Browse Prior Art Database

Transparent Data Structure Changes

IP.com Disclosure Number: IPCOM000079281D
Original Publication Date: 1973-Jun-01
Included in the Prior Art Database: 2005-Feb-26
Document File: 6 page(s) / 63K

Publishing Venue

IBM

Related People

Sanborn, JL: AUTHOR

Abstract

In a large file-based system with a complex file structure definition, it is impossible to define a permanent structure which will support all present and future requirements of the system. Described herein are methods which will handle two types of data structure changes, without a mandatory conversion of all active files.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 24% of the total text.

Page 1 of 6

Transparent Data Structure Changes

In a large file-based system with a complex file structure definition, it is impossible to define a permanent structure which will support all present and future requirements of the system. Described herein are methods which will handle two types of data structure changes, without a mandatory conversion of all active files.

A typical file consists of a number of different item types to define different sets of data fields, and a name field for each usage of an item type to distinguish different copies of the same type of data. The physical storage of the data entries is not pertinent to this description and will not be discussed. Rather, the way in which a data entry appears after retrieval from the file will be discussed.

A typical data entry from the file has four major sections:
1) The data entry length, which gives the total length of the

current entry, including the two-byte field containing the

length.
2) A set of fixed-length fields which are required for every

data entry handled by the access method.
3) A fixed area consisting of a set of fixed fields which are

defined for a particular item type, and are always the same

length for an item type in a particular file.
4) A variable area which may contain a number of variable

fields defined for an item type, but which may also be

completely absent from a particular entry.

There are four required fields in every data entry. The length of these fields is fixed for all entries to be processed by the system access method:
Item type - A one-byte field which distinguished this entry from another entry with the same name, but a different set

of data fields.

DESC - A two-byte field which provides a selection criterion for different variations of the same data entry.

USN - A two-byte field which identifies the level of a particular data entry and is used to carry history on

the file.

FLAGS - A one-byte field which identifies the type of entry in the file, e.g., add or delete.

Each item type defined in the data structure has a number of fixed fields (sometimes overlapping), which are considered present in every data entry with this item type. These fields may be of two varieties - a fixed length for every file, and a length which is variable between files but is fixed for every occurrence in a single file. The only changes currently possible in the definition of these fields in an existing item type without file conversion are; 1) to add a field with zero length in existing files and nonzero length in new files; 2) create subfields which correspond to the length of an existing field; and 3) replace an existing field with a new field (or fields) if the existing field has never contained any data.

1

Page 2 of 6

The fixed fields by item type may be grouped into three sets:

Naming fields - The first set of fixed fields overlaps with

the naming field, which is used to distinguish

between data entries with the same values in

the required fields, including ite...