Validation

Valid data

Data that is within the limits of what a user might sensibly or reasonably enter.

Validation

Automatic checking of data (checked by system), to ensure only possible, sensible, reasonable data is accepted.

How should good code approach validation?

Code should always be tested with data samples from as many of the following types: normal, boundary, abnormal, erroneous. It data is rejected, a message should explain why, and another opportunity to answer should be given.

Normal data

Data that is sensible and should be accepted.

Boundary data

Data that is right at the very edge of the accepted range.

Abnormal data

Valid data that falls outside of the accepted range.

Erroneous data

Data that is outside of the accepted range.

Range

Number must be in a specified range, e.g. between 0 and 100 inclusive.

Length

Length of inputted characters must match specified number of characters, e.g. 8 characters long / between 2 and 30 characters long

Type

The data inputted must be of the correct data type: it must follow rules related to whether it is numbers or letters, e.g. number of siblings must be an integer

Look-up

A specific character must be / must not be included, e.g. an @ in an email address / a % in name.

Often coded with lists of data

Format

Ensures the data entered is in the correct format required, e.g. an date as 00.00.0000

Presence

Ensures some data has been entered, e.g. a phone number in a form

Uniqueness

Used to check data such as an email address or username are unique within a system.

Where is validation written into?

Programmes for: Database data entry, Online data entry to forms on websites, Data entry to kiosks (at train stations / ATMs)

ANSWERING TEST DATA QUESTIONS

1. State the type of data (e.g. normal)

2. Give an actual example

3. State whether it will be accepted/rejected.

"Abnormal data would be 20 and it should not be accepted."