Options

Problem XML => MQ => BizTalk 2002 Special Characters

wj_mmwj_mm Member Posts: 23
edited 2005-04-15 in Navision e-Commerce
Hello together,
i have a problem with special characters (äöü). the situation is so:

i create a xml with Navision XML DOM Automation Control and write it
in a public Message Queue over Navision MQ Bus Adapter.

If i read the file with navision from the queue the special characters are ok
if the biztalk server get the file from the queue the special characters are
:(.
if the biztalk server get the file from the harddisk the special characters are ok.

have someone a solution??

i think the problem is that navision can only write in ASCII mode and biztalk server needs a file in UTF-8 format. But why the file polling function works correct?

thanks for every help

Jürgen

Comments

  • Options
    pduckpduck Member Posts: 147
    i don't think so ... which header has your xml ? encoding="ISO-8859-1" ? wht are the country setting on your navision machine (where the client runs), your navision server and your biztalk server ?
  • Options
    wj_mmwj_mm Member Posts: 23
    My XML has no Header with a encoding section, because this is in the
    biztalk server dokument definition. if i give the document a encoding information the biztalk server brings a error (bad dtd schema)

    The NAS is Running on a WIN2000 Server (English)
    The BizTalk Server has the same OS
  • Options
    pduckpduck Member Posts: 147
    mmmh i had a look in the msxml sdk under "encoding" and it really confused me :).

    what about replacing the special characters with their XHTML entity? for example this would be : Ä for Ä.

    we use this and it works fine. bit we also haven't any biztalk :)
  • Options
    wj_mmwj_mm Member Posts: 23
    this works right with biztalk too. but we have not checked if you have a xml definition of a field with length 10 and you have some special characters and the field will be to long.

    this problem makes me crazy!!
  • Options
    pduckpduck Member Posts: 147
    mmmh but this shouldn't be a problem beacuse the import parser should replace the entitys with their "real" characters ... and this is done before any check.
Sign In or Register to comment.