Thai Language - double byte?

We have a requirement to store Thai characters in a live JDE environment.

2 questions: 1 - Is Thai a double byte language? 2 - What will the impact be of having to convert a live system from single byte to double byte?

I can't give specific advice without knowing your OneWorld or E1 release and underlying DB platform.

As far as I know the Thai language is not supported on E1. I have worked with Thai data under World using a specific CCSID and a dedicated environment for Thai users. With that said if you are running E1 with Unicode you may be able to store Thai data. I have successfully stored and accessed Hebrew data under Unicode E1 although it is not a supported language. The Hebrew data is used in only a limited fashion for storing address information for customers in Israel that is then placed on invoice documents.

Thai is not a typical "double-byte" language in the CJKV group -- that is Chinese, Japanese, Korean and Vietnamese) Thai has accent characters than can be applied to the alphabet. The total number of character to accent combinations exceeds 255. Some systems store Thai data using a double-byte scheme to deal with this while others deal with it during presentation where the character and accent are each stored separately in single-bytes and merged when presented. On the iSeries there are both single-byte and double-byte CCSID's for Thai. I have only worked with the single-byte Thai CCSID 838.

All that said if you are speaking about E1 Unicode then storage will be UTF-16 and take two bytes per logical character. This will obviously increase your disk storage requirements. (Of course that is if you can get Thai data to work for you at all under E1.)

Thanks Justin. I should have provided more information:

We're on E1 8.11 SP1, tools release 8.96. We have indeed enabled Unicode UTF-16.

What I'm not clear on is whether there needs to be any kind of conversion to double byte, or whether UTF-16 means the DB already is setup for double byte and there is no conversion required. So in theory we should be able to install a Thai font, plug in a Thai keyboard, and off we go?

While you have configured your database for UTF-16 you may not be configured for Unicode within E1. Is your 8.11 SP1 instance a new install or an upgrade? If the business data datasource in question is configured for Unicode then any data you enter into the system will be stored in Unicode. No conversion from double-byte to Unicode needs to be done. If you have Thai language support and fonts installed you should be able to enter and store Thai text.

Do you have existing Thai data somewhere that you want to load into the system? (Other than typing it in via the web client)

Since I don't like to give advice regarding something I haven't done I decided to do a quick experiment with Thai text in my 8.11 SP1/8.96 lab. I have attached a document with my results. In short I was able to both input Thai text and output Thai text to PDF after tweaking some configuration. The approach in my document is the same one I took to enable Hebrew. As long as you are willing to do this without support from JDE I think it will work for you.

In truth some languages that are fully supported today started as "tier 3" languages where a JDE business partner used the OneWorld toolset to create a language pack. Arabic used to be a Tier 3 language. The steps I used in my document are some of the same steps that would have been used in that case.

Hey mate,

Worked like a charm! Thanks for your help - really appreciate it!

We installed a font called MODSSS.ttf to resolve the UBE issue. Works well.

Next step is to learn Thai!