Today's featured topic on support.sas.com teaches you how to use SAS to work with multiple languages and character sets in a single SAS session. The ability to switch locales and languages "on the fly" depends on the improved support for Unicode within SAS 9.2.
Although it's a less heralded component of this feature, the SAS program editor (known as the "enhanced editor" on Windows) also added Unicode support in SAS 9.2. This means that for the first time, you can view and edit SAS program content across multiple languages within your SAS applications. I'm talking about cultural languages here -- such as Chinese and Polish -- not programming languages like DATA step and macro (but those still work too!).
Is this a big deal? Let's illustrate the impact by reviewing some before-and-after pictures. Here is a SAS program that features multiple languages as it would appear in SAS 9.1.3 or SAS Enterprise Guide 4.1:
Notice all of the question marks. That's what your content looks like when national characters cannot be rendered in the application's current encoding setting. Now look at the same program as viewed in SAS 9.2 or SAS Enterprise Guide 4.2:
Does this mean that these multilingual programs will work correctly when you run them in SAS? When using the SAS Unicode server, as described in Manfred's paper, the answer is Yes. But if your SAS session is pinned to an encoding for a specific language (such as latin1 for English, or shift-Jis for Japanese), then SAS will still have trouble interpreting the "out-of-range" characters that you send across. Remember: multilingual content works best with the multilingual Unicode SAS session.
1 Comment
Pingback: I18N and L10N: Lazy terms for important work - The SAS Dummy