Hi Sebastian,
First off, there are a few features in XML and XSLT that can assist you with
localization, e.g...
In XML there is a pre-defined attribute of xml:lang which can contain
language (and optionally, national dialect) values. The useful thing about
utilising the xml:lang rather than defining your own 'lang' attribute is
that the xml:lang is decended down the tree for you until it is overridden.
In XSLT you have the lang() function - which can be used to query/test XML
based upon it's xml:lang attribute. The lang() function also gives you a
certain amount of fallback functionality (see below) and also is one of the
few parts of XML that is not case-sensitive.
So given those features there are several ways in which you can go about
providing a localization system for use in XSLT.
As an example...
Say you had a single vocabulary XML document that defined all of your
'tokenized' words, e.g. something like...
== Vocabulary.xml =============== =============== =
<?xml version="1.0"?>
<vocab xml:lang="EN-GB"> <!-- this lang gives the default/fallback
language -->
<language name="English (GB)">
<token name="bin">Dust bin</token>
<token name="trousers" >Trousers</token>
<token name="errtest"> Something</token>
</language>
<language name="English (US)" xml:lang="en-us">
<token name="bin">Tras h can</token>
<token name="trousers" >Pants</token>
</language>
<language name="Deutsch" xml:lang="de">
<token name="bin">Müll eimer</token>
<token name="trousers" >Hose</token>
</language>
</vocab>
== end of Vocabulary.xml =============== =========
using the lang() function in XSLT if you did used the XPath selection...
/vocab/language/token[@name = 'bin'][lang('en')]
it would give you the nodes...
<token name="bin">Dust bin</token>
<token name="bin">Tras h can</token>
because both are defined under an element whose xml:lang is "en" (i.e.
"en-gb" and "en-us").
So if you do a more specific selection of...
/vocab/language/token[@name = 'bin'][lang('en-gb')]
this would give just the node...
<token name="bin">Dust bin</token>
So utilising that in XSLT could look something like...
== TestVocab.xsl =============== =============== =
<?xml version="1.0"?>
<xsl:styleshe et version="1.0"
xmlns:xsl="http ://www.w3.org/1999/XSL/Transform">
<xsl:param name="lang-select" select="'en'"/>
<xsl:output method="text"/>
<!-- get the vocabulary document -->
<xsl:variable name="vocab-doc" select="documen t('Vocabulary.x ml')/*"/>
<!-- pick the token words from the vocab document according to the selected
language -->
<!-- also providing fallback when the selected language does not exist -->
<xsl:variable name="vocab"
select="$vocab-doc/language[lang($lang-select)]/token |
$vocab-doc/language[lang(/vocab/@default-lang)][not(/vocab/language[lang($la
ng-select)])][1]/token"/>
<xsl:template match="/">
<xsl:value-of select="$vocab[@name = 'bin']"/>
</xsl:template>
</xsl:stylesheet>
== end of TestVocab.xsl =============== =========
NB. Selecting the approriate tokens into the $vocab variable once saves you
keep specifying the language in later XPath expressions.
Also, some people like to keep each language in a seperate XML document - in
which case you could use system entities to accomplish this and still only
refer to one document() in your XSLT, e.g.
== Vocabulary.xml =============== =============== =
<?xml version="1.0"?>
<!DOCTYPE includes [
<!ENTITY en-gb SYSTEM "VocabEN-GB.xml">
<!ENTITY en-us SYSTEM "VocabEN-US.xml">
<!ENTITY de SYSTEM "VocabDE.xm l">
]>
<vocab xml:lang="EN-GB"> <!-- this lang gives the default/fallback
language -->
&en-gb;
&en-us;
&de;
</vocab>
== end of Vocabulary.xml =============== =========
and the the individual language XML documents...
== VocabEN-GB.xml =============== =============== =
<language name="English (GB)" xml:lang="en-gb">
<token name="bin">Dust bin</token>
<token name="trousers" >Trousers</token>
<token name="errtest"> Something</token>
</language>
== end of VocabEN-GB.xml =============== =========
== VocabEN-US.xml =============== =============== =
<language name="English (US)" xml:lang="en-us">
<token name="bin">Tras h can</token>
<token name="trousers" >Pants</token>
</language>
== end of VocabEN-US.xml =============== =========
== VocabDE.xml =============== =============== =
<language name="Deutsch" xml:lang="de">
<token name="bin">Müll eimer</token>
<token name="trousers" >Hose</token>
</language>
== end of VocabDE.xml =============== =========
Using such a system it would then be fairly 'easy' to write a stylesheet
that
checked the existence of each token in the languages, e.g.
