A Study On Temporal Extension And Query Rewriting Technology Of SQL

With the rapid development of information technology and universal application ofcomputer technology, there are increasing demand for dealing with the data of temporalinformation. However, the traditional relational database can only reflect the development ofone object in the entire process at a certain moment state (snapshot), it could not meetpeople’s temporal data manipulation well due to the operating characteristic of itself. So today,there is a new idea being proposed: setup temporal database to meet the treatment requirementof temporal data. The theories of Temporal Database provide a better solution for thissituation. Over the past20years, researches on temporal database were very fruitful.Thousands of papers and literatures have been published in this field. At the same time,though temporal data processing technology has come into application era, almost no DBMS(Database Management system) with all functionalities of temporal data processing have befound in practice until now.The temporal operation language has many versions in the past two decades, such asTempSQL, TQuel, TSQL2and ATSQL2.The ATSQL2is based on the standard SQL, whichnot only includes the bitemporal query language, but also includes bitemporal modification,data definition and constraints language. The TimeDB of bitemporal relational databasemanagement system realized the ATSQL2temporal operation language. Firstly, this paper isbased on ATSQL2, explains the grammar of it, analyzes its syntax and makes improvementson its time arithmetic operation. For instance, the period comparison predicates (precedes,overlaps, meets, contains, etc)of ATSQL2,whose results are Boolean type,so these predicatesare only used in "where" clause and lack some temporal comparison predicates such as aboutsimultaneous,and separates asynchronous and etc. Nevertheless, in the practical queryapplication, we sometimes need to get query results as period, or more complex, flexibletemporal operation, but the ATSQL2can only complete these operations by more complexquery statement. So this paper discusses the way of expanding comparison operation betweenperiod and instant to enrich temporal arithmetic operational semantics and make expressionalability stronger. Secondly, this paper focus on the use of layered extending relational database management system (RDBMS) to implement the application of temporal data. According tothe study of the algorithm of transforming ATSQL2to standard SQL and adding the "temporalquery language transform layer" to implement ATSQL2query rewriting, the standard SQLstatement can be directly implemented on RDBMS.This method not only encapsulates somedetails of treatment of the time dimension, but also improves the efficiency of the programmerto operate the application of the temporal data. All these work makes a basic foundation forthe temporal database implementation.