FILE ORGANISATION AND FILE ACCESS

Review:
Storage devices can be categorized according the manner
in which they access data. There are two types of access:
1. Sequential access (the retrieval of data in the order in
which it was stored). Example of sequential access
storage media - MAGNETIC TAPE. The data stored
on magnetic tapes have to be read in sequence. It is
impossible to jump the middle of the tape to find a
particular data item.
Random access (the ability of the computer to go
directly to a specific piece of data on the storage
media without having to access any other data).
Examples of direct access storage media Most other
secondary storage media such as floppy disk and hard
disk that allow software medium without having to
access any other data.
Files of Records
In any computer system data is stored in files. There are
different types of files:
Text files contain just the characters that you
include in them.
Word Processing files contain text but they also
contain how the document should be laid out.
Files of records.
A file of records is the simplest form of database. A
record is a collection of all the data about a single item in a
file. For example, the details of the members of a club may
be held in a file. Each record would contain the data about
just one club member. The file would hold a record for
each member of the club.
The data about each member would fall under a series of
headings known as fields. The fields could include
surname, first name, and the date when they joined the
club.
Each row is a record in a file. Each record will have data
in one or more of the fields. Files of records are used by
many different software applications. For example, the
phone directory on a cell phone is a file of records.
File Organisation
Three main methods of file organisation are:
1. Sequential
2. Indexed-Sequential
3. Random

Sequential files can be used with sequential access or direct

access media. Indexed-sequential and random files can
only be used with direct access media.
Sequential File Organisation
In a sequential file, all the records are stored in order. That
means that the records have been sorted in some way. One
of the fields will be used for ordering, that is sorting the
file. This field is known as a key or key field.
Sequential files are searched using sequential access. That
means that if you want to find a record in a sequential file
you have to start at the beginning and work your way
through until you find it. (N.B. Serial files are files with
unordered records stored in a sequential manner while
sequential files are files with ordered records stored in a
sequential manner.)
Indexed-Sequential File Organisation
If a file is quite large you may have to wait a long time
while the system searches through the records if a
sequential search is used. An indexed-sequential file is
also a file that has been sorted in some order. Another file
is also set up to act as an index.
Letter
A
B
C

W
X
Y
Z

Position
1
20
49
437
468
469
475

The Indexed file

Position
437
438
439
Etc.

Surname
Waldron
Walker
Walsh

First Name
Pauline
George
Joel

Date Joined
14 Apr 1990
23 Aug 1985
1 Sep 1989

The Sequential file

Supposed the club file mentioned above has very many
records. An index file could be set up (see first table). The
letter field refers to the first letter of the names in the
sequential file. The position field states where the first
surname beginning with the letter is to be found. So if you
wanted to search for the surname Williams, the system
would first search through the index file to find the letter

W. It would jump to that position in the main sequential

file. It would then search though all the surnames
beginning with W until it found Williams. The indexedsequential file method makes it possible to search large
sequential files very quickly.
Indexed-sequential file organisation cannot be used with
sequential access media such as magnetic tape. It can only
be used with direct access media such as disks. The main
sequential files containing all the records could be stored
on a hard disk.
Random File Organisation
Like indexed-sequential files, random files can only be
used on direct access media, such as disks. Random file
organization is the most efficient way of storing extremely

large files, such as national databases. Random files use a

rather clever method of providing direct access to records.
Each record will have its own specific position (or address)
on the disk but the records may not be stored in sorted
order at all. The position allocated to each record will be
calculated by a special formula called a hash formula. This
formula will use some of the data in the record and then
convert it into the address of a position on the disk.
When a new record is added to the file, the formula works
out the address where it should be stored on disk. When
someone wants to search for a record, once again the
formula will work out where to find it.
This method means no time is wasted when searching for a
record since the system can go directly to the right record.

Choosing appropriate file organisation

When new application software is being developed, the software engineers and programmers have to decide on the
method of file organization that they will use. There is rarely one way of solving the problem, but here are some typical
examples.
A payroll is a list of all the employees in an organization, together with their pay details. Every month a payroll
application will calculate how much each person should be paid. It will print payslips and it may send the details of
payments directly to the companys bank. The payroll file would normally be stored sequentially by surname. When
the application is run each month it will access the record for each employee one at a time, so a simple sequential file
is needed.
A shop uses a computerized stock system. The stock records hold details of all the items for sale in the shop. Each
item is identified by a stock code. The stock record for each item also holds the price of the item and the number in
stock. When a customer wants to buy an item the sales assistant enters the stock code at a point of sale terminal. (A
barcode reader can be used to scan in the stock code from a label on the item.) The stock system then finds the record
for the item and sends the price to the point of sale terminal. It also updates the stock record to show that the number
in stock has gone down. There will be a large number of records in the stock file, and each item has to be retrieved
quickly, so the best way to organize the file would be as a random file.
A library has a computer-based catalogue of all its books. The catalogue is sorted by the ISBN (International Standard
Book Number) which you can find on the back of any book that has been published. There is usually a barcode that
gives the ISBN as well. When someone borrows the book the librarian scans the ISBN barcode and this brings up the
details of the book on screen. The catalogue file is large but is stored sequentially. Index-sequential file organization
will give fairly quick access to any record.

(From Heinemann Information Technology)