CHANGESTR.REX: Difference between revisions

From Rosetta Code
Content added Content deleted
m (fixed underscore HTML tag usages.)
m (→‎the CHANGESTR source: changed a comment to reflect its IF statement.)
Line 38: Line 38:
if \datatype(t, 'W') then signal syntax /*4th arg not an integer. */
if \datatype(t, 'W') then signal syntax /*4th arg not an integer. */
if \datatype(b, 'W') then signal syntax /*5th " " " " */
if \datatype(b, 'W') then signal syntax /*5th " " " " */
if \datatype(p, 'W') then signal syntax /*5th arg " " " */
if \datatype(p, 'W') then signal syntax /*6th arg " " " */
if t<0 then signal syntax /*4th arg not non-negative*/
if t<0 then signal syntax /*4th arg " non-negative*/
if b<1 then signal syntax /*5th arg not positive. */
if b<1 then signal syntax /*5th arg " positive. */
if p<1 then signal syntax /*6th " " " */
if p<1 then signal syntax /*6th " " " */
L=length(o) /*length of OLD string. */
L=length(o) /*length of OLD string. */

Revision as of 21:04, 6 September 2016

the CHANGESTR function

This a RYO version of the REXX function   changestr   (change string).

It is included in some of the more modern Classic REXXes, but older versions of
Classic REXX don't have this function as a   BIF   (built-in-function).


This version of the   changestr   BIF has more functionality than the standard BIF.

the CHANGESTR source

<lang rexx>/*REXX program emulates the CHANGESTR BIF (built-in function) for older REXXes.*/ /*──────────── This version has more functionality: limit the number of changes. */ /*──────────── start of change occurrence#. */ /*──────────── start of change position. */

 /* ╔═══════════════════════════ CHANGESTR function ════════════════════════╗
  ╔═╩═══════════════════════════════════════════════════════════════════════╩═╗
  ║ The  CHANGESTR  function is used to replace some or all occurrences of an ║
  ║ (old)  string in a haystack  with  a new string.   The changed string is  ║
  ║ returned.    If the haystack  doesn't  contain the  old  string,  the     ║
  ║ original haystack is returned.   If the old string is a   null   string,  ║
  ║ then the  original string  is  prefixed  with the  new string.            ║
  ║                                                                           ║
  ║        new string to be used ►──────┐ ┌──────◄ limit of # changes (times).║
  ║   original string (haystack) ►────┐ │ │         [default:  ≈ one billion] ║
  ║    old string to be replaced ►──┐ │ │ │ ┌────◄ begin at this occurrence # ║
  ║ {O, H, and  N  can be null.}    │ │ │ │ │ ┌──◄ start position (default=1) ║
  ╚═╦═════════════════════════════╗ │ │ │ │ │ │ ╔═══════════════════════════╦═╝
    ╚═════════════════════════════╝ │ │ │ │ │ │ ╚═══════════════════════════╝
                                    ↓ ↓ ↓ ↓ ↓ ↓                                   */

changestr: parse arg o,h,n,t,b,p, ,$ f /*T, B, P are optional. */

          t=word(t 999999999, 1)                        /*maybe use the default?  */
          b=word(b 1        , 1)                        /*  "    "   "     "      */
          p=word(p 1        , 1)                        /*  "    "   "     "      */
          if arg() < 3             then signal syntax   /*not enough arguments.   */
          if arg() > 6             then signal syntax   /*too  many  arguments.   */
          if \datatype(t, 'W')     then signal syntax   /*4th arg not an integer. */
          if \datatype(b, 'W')     then signal syntax   /*5th  "   "   "    "     */
          if \datatype(p, 'W')     then signal syntax   /*6th arg  "   "    "     */
          if t<0                   then signal syntax   /*4th arg  "  non-negative*/
          if b<1                   then signal syntax   /*5th arg  "   positive.  */
          if p<1                   then signal syntax   /*6th  "   "      "       */
          L=length(o)                                   /*length of  OLD  string. */
          if L==0  &  t\=0         then return n || h   /*changing a null char?   */
          if p\=1  then do                              /*if P ¬= 1, adjust F & H.*/
                        f=left(h, min(p-1, length(h)))  /*keep first part intact. */
                        h=substr(h, p)                  /*only use this part of H.*/
                        end                             /*now, proceed as usual.  */
          #=0                                           /*# of changed occurrences*/
                   do j=1   while  # < t                /*keep changing, T times. */
                   parse var  h  y  (o)  _  +(L)  h     /*parse the haystack ···  */
                   if _==  then return f || $ || y    /*no more left,  return.  */
                   $=$ || y                             /*append the residual txt.*/
                   if j<b    then $=$ || o              /*append OLD if too soon. */
                             else do                    /*met the occurrence test.*/
                                  $=$ || n              /*append the  NEW  string.*/
                                  #=#+1                 /*bump  occurrence number.*/
                                  end
                   end   /*j*/                          /*Note:  most REXX  ···   */
                                                        /* CHANGESTR BIFs only ···*/
          return  f || $ || h                           /* support three options. */</lang>