Algorithm 568: PDS - A Portable Directory System

Abstract

PDS is a set of procedures that provides a machine-independent method of file specification.PDS provides capabilities beyond those provided by many vendorsupplied systems.In addition, because PDS is portable, additional capabilities, such as protection schemes, file usage statistics, or some of the functions of a source code control system [5], can be added easily.The basic function of PDS is to maintain a useful directory structure and provide a set of primitives for manipulating that structure.The PDS directory structure is identical to the tree structure of the UNIX [4] file system, and many of the PDS primitives are identical to UNIX primitives.PDS is, in large part, a portable implementation of the UNIX directory system.Besides PDS's machineindependence, the major differences are the extensibility of PDS and, as described in the next section, its i/o independence.In the simplest terms, PDS provides a directory structure and a mapping from machine-independent file names to machine-dependent names.It deals only with the information describing a file; it does not use or manipulate actual files in any way.The importance of this approach is that PDS is used to specify a file but does not participate in the actual i/o to that file.Consequently, there is no impact on i/o efficiency when PDS is used.PDS manipulates a rooted tree structure in which the leaves are files or directories and the nodes are directories.A directory is simply a list of files and directories.An example is shown in Figure 1, in which circles indicate directories and squares indicate files.The root of the tree is denoted by "/", and files and directories are denoted by their "path," which specifies their absolute position in the tree.A path is composed of the names of the nodes on the path from the root