Date of Award:

2016

Document Type:

Thesis

Degree Name:

Master of Science (MS)

Department:

Computer Science

Advisor/Chair:

Curtis Dyreson

Abstract

Sequenced and nonsequenced semantics are the two previously researched semantics for the evaluation of an operation in a temporal database such as a query or data modification. Sequenced semantics evaluates an operation in each time instant using only the data alive at that time. Nonsequenced semantics, in contrast, means that an operation explicitly references and manipulates the timestamps in the data.

In this thesis we propose a new framework that shows both semantics are variants of a general temporal semantics. We present the general semantics and show how additional semantics, such as preceding semantics can be realized. The semantics are specified using annotations.

The primary contribution of this theses is the translation from temporal SQL to nested SQL. We focus on SQL's SELECT statement, which is used to query data. Temporal SQL is SQL annotated with temporal semantics. Nested SQL is SQL for non-1NF data, with additional operations, such as COGROUP and FLATTEN to create and un-nest, respectively, bags of tuples (non-1NF data). This thesis develops a denotational semantics for translating from temporal to nested SQL. We implemented the denotational semantics for an SQLite ANTLR grammar, and the thesis also reports on the implementation.

Checksum

62a6b940d6d4434119de7c32bfa4152c

Share

COinS