<directory element>

──┬─<pointer expression>─┬─────────────────────────────────────────────┤
  ├─<array row>──────────┤
  └─<string literal>─────┘

A directory element is a file name, a directory name, or both a file name and a directory name. A directory name references a group of files. For example, the following files are all in the directory named JAMES. The first six files are in the directory named (JAMES)OBJECT, and the first five files are in the directory named (JAMES)OBJECT/TEST. Note that (JAMES)OBJECT/TEST/PRIMES is both a file name and a directory name.

(JAMES)OBJECT/TEST/COMM
(JAMES)OBJECT/TEST/SORT
(JAMES)OBJECT/TEST/PRIMES
(JAMES)OBJECT/TEST/PRIMES/1
(JAMES)OBJECT/TEST/PRIMES/2
(JAMES)OBJECT/LIBRARY1
(JAMES)MEMO

In the CHANGEFILE statement, the second directory element, the target, designates the name to which the first directory element, the source, is to be changed. If the change applies to files on pack, and a family substitution specification is not in effect (either by default through the USERDATA file or by specification in either CANDE or WFL), the target must include ON <family name>, and the source must not include a family name. If a family substitution specification is in effect, ON <family name> is not required; if ON <family name> does not appear, the family substitution specification is used to determine the family on which the files reside.

To change a filename that contains long filename nodes, refer to the CHANGELFILE procedure in the System Software Utilities Operations Reference Manual.