The text read off it
[page 63]
Before proceeding with the details of the data description it is only necessary to define three terms as they are used in the Commercial Translator system: "file," "record," and "field." The exact implication of these terms varies from person to person and company to company. In this manual they are used to distinguish between kinds of data that can be operated on by different Commercial Translator instructions. Thus, the verbs OPEN and CLOSE can operate only on files, and a file must therefore be regarded as a body of data organized for use within the scope of those commands. Similarly, the verbs GET and FILE can operate only on records, so that a record can be defined as a body of data which can be governed by those commands. Fields, on the other hand, are subordinate units of data and are not readily definable in terms of the commands that affect them. While these definitions may seem somewhat arbitrary, it will be found that in practice they correlate with normal principles of data organization.
Files
A file is a body of data stored in some external medium which can be made accessible to the system by the use of the verb OPEN. In this sense, an external medium is any medium which provides input to the system from without. Data in such a medium cannot be used by the system until it has been brought into it, which is the basic purpose of an OPEN command. The usual external medium of electronic data processing systems, of course, is magnetic tape.
The concept of a file as "external" to the system carries certain implications which should be considered. These arise, not from the definition of a file, but rather from certain practical considerations.
For example, a file is usually a relatively large body of data, although there is no implied relationship between the length of a file and the storage capacity of a tape. Thus, there may be more than one file on a tape, and, on the other hand, a file may extend over a number of tapes.
A file is commonly understood to consist of a number of individual records, the records being generally similar to each other in size, content, and format. (Sometimes a series of differing records may be grouped together in a file; in such a case, the programmer must make provisions for distinguishing among them.) The file itself may be unique, or it may closely resemble other files. Thus, a file of actuarial data might be the only one of its kind in a library, while a file of insurance policy data might differ from another chiefly because it covers a different area or a different period of time. In this sense, a file serves a function like that of a ledger, or a filing cabinet containing a series of similar papers. It is usually a rather large body of related information. Thus it is convenient in most cases to treat such large bodies of data as complete and separate units which can be identified externally by direct reference to the physical media in which they are stored. Such media may be controlled only by the OPEN and CLOSE instructions, and this fact accounts for the definition of a file used in the Commercial Translator system.
Many files contain special records which are used primarily to identify the file and its contents. These records are called "labels." If labels are used, one is normally placed at the beginning of the file, and another is placed at the end. These labels correspond in many ways to the labels on file drawers, except that certain technical information is required because of the nature of an electronic system. Many of the details of labeling are handled automatically by the "input-output package" provided for each system. Most labeling procedures are prescribed in the environment description for each machine system, and will be discussed, therefore, in the publication covering the Commercial Translator processor for each particular system.