Last: bibliography Up: ../karrtn.html Next: background


INTRODUCTION

          PROPOSAL FOR A STANDARD SET OF PRIMITIVES
                   FOR MACHINE-INDEPENDENT 
              CHARACTER MANIPULATION IN FORTRAN
 
                              by
 
                      Nelson H.F. Beebe
             Departments of Physics and Chemistry
                      University of Utah
                   Salt Lake City, UT 84112
 
                     Tel: (801) 581-5254
       
 
 
                         INTRODUCTION
                         ============
 
          At  the 1979 NRCC Conference  on Software Standards
 in Chemistry  held at  the University  of Utah,  the  author
 proposed that  a standard  set of  primitives for  character
 manipulation in  chemical software  written in  the  FORTRAN
 language be  adopted, and set forth  a selection of routines
 which  were  felt  to  serve  the  purpose.   The  suggested
 routines followed a  uniform naming convention of a mnemonic
 prefix CHR,  to which some opposition  was voiced by several
 participants who felt that since many of the primitives were
 INTEGER functions, their names should conform to the FORTRAN
 default by which variable names beginning with the letters I
 through N are typed as INTEGERs.  Although I personally feel
 that good  names are  more important  than the  question  of
 whether  an explicit  type  statement  must be  inserted,  a
 compromise prefix  KAR  was suggested  and accepted  by  the
 working group which considered the issue.
 
          The discussions  at the  Conference and  subsequent
 communications  showed  the desirability  of  adding  a  few
 additional primitives to facilitate programming and increase
 the   general  utility  and  efficiency   of  the  character
 primitives.    Several  participants   also  suggested  that
 standards  may already  exist in the  literature for FORTRAN
 character primitives,  and  that  such standards  should  be
 seriously considered for  adoption, if they existed.  Unlike
 the case  of bit primitives, no such  standard seems to have
 been proposed.