Group change

from Wikipedia, the free encyclopedia

Group change ( English control break ) is a term used in data processing and denotes the fact that a data record to be processed from a sorted, sequentially processed input database does not belong to the same grouping as the one processed immediately before. The associated processing principle is called group control , group change logic or group change processing . Group change processing is usually part of a batch program , usually a file is processed. A group change occurs when the classification of a data record differs from that of the previous record. Such a change often requires special processing steps such as the output of a total line for the preceding group and the initialization of variables for the following group. Some high-level programming languages (such as RPG or ABAP ) offer integrated support for group processing. Group change handling is also supported by most report generators .

One speaks of a multi-level group change if more than one classification term is to be used. The classification terms must show a hierarchical dependency, i.e. the data records are first sorted according to the first classification term and within the first classification term according to the second, etc. Ensuring the correct sorting order of the input data and determining the correct reading order (when processing several input files at the same time ) are closely related to group change processing for some problems (see also standardized programming ).

history

The tabulating machine already has the principle of group change processing, whereby the data records are in the form of punch cards . The first tabulating machine that enabled punch cards to be handled in groups was developed by Herman Hollerith in 1914. This was able to show numerical material sorted according to configurable criteria, which was particularly useful for bookkeeping . In 1936 the DEHOMAG company launched the D11 on the German market, a tabulating machine that supported a three-stage group change. The punch cards had to be sorted and mixed according to the group criteria . Depending on the level of group change, the D11 could create sums or cause a page feed in the print list output. The group criteria were configured using control panels.

Group changes can also be found at the punch card mixer

With the beginning of the computer age, when batch processing programs were the predominant form of data processing from the 1950s, group change processing was a constantly recurring processing principle. For this reason, standardized solution methods were designed, such as DIN 66260 . Because of the increasing importance of interactive dialog processing since the 1980s, group change processing plays a less dominant role.

Solution principle

Structogram of the program logic for a two-stage group change with sequential processing of a file

The usual solution procedure provides two switched loops for a one-step group change . The termination criterion of the outer loop is that there are no more data records, i. H. in the case of sequential file processing, that the end of the file has been reached. The first data record is to be read before the loop. Within this loop, the so-called group advance is first carried out; this mainly includes the recording of the order criterion of the current data record (for example the personnel number) in order to be able to recognize a change. In addition, all variables that relate to the group must be initialized at this point, for example counters for totalizing and information that will be required in later stages (e.g. name and address of the employee).

The inner loop follows the group forward. As with the outer loop, the termination criterion is the end of the file, as well as the change in the group's classification criterion. The actual processing of the record takes place in the inner loop; this can be the output of a data line on a printer or the updating of sum variables for the next higher group level. At the end of the inner loop, the next record is to be read from the file.

If the termination criterion of the inner loop has been reached, the group has changed; the current data record already refers to the following group. After the inner loop, the so-called group follow-up takes place (sometimes also called “group change”); this includes, for example, the output of a total line for the group.

After the termination criterion of the outer loop has been reached, final processing can be carried out, for example the output of a total line.

In the case of a multi-level group change, the processing part of an outer group consists of the group change sequence of the inner group with a lead, a further nested loop and a follow-up. The termination conditions of the inner loop always also include those of the outer, i.e. also the change in the superordinate group criteria. Reading the next sentence takes place at the end of the innermost loop.

literature

  • Daniel D. McCracken, Donald G. Golden: COBOL: Structured Programming with Cobol 85. 7th edition, Oldenbourg Wissenschaftsverlag, Munich 2001, ISBN 978-3-486-25625-3 .

Web links

Individual evidence

  1. ^ A b Hans Herbert Schulze: The Rororo Computer Lexicon. Rowohlt Taschenbuch Verlag, Reinbek bei Hamburg 1984, ISBN 3-499-18105-3 , page 253.
  2. ↑ Ulrich Mende: Software development for R / 3: Data Dictionary, ABAP / 4, interfaces. Springer Verlag, Berlin 1998, ISBN 3-540-62849-5 , page 136.
  3. ↑ Hans Heger: The history of machine data processing. Volume 1. Page 37, IBM Deutschland GmbH, Stuttgart 1990
  4. ↑ Franz Haurenherm: From the Hollerith machine for Computer: IBM Data Processing and Administration. Page 18f, Diplomica Verlag, Hamburg 2011 ISBN 978-3-8428-6889-2