S-NAR: generating narrations of SQL queries using regular expressions

Abstract

Structured Query Language (SQL) is popular with relational databases. Despite the simple and highly structured nature of SQL, end users often find it difficult to comprehend written queries and/or write new queries. Hence, a number of tools have been proposed to aid the comprehension of SQL queries via visualisation. In this paper, we present a different comprehension aid called narrations. Narrations are non-technical, textual descriptions of SQL queries. Using a representation of regular languages, regular expressions, we have designed a tool, called S-NAR (or Sql-NARrator), that reads queries and generates a narration of these queries using pre-defined templates. S-NAR was tested with 5,000 queries and narrated 96% of the queries. The generated narrations should find applications in aiding comprehension in the teaching of database courses at higher institutions of learning, and assisting the comprehension of legacy queries in data-centered industries.