ILE RPG ReplaceXmlEntities()

Use this subprocedure when you need to escape common XML entities & (ampersand), < (less than), > (greater than), ‘ (single quote), and ” (double quote). Several web reports I generate using CGIDEV2 are in HTML or XML form, and those special XML characters must be escaped.

      // Variables for XML entity replacement.
     d xmlentities     s              1a   dim(5) ctdata
     d xmlreplacement  s              4a   dim(5) alt(xmlentities)

      // Prototype.
     d ReplaceXmlEntities...
     d                 pr           400a   varying
     d  invalue                     200a   options(*varsize) const

      // Subprocedure.
     p ReplaceXmlEntities...
     p                 b

     d ReplaceXmlEntities...
     d                 pi           400a   varying

     d  invalue                     200a   options(*varsize) const
     d lastpos         s             10i 0 inz
     d start           s             10i 0 inz
     d workvalue       s            400a   inz varying
     d z               s             10i 0 inz

      /free

         // See the XMLEntities CTDATA array for list.
         // Ampersand (&) should be first because that is what all the
         // other replacements use.
         workvalue = invalue;

         for z = 1 to %elem(xmlentities);
             lastpos = 0;
             start = 1;
             lastpos = %scan(xmlentities(z):invalue:start);
             dow lastpos > 0;
                 workvalue = %replace('&'+%trim(xmlreplacement(z))+';' :
                   workvalue:lastpos:1); // ":1" overwrites the existing "bad" char
                 start = lastpos+1;
                 lastpos = %scan(xmlentities(z):invalue:start);
             enddo;
         endfor;

         return workvalue;

      /end-free

     p ReplaceXmlEntities...
     p                 e

      // Compile-time data.
      // This could also be stored in a D-spec array.
**CTDATA xmlentities
&amp
<lt
>gt
'apos
"quot

Feel free to change the size of the invalue, workvalue, and return sizes. I use the procedure over single fields of data which are not very large.