Hi Thomas,

Arabic *and* chinese characters in one file ..?? Puh, I don't know if
there is any text-editor or text-component which is able to display
something like that correctly ... Am I right, arabic-language is read
from right to left and chines from top to bottom ... ?

yes , you are rigth but this is not my problem.
I try to explain better what is my problem.
I have an application ( Multiterm ) that store in a proprietary format terminology records in all language
even chinese arabic russian etc.
This application can export and import terminology record *only* in text files.
When i export a record from Multiterm this is the result :

**
<Entry Number>59877
<Status>Temporary
<Reliability>1
<Source Language>en
<Category>Terminology
<Subject>AGRICULTURE
<English>site-specific management
<Remarks>Short denomination.
<English>SSCM
<Form>Abbreviation
<French>agriculture de précision
<Remarks>Dénomination alternative.
<Source>^INRA^, France, 2005.
<French>aménagement spécifique à un site
<Remarks>Dénomination alternative.
<Source>^INRA^, France, 2005.
<Spanish>manejo sitio específico
<Source>Sitio Web de la Agricultura de precisión, 2005 (http://www.agriculturadeprecision.org/presfut/GlosarioAgPrec.htm )
<Spanish>ordenación específica para cada lugar
<Source>Unasylva No. 177, FAO, 1994 (T2230)
<Spanish>agricultura de precisión
<Arabic>ÒÑÇÚÉ ãõÍúßãÉ
<Arabic Source>International Centre for Agricultural Research in the Dry Areas, ICARDA, FAO Terminology Project, 2003.
<Chinese>¾«×¼Å©Òµ
<Chinese>¾«È·Å©Òµ
<Italian>agricoltura di precisione
<Source>Istituto sperimentale per la meccanizzazione agricola (ISMA), Italia, 2005 ( http://www.ingegneriaagraria.it/home/index.php?interno=ricerca.php&idricerca=19); GB, FAO, 2005.
<Italian>coltivazione in funzione della superficie
<Remarks>Traduzione.
<Source>GB, FAO, 2005.
**
 
in this case the Arabic and the Chinese entry are *not* visible in a text editor, but i can see strage character because the encoding of the text editor is wrong.
Well, now I have a terminology database stored in xml format and encoded in utf-8.
I have defined my xslt transformation to import my xml records in Multiterm, but after the transformation the record that i see in a text editor and in Multeterm too is in this form :

**
<Entry Number>59877
<Status>Temporary
<Reliability>1
<Source Language>en
<Category>Terminology
<Subject>AGRICULTURE
<English>site-specific management
<Remarks>Short denomination.
<English>SSCM
<Form>Abbreviation
<French>agriculture de précision
<Remarks>Dénomination alternative.
<Source>^INRA^, France, 2005.
<French>aménagement spécifique à un site
<Remarks>Dénomination alternative.
<Source>^INRA^, France, 2005.
<Spanish>manejo sitio específico
<Source>Sitio Web de la Agricultura de precisión, 2005 (http://www.agriculturadeprecision.org/presfut/GlosarioAgPrec.htm)
<Spanish>ordenación específica para cada lugar
<Source>Unasylva No. 177, FAO, 1994 (T2230)
<Spanish>agricultura de precisión
<Arabic>????????????
<Arabic Source>International Centre for Agricultural Research in the Dry Areas, ICARDA, FAO Terminology Project, 2003.
<Chinese>????????
<Chinese>?????????
<Italian>agricoltura di precisione
<Source>Istituto sperimentale per la meccanizzazione agricola (ISMA), Italia, 2005 (http://www.ingegneriaagraria.it/home/index.php?interno=ricerca.php&idricerca=19); GB, FAO, 2005.
<Italian>coltivazione in funzione della superficie
<Remarks>Traduzione.
<Source>GB, FAO, 2005.
**

I know that the question marks are not *really* question marks anyway this record is different from the first one and this diofference is noticed
by Multiterm that display all the arabic and chinese entry with ??????.

Maybe now my problem is more clear,
thank's a lot

Giuseppe.