== CheckVocab.xsl =============== =============== =
<?xml version="1.0"?>
<xsl:styleshe et version="1.0"
xmlns:xsl="http ://www.w3.org/1999/XSL/Transform">
<xsl:output method="html" indent="yes"/>
<xsl:variable name="vocab" select="/vocab/language[lang('invalid')]/token |
/vocab/language[lang(/vocab/@xml:lang)][not(/vocab/language[lang('invalid')]
)][1]/token"/>
<xsl:key name="kXMLLangA tts" match="@xml:lan g" use="'x'"/>
<xsl:key name="kDistinct Languages" match="@xml:lan g"
use="translate( substring(.,1,2 ),'ABCDEFGHIJKL MNOPQRSTUVWXYZ' ,'abcdefghijklm n
opqrstuvwxyz')"/>
<xsl:key name="kDistinct LangDialect" match="@xml:lan g"
use="translate( .,'ABCDEFGHIJKL MNOPQRSTUVWXYZ' ,'abcdefghijklm nopqrstuvwxyz') "
/>
<xsl:key name="kDistinct Tokens" match="token" use="@name"/>
<xsl:variable name="ucase" select="'ABCDEF GHIJKLMNOPQRSTU VWXYZ'"/>
<xsl:variable name="lcase" select="'abcdef ghijklmnopqrstu vwxyz'"/>
<xsl:variable name="vDistinct Languages"
select="key('kX MLLangAtts','x' )[generate-id() =
generate-id(key('kDistin ctLanguages',tr anslate(substri ng(.,1,2),$ucas e,$lcas
e)))]"/>
<xsl:variable name="vDistinct Tokens"
select="/vocab/language/token[generate-id() =
generate-id(key('kDistin ctTokens',@name ))]"/>
<xsl:template match="/">
<html>
<head>
<title>
<xsl:text>My Vocabulary Checks</xsl:text>
</title>
</head>
<body>
<h2>
<xsl:text>Vocab ulary Checks</xsl:text>
</h2>
<hr/>
<table border="1">
<tr>
<th rowspan="2" align="center" valign="bottom" >
<xsl:text>Token </xsl:text>
</th>
<th rowspan="2" align="center" valign="bottom" >
<xsl:text>Defau lt</xsl:text>
<br/>
<xsl:text>Value </xsl:text>
</th>
<xsl:for-each select="$vDisti nctLanguages">
<xsl:variable name="this-lang"
select="transla te(substring(., 1,2),$ucase,$lc ase)"/>
<xsl:variable name="no-dialects"
select="count(k ey('kDistinctLa nguages',$this-lang)[contains(.,'-')][generate
-id() =
generate-id(key('kDistin ctLangDialect', translate(.,$uc ase,$lcase)))])+1"/>
<th align="center" colspan="{$no-dialects}">
<xsl:text>Langu age: </xsl:text>
<xsl:value-of select="$this-lang"/>
</th>
</xsl:for-each>
</tr>
<tr>
<xsl:for-each select="$vDisti nctLanguages">
<xsl:variable name="this-lang"
select="transla te(substring(., 1,2),$ucase,$lc ase)"/>
<th>
<xsl:value-of select="$this-lang"/>
</th>
<xsl:variable name="dialects"
select="key('kD istinctLanguage s',$this-lang)[contains(.,'-')][generate-id()
= generate-id(key('kDistin ctLangDialect', translate(.,$uc ase,$lcase)))]"/>
<xsl:for-each select="$dialec ts">
<th>
<xsl:value-of select="transla te(.,$ucase,$lc ase)"/>
</th>
</xsl:for-each>
</xsl:for-each>
</tr>
<xsl:for-each select="$vDisti nctTokens">
<xsl:sort/>
<tr>
<th align="left">
<xsl:value-of select="@name"/>
</th>
<td>
<xsl:value-of select="$vocab[@name = current()/@name]"/>
</td>
<xsl:variable name="this-name" select="@name"/>
<xsl:for-each select="$vDisti nctLanguages">
<xsl:variable name="this-lang"
select="transla te(substring(., 1,2),$ucase,$lc ase)"/>
<xsl:variable name="dialects"
select="key('kD istinctLanguage s',$this-lang)[contains(.,'-')][generate-id()
= generate-id(key('kDistin ctLangDialect', translate(.,$uc ase,$lcase)))]"/>
<td>
<xsl:variable name="this-token"
select="/vocab/language[lang($this-lang)]/token[@name = $this-name]"/>
<xsl:choose>
<xsl:when test="$this-token">
<xsl:value-of select="$this-token"/>
</xsl:when>
<xsl:otherwis e>
<xsl:attribut e name="bgcolor">
<xsl:text>Red </xsl:text>
</xsl:attribute>
<xsl:text>(unde fined)</xsl:text>
</xsl:otherwise>
</xsl:choose>
</td>
<xsl:for-each select="$dialec ts">
<td>
<xsl:variable name="this-token"
select="/vocab/language[lang(current())]/token[@name = $this-name]"/>
<xsl:choose>
<xsl:when test="$this-token">
<xsl:value-of select="$this-token"/>
</xsl:when>
<xsl:otherwis e>
<xsl:attribut e name="bgcolor">
<xsl:text>Red </xsl:text>
</xsl:attribute>
<xsl:text>(unde fined)</xsl:text>
</xsl:otherwise>
</xsl:choose>
</td>
</xsl:for-each>
</xsl:for-each>
</tr>
</xsl:for-each>
</table>
</body>
</html>
</xsl:template>
</xsl:stylesheet>
== end of CheckVocab.xsl =============== =========
HTH
Marrow
http://www.marrowsoft.com - home of Xselerator (XSLT IDE and debugger)
http://www.topxml.com/Xselerator
"Sebastian Kerekes" <se************ ***@gmx.net> wrote in message
news:10******** *******@newsmas ter-03.atnet.at...
Marrow wrote:
Hi,
Using variables and including them? That seems a long way to do things?
Show a short example of what you are doing - perhaps there's a better
way of achieving your goal. ;)
Ok, here's a very simple and short example:
Somewhere at the top of the document i have my variables. The variables
need to have different content depending on the language I want to use.
So my XSL is a JSP file that gets a "language" parameter and includes
the file with the right language.
When the JSP finishes processing, the result is a XSL with for the
specific language passed with "language". For "en" the result is:
<xsl:variable name="text.gree tings" select="'Userna me'" />
<b><xsl:value-of select="$text.g reetings" /></b>
for "de" it's:
<xsl:variable name="text.gree tings" select="'Benutz ername'" />
<b><xsl:value-of select="$text.g reetings" /></b>
and so on.
I found this approach in a book, and it seems ok to me, but if you have
any other (more common, more elegant) solutions, feel free to post them ;)
> To your original question...
> You could use XSLT itself to perform the checks to ensure that a word
is > defined in each language.
Uuuhm .. how? How would you solve this?
Greetings, Sebastian
