Managing Fonts

[atuldalvi]

Hi,

I have a requirement like we have to enter input data in few fields in arabic language on screen and have to store all these info in the database aswell. Database will have mixture of data i.e. English and Arabic data

I am using 10.2B OpenEdge studio.

How do I accomplish this requirement ?

Do we need to install Arabic font on our local machine for development & testing ?

 

[RealHeavyDude]

As soon as you have users using more than one client code page ( English and Arabic ) you should convert your database to UTF-8. You can't store Arabic data in a database with an "English" code page like iso8859-1, ibm850 or 1252 without any problem. Doing so is a recipe for disaster.

You should also think to have your clients running with UTF-8: Imagine an user using an English code page viewing a record that was entered using an Arabic code page. All characters that can't be displayed with the English code page with be displayed as question marks. Even worse, if the English user then goes and updates the data ...

I would strongly suggest to have the Arabic fonts installed on your local systems for testing. You migh find all sorts of differences that you wouldn't see otherwise.
For example:

• English writes left right whereas Arabic writes right to left.
• Arabic labels and text might take up much more space on the screen than English ones

Heavy Regards, RealHeavyDude.

 

[TheMadDBA]

Just about to post basically the same thing RHD ;-)

Spot on

 

[atuldalvi]

Thanks for ur reply guyz. I got some idea now. There is 1 doubt as well as below
we are using oracle as a backend which is already UTF-8 and we are connecting oracle using progress schema holder.
Do we need to set schema holder to UTF-8 also ?

 

[RealHeavyDude]

Basically I can't answer whether you should convert your schema holder databse to UTF-8 too. As far as I know there is no data stored in the schema holder - from that point of view it should net be relevant. But, UTF-8 is not only about character set and code page, there is also a collation ( mostly sort order ) involved which might be different between schema holder and Oracle database then. Plus, I don't think it can hurt to convert the schema holder to UTF-8 too. If I were you I would just try it ( making a backup of your schema holder first, of course ).

Heavy Regards, RealHeavyDude.

 

[TheMadDBA]

Not necessarily the schema holder DB, but everything that connects to Oracle via the schema holder should be UTF-8.

Might as well do the schema holder as well just to simplify things.

 

[atuldalvi]

OK thanks
I will check and update you.

 

[atuldalvi]

Hi Guyz,

we converted schema holder to UTF-8 and trying to connect to oracle through schema holder but we are receiving below error
"
A "UTF-8" enabled schema holder code page requires a UTF-8 format for NLS_LANG in order to match the character sets of the client and OCI driver.
"

Then I set environment variables as below
set NLS_LANG=.AL32UTF8 or set NLS_LANG=.UTF8

No extra territory information added but still getting same error.

Pls suggest.

 

[atuldalvi]

We have setting in oracle parameter table as attached

 

[atuldalvi]

Hi Guyz,
Pls help

 

[Cringer]

Please don't bump your posts. I can understand you're desperate for this to work, but everyone here is a volunteer and posts in their own time so it can come across as pretty rude to pester.
If your problem is really urgent then you should consider contacting Progress Support (assuming you have a support contract with them (which you should!)).
Thanks.

 

[TheMadDBA]

You did make a backup of the schema holder first or try this with a test copy right??

Please post the full environment settings for the process and script that are trying to start the dataserver broker... as well as complete OS version and 10.2B service pack.

 

[atuldalvi]

Hi Guyz,

We have followed all the steps to create UTF-8 database and UTF-8 schema holder as per progress docs available on progress site still no luck.

Progress KB - How to create a Schema Holder using UTF-8 codepage to enable Unicode support?

Now we are able to update arabic in DB from data server without using -DataService ( ORA broker name ) parameter and port no in the connection string. It is not working when we put -DataService ( ORA broker name ) parameter and port no in the connection string . Data Server and Oracle both are installed on same server.
We need to know what is the default -DataService and port no which is used by progress session while establishing connection with Oracle DB. Also let me know if there is any property has to be set under ORA broker to handle UTF-8.