Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Action Sequence Table Management

IP.com Disclosure Number: IPCOM000047482D
Original Publication Date: 1983-Nov-01
Included in the Prior Art Database: 2005-Feb-07
Document File: 2 page(s) / 14K

Publishing Venue

IBM

Related People

Shovelin, J: AUTHOR

Abstract

One of the enhancements provided for the IBM 8775 Data Entry RPQ permits an Action Sequence Table (AST) to be sent to a device incrementally, i.e., instead of loading a large AST at the start of the session (with subsequent delays), segments of the AST may be loaded as required. This is particularly useful where the path through the map group is dependent on the data. It may not be possible in advance to predict the path which will be taken through the map group(s) and may be unrealistic to attempt to deal with every possible map by creating a large AST since such an AST may be too large to be accommodated by the device. It is more efficient to load a "common subset" AST initially and then to load a small AST to handle the exception conditions of each subsequent map, as required.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Action Sequence Table Management

One of the enhancements provided for the IBM 8775 Data Entry RPQ permits an Action Sequence Table (AST) to be sent to a device incrementally, i.e., instead of loading a large AST at the start of the session (with subsequent delays), segments of the AST may be loaded as required. This is particularly useful where the path through the map group is dependent on the data. It may not be possible in advance to predict the path which will be taken through the map group(s) and may be unrealistic to attempt to deal with every possible map by creating a large AST since such an AST may be too large to be accommodated by the device. It is more efficient to load a "common subset" AST initially and then to load a small AST to handle the exception conditions of each subsequent map, as required. Each increment of an AST may redefine a key which was used in a previous increment, and it is this latest definition which prevails. Therefore, the chronological order of AST increments must be maintained. A further enhancement enables the application to specify that an AST is not owned by (and solely accessible from) a single partition but rather is owned by a device and accessible from every partition; this is known as Action Table Sharing. A device level AST has somewhat different deletion characteristics from partition level ASTs as it is impervious to the dynamic creation and destruction of partitions. The device level AST is only erased by a small number of conditions, specifically BIND, or by the application requesting a device level CLEAR before loading a different "common subset" for a new task. There is no restriction on the order in which device and partition level ASTs and their increments may be sent to the device. Since many ASTs and their increments may exist at partition and device level, a search order must be defined to locate the definition of a given key. The key for a field is defined in the extended field attribute. This extended field attribute must reside in either an explicitly created partition or in implicit partition zero. The partition identifier (ID) of the partition in which the field is defined is used to identify the group of AST increments and these increments are searched in reverse chronological order. If the key is not defined at the partition level, then the device level increments are searched in reverse chronological order. If the key is still not defined, then an exception condition is raised. For performance reasons no buffer management is peformed when an AST is received unless a CLEAR is requested. For this reason there is no difference between the first AST at a given level and any subsequent increments. This implies that there must be one control block entry per increment. Therefore, the fixed length control block of the IBM 8775 Data Entry RPQ now becomes a variable length control block, and the control block entry is expanded to include the ID of t...