slice:>" tag is an alternative to this
       shebang-line.

       Before you can use a language, you have to define the corresponding
       tags via "<lang:new>". For instance when you want to use the languages
       english and german, use:

        <lang:new id=en>
        <lang:new id=de>

       Then the following tags are defined:

        <lang:en>...</lang:en>
        <lang:de>...</lang:de>
        <lang:en: ...>
        <lang:de: ...>

       i.e. for both languages a container tag and a simple tag is defined.
       The container tag is more readable while the simple tag is nicer for
       short variants. When the names "lang:xx" are still to large for you,
       you can use the "short" attribute to "<lang:new>"

        <lang:new id=en short>
        <lang:new id=de short>

       when then leads to the defintion of the shortcut variants:

        <en>...</en>
        <de>...</de>
        <en: ...>
        <de: ...>

       Additionally you always have the "<lang:area>"..."</lang:area>"
       container tag available which provides an alternative way of selecting
       the language in its body. It automatically surrounds the data between
       ‘"(xx)"’ start tags with the corresponding "LANG_XX" slice.

       The following are equal:

        <lang:xx: Foo><lang:yy Bar>
        <lang:xx>Foo</lang:xx><lang:yy>Bar</lang:yy>
        <lang:area>(xx)Foo(yy)Bar</lang:area>

       Because these three lines internally get expanded to

        [LANG_XX:Foo:][LANG_YY:Bar:]
        [LANG_XX:Foo:][LANG_YY:Bar:]
        [LANG_XX:Foo:][LANG_YY:Bar:]

       There is one additional special tag: "<lang:star:>".  This tag expands
       its attribute line like the "<lang:xx:>" tags but multiple times.
       Actually as much as defined languages exists ("<lang:new>"!).  And in
       each expansion the asterisks (=stars) in the data gets replaced by the
       language identifier.

       Is is sometimes convenient to use another wildcard, e.g. when defining
       navigation bars. The "<lang:set-wildcard>" tag does the job.  The
       attribute becomes the wildcard used in future substitutions. Without
       attribute, the default value is restored. You may specify any regular
       expression, and do not forget to escape special characters (the
       astersisk is in fact ‘‘\\*’’).

         <lang:set-wildcard "%">
         <lang:star: index.%.html>
         <lang:set-wildcard>

       There is a more specialized variant named "<lang:star:href:>" which is
       similar to "<lang:star:>" but treats its attribute value as a URL part
       and tries to check if it already exists. If it doesn’t exist the tag
       expands the value without the star or an alternative value which can be
       appended with ‘‘|alt-value’’.

       The "<lang:star:slice:>" is another variant to help writing multi-
       lingual files quickly. It must come after all occurences of
       "<lang:new>" tags.

         <lang:star:slice: index.html.*>

       The ‘%BASE’ form is recognized (see wml(1)) and an empty argument is
       equivalent to the string ‘"%BASE.*.html"’.  But note that the use of
       this tag instead of the WML shebang line prevents WMk from doing its
       job, because WMk can not guess output filenames in this case.

       For complex multi-lingual documents, you may want to know in which
       language text is currently processed.  This is achieved with

         <lang:current>

       which always returns current language (as defined in "<lang:new>" or an
       empty string when outside of any language portion. The macro

         <lang:list>

       prints the newline separated list of defined languages.

EXAMPLE

       The following is an example of a webpage "index.wml" with a multi-
       lingual header and hyperlink:

        #use wml::std::lang
        #use wml::std::href

        <lang:new id=en short>
        <lang:new id=de short>
        <lang:star:slice: index.html.*>

        <h1><en: Welcome><de: Willkommen></h1>

        <href name="The Hyperlink" url="<lang:star: index.*.html>">
        <href name="The Hyperlink" url="<lang:star:href: index2.*.html|inde.html>">

       When processed via

         $ wml index.wml

       The following two output files are generated (assuming that inde.html
       and only index2.de.html exists):

       index.html.en:

         <h1>Welcome</h1>
         <a href="index.en.html">The Hyperlink</a>
         <a href="inde.html">The Hyperlink</a>

       index.html.de:

         <h1>Willkommen</h1>
         <a href="index.de.html">The Hyperlink</a>
         <a href="index2.de.html">The Hyperlink</a>

AUTHOR

        Ralf S. Engelschall
        rse@engelschall.com
        www.engelschall.com

        Denis Barbier
        barbier@engelschall.com

REQUIRES

        Internal: P1, P2, P6, P9
        External: --

SEE ALSO

       slice(1)