• Print  |
  • Feedback  |

FOCUS AREAS

Hot Topics

Related Links

Migration

National Language Support

National Language Support (NLS), sometimes called software globalization, enables SAS to function properly outside the United States. This topic covers a few enhancements and migration issues. For full details, read the SAS National Language Support (NLS): Reference Guide in the SAS online documentation.

SAS®9 Enhancements

You'll find that SAS®9 offers enhanced multilingual capabilities as well as easier setup and data transfer. Here are a few highlights:
VALIDMEMNAME=
is new for SAS 9.3. Naming rules are extended for more functionality in SAS data set names, view names, and item store names. If you need to enforce behavior that is compatible with previous releases, set the system option VALIDMEMNAME=COMPATIBLE. Messages are printed in the log when names cannot be transcoded. For details, see the chapter about rules for SAS data set names, view names, and item store names in SAS Language Reference: Concepts.

SAS Unicode server
enables multilingual storage and processsing. See these papers:
LOCALE= system option
is enhanced in SAS®9 to set many of the locale-specific parameters in SAS:
  • default session encoding based on environment
  • TRANTAB= system option
  • DFLANG= system option, if available
  • DATESTYLE= system option
  • PAPERSIZE= system option.

NONLSCOMPATMODE System Option
is the default in SAS 9.1 for the z/OS environment. This is a much cleaner way to run SAS, because it removes ambiguities.

Locale-sensitive formats and functions
are introduced for numeric, date, time, and monetary formatting.

Migration Issues

Here are a few possible problems and their solutions.

Using the CPORT and CIMPORT Procedures with Encoding

SAS 9.1.3 customers who use transport files to migrate from a previous release with an incompatible encoding will encounter an error. For more details, see SAS Note 20578.

Updating Substitution Characters under OS/390 (z/OS)

Under OS/390 or z/OS, when the default NONLSCOMPATMODE is in effect, SAS does not support substitution characters in SAS syntax. If you run SAS with NONLSCOMPATMODE, you will have to update existing programs to use national characters in place of those substitutions. For example, Danish customers who have substituted the Å for the $ character in existing SAS programs will have to update the SAS syntax to use the $ in their environments.

Overriding the Encoding Attribute on Data Sets

In a SAS®9 session, data sets are stamped with an encoding attribute when they are output, replaced, or updated (but not when input only). SAS normally uses the encoding attribute to determine whether CEDA transcoding is required.

This is a wonderful enhancement for SAS®9, but you must take care when processing existing SAS 7 and 8 data sets for the first time, because those data sets do not have an encoding attribute. If you are working under the same operating environmnt where a data set was created, and the data set does not have an encoding attribute, then the session encoding is stamped on the data set by default. This could result in an incorrect encoding stamp. You can override the data set's encoding by using the DATA step option ENCODING= or the LIBNAME options INENCODING= and OUTENCODING=.

Using the %LS Macro for Compatibility

Starting with SAS 9.1, enhancements to the LOCALE= system option have made the Locale Setup Window unnecessary, and therefore it is no longer provided. For compatibility with servers that are running an earlier relase of SAS, use the %LS macro. This macro replaces the functionality of the Locale Setup Window and sets up the remote SAS environment for data transfer. If you are running SAS®9 on both your client and server sessions, the %LS macro is not necessary.

Migrating Data Sets That Contain ASCII-OEM Characters

When a data set uses the ASCII-OEM character set, the MIGRATE procedure does not correctly translate any non-English characters. Currently, only the ASCII-OEM character set is known to have this issue. See the special migration instructions for these data sets.