HP 3000 Manuals

HP FORTRAN 77/iX Reference

NLS Directive 

The NLS (Native Language Support) directive supports special processing
to handle foreign language text and data.

Syntax 

$NLS [LITERALS] [ON ]
     [COMPARE ] [OFF]

ON                  turns on both LITERALS and COMPARE.

OFF                 turns off all NLS processing.  By default, the
                    Native-Computer character set and collating sequence
                    is used.

LITERALS            enables the handling of native language characters in
                    strings and comments during compilation of a source
                    program.  Run-time native language I/O is also
                    enabled.

COMPARE             enables all operators that deal with string
                    comparisons (LGE, LGT, LLE, LLT) to compare string
                    variables and string constants using the collating
                    sequence corresponding to the specified NLDATALANG
                    JCW. Run-time native language I/O is also enabled.

Default               Off; the Native-Computer character set and
                      collating sequence is used.

Location              The NLS directive must appear before any
                      nondirective statements in a program unit.

Toggling/ Duration    Cannot be toggled after the appearance of
                      nondirective statements in a program unit.

Impact on             Using the LITERALS option decreases compile time
Performance           performance.

Additional Information 


NOTE  $NLS LITERALS replaces the NLS_SOURCE compiler directive.  Any
      occurrences of $NLS_SOURCE in FORTRAN source programs should be
      replaced with $NLS LITERALS.


To use the NLS directive, NLUSERLANG and NLDATALANG must be set.
NLDATALANG determines the language used for string comparisons and
scanning FORTRAN source programs.  NLUSERLANG determines the language
used to output compiler error messages.  For example, to set NLDATALANG
and NLUSERLANG, specify the following:

     :SETJCW NLDATALANG 221
     :SETJCW NLUSERLANG 0

In the example above, 221 is the JCW value for Japanese and zero is the
JCW value of Native-Computer (the default value).  Refer to the Native 
Language Programmer's Guide for a complete list of JCW values.

Examples 

Following is a FORTRAN source file called test:

     $NLS ON
           PROGRAM testnls
           CHARACTER*10 st1,st2
           st1 = 'coin'
           st2 = 'change'
           IF (LLT(st1,st2)) PRINT *,'This is the Spanish language.'
           IF (LGT(st1,st2)) PRINT *,'This is the English language.'
           STOP
           END

Following are examples of setting the MPE JCWs for test, followed by the
output from the program for each setting:

     :SETJCW NLUSERLANG 0     Tells the compiler to print compile-time messages using the default message catalog. 

     :SETJCW NLDATALANG 12     Tells the compiler to do lexical comparisons in Spanish and to expect Spanish characters in the 
                               source file. 

     :FTNiXLK TEST               
     :SAVE $OLDPASS,NLSPROG
     :NLSPROG

Output:

     This is the Spanish language.

     :SETJCW NLDATALANG 0
     :NLSPROG

Output:

     This is the English language.


NOTE  The Spanish alphabet has both the letters "c" and "ch".  Because
      "c" comes before "ch" in the Spanish alphabet, coin is considered
      to be lexically less than change.  In English, change is considered
      to be lexically less than coin.

MPE/iX 5.0 Documentation