Structured query language script analysis method, device and system

Abstract

The invention discloses an SQL script analysis method, an SQL