Text operations for a relational database programming language

Abstract

In this thesis we introduce a method of storing static text data, and algorithms for operations on the data, in a relational database programming language. We introduce a text data type, and implement relational algebra operations on the text data type. Our method stores the text data in text files external to the relations, and maintains pointers to the text data within the relations. Our algorithms minimize accesses to the actual text data so as to maintain the efficiency of database operations. We also implement a text processing mechanism where a text script can be joined to a relation, producing an individualized text script for each tuple in the relation. Our implementation includes queries involving pattern searches within text attribute values. All operations on text data are relational algebra operations, requiring the text data to be in relations, and returning relations as results.