Umlaute Charset

[philippr1991]

Hello everyone,

I am relativly new to qliksense and have a Topic that is bothering me for months.

Im getting the Data from a Sybase Database but german letters like ä,ö,ü arent displayed correctly

From reading trough Forums so far i learned that it is a charset issue and i Need to Change it to like "UTF-8" for example. Is that correct?

But how would i type it correctly in my loading Script?

Lets say i have the table "Customer" and i want the row "Name" to Display ä,ö….. correctly

LTB Connect to…...

Load Name;

[Customer]:

Select Name

from db…...

Now how do i have to Change the "load Name" part?

I would really appreciate some help here

Kind regards: Philipp Riewaldt

[marcus_sommer]

I think I would check which content is really there within these fieldvalues. In Qlik you could use something like this for it:

load
   recno() as RecNo, iterno() as IterNo,
   Name, len(Name) as NameLen, ord(mid(Name, iterno(), 1)) as ASCII
from Source while iterno() <= len(Name);

and to identify respectively translate the ASCII results just query for ASCII in google and you will find multiple listings about it. If those chars are missing or replaced with a certain dummy-char you will defintely need to change the driver. If they have proper ASCII chars you could transform then with my suggestion from above.

- Marcus

[philippr1991]

Thank u that helped me a lot.

With the Skript u suggested I managed to find the Problem. It is indeed a charset issue. The Problem is/was that i tried to apply MS dos Codepages as Windows Codepages. So far I only found MS Dos Codepages that fit to the signs. Is there a way to use these instead of Windows Codepages?

(I Need MS DOS Codepage 850) But Applycodepage(table1, 850) uses Windows Codepage 850 if im not mistaken?

[marcus_sommer]

Probably it will be windows codepages whereby I assume that there are one which are interchangeable respectively suitable on your data, too.

Beside this you could try to apply oem as character-set.

- Marcus

[philippr1991]

im afraid i couldnt find a Windows Codepage that would work. 

oem as character set looks promising since the signs seem to fit! (according to tables in Google)

but i have Trouble changing the Skript correctly and also couldnt find the correct way in qlik comunity.

Character-set example is:

LOAD * from a.txt (codepage is 10000, txt, delimiter is ',' , no labels)

i switched it into

 LOAD * from "Name column in Datatable" (oem, txt, delimiter is ',' , no labels);

["Name Table"]:

Select "Name column"

From "Name Database";

This is prolly wrong on multiple lvls? But i couldnt find any more examples/help to figure it out myself.

Many thanks for ur help so far already!

I wish u a nice Weekend!

[marcus_sommer]

If you load these data from a database you will need to apply an appropiate driver and/or charset directly within your connect-statement.

Otherwise you must take the results from the sql load and store them as txt file and trying to apply your needed charset on a following load. An alternatively to the txt file could be to concat the needed sql fields into a single field and then using a from_field loading.

Another possibility is the above mentioned translating with a mapsubstring(). In all these cases it depends on the remaining chars (because multiples loading/storing of the data might change them) which ways are possible and most suitable for your case.

- Marcus