File Organisation (Cambridge International A-Level Computer Science)

What is file organisation?

• Files store data permanently on secondary storage as a collection of records, where each record represents one item (a customer, a transaction, a product)
• File organisation is the way those records are arranged within the file
• The method chosen affects how quickly records can be searched, added, updated and deleted, so it is picked to suit the task

Choosing a method

• There are three organisation methods you need: serial, sequential and random
• The organisation you choose also decides which access methods are possible (see File Access)

File organisation methods

Serial file organisation

• Records are stored one after the other, in the order they arrive (chronological order, no sorting)
• New records are appended to the end of the file
• Simple to implement, but slow to search because there is no order to exploit

Sequential file organisation

• Records are stored in sorted order, based on a key field
• To update the file, a new version of the file has to be created (records cannot easily be inserted in place)
• More efficient to search than serial, because a search can stop once the key is passed

Random (direct) file organisation

• Records are stored in no particular order; there is no sequencing in where a record is placed
• There is a relationship between the record's key and its location: a hashing algorithm is applied to the key to work out the address
• Updates can be carried out directly, without rewriting the file

Summary

Exam Hint

• For serial, both "records stored in the order they arrive" and "new records are appended to the end" are separate mark points. Give both.
• Do not confuse organisation (how records are arranged) with access (how records are read). A question can ask about either, so keep the two ideas separate in your answer.