INFSCI 1022
Database Management Systems

Syllabus: Term Project

The project for this course involves building a simple, single user relational database in a structural programming language, such as C or Pascal. The database will consist of a collection of routines that will be accessible through a simple interactive user interface. The specification of these routines follows below. You may, although you do not need to, add other features to your database. Please make sure that you have implemented the required parts first. The deliverables include the source code in electronic and printed form and a report describing your system and its capabilities. The project will be done in groups of normally four students. The groups will be formed during the first class meeting. Each group will prepare a written statement listing its members, the group leader, and the projected responsibilities of each individual member. About a week later, the group will submit a written design of the system. From that point on, the group will proceed with implementation. Each member of the group carries the responsibility for the success or failure of the entire group. Individual group members will have the opportunity to evaluate the contribution of each of their colleagues in the group to the final result. To make your project successful, you will have to meet regularly and coordinate your activities. The due dates for the project statements and the completion of the project are listed in the syllabus. Projects will be demonstrated during one of the last classes. The best project will be rewarded by a cup of coffee (if you do not drink coffee, a cup of tea, a Coke, or equivalent) and a doughnut for each member of the group and an accompanying certificate.

Project Specification:

The following functions have to be included in your database:

Operations on relations

The database contains a currency register that points to one relation in the database that is considered to be current. This is useful in operations on tuples and allows a user to omit references to a relation (all operations refer to the current relation). There is no current relation when the database is started.

CREATE relation-name

Creates a new relation named relation-name. Prompts the user to interactively input the attribute names and types of the created relation. Attributes specified by the user may contain two indicators:

• An asterisk before the attribute name (e.g., *attribute-name) denotes that the attribute is the primary key of the relation.
• A hash followed by a relation name and an attribute name (e.g., #relation-name0:attr-name0) specifies a referential integrity constraint: the specified relation (relation-name0) and the attribute name (attr-name0) must exist (i.e., must have been created prior to creating the current relation).

The relation relation-name becomes current.

DELETE relation-name

Deletes the relation named relation-name from the database. This operation does not affect the contents of the currency register, unless relation-name was current, in which case there is no current relation after performing delete.

CURRENT relation-name

Makes the relation named "relation-name" current.

Operation on tuples

Similarly to the currency pointer for the relations, there is a currency pointer for the current tuple in each of the relations. This is useful in operations on tuples such as display. There is no current tuple when a relation is created.

insert

Prompts the user to enter values of the attributes of the current relation and inserts the complete tuple into the current relation.

first

Makes the first tuple in a given relation the current tuple.

next

Makes the tuple that is next after the current tuple the current tuple.

previous

Makes the tuple that is right before the current tuple the current tuple.

last

Makes the last tuple in a given relation the current tuple.

remove

Removes the current tuple from the current relation.

modify

Prompts the user interactively to modify the values of the attributes of the current relation and replaces the original tuple with the modified tuple in the current relation.

Input-output

read file-name

Reads the tuples of the current relation from a disk file rather than interactively from a keyboard.

write file-name

Writes all tuples of the current relation to a disk file.

Querying

SELECT attr-name1 operator attr-name2|constant from relation-name1 giving relation-name2

Creates a new relation relation-name2 consisting of tuples from relation-name1 that satisfy the condition attr-name1 operator attr-name2|constant (e.g., name="Smith") (this is a regular relational select operation). Relation-name2 becomes current.

project relation-name1 on attr-name1 attr-name2 ... giving relation-name2

Creates a new relation relation-name2 consisting of the attributes attr-name1 attr-name2 of all tuples of the relation relation-name1 (this is a regular relational project operation). relation-name2 becomes current.

join relation-name1 relation-name2 where attr-name1 =attr-name2 ... giving relation-name3

Creates a new relation relation-name3 consisting of tuples from relation-name1 and relation-name2 such that their attribute values specified are equal (this is a regular relational equijoin operation). relation-name3 becomes current.

Evaluation criteria:

Your database has to be functional - this is a necessary condition for completing the project. Your project will be judged on the software design methodology (e.g., data organization and structure, whether you used indexing or any efficient access mechanism), the flexibility of your database (e.g., limitations on the number of attributes each relation can have, robustness to the input format), clarity of the messages generated by your system (including error messages), maintenance of integrity constraints, on-line help provided to the user, and the manual that you will provide with your database.

INFSCI 1022 main page
Marek Druzdzel's teaching page
Marek Druzdzel's home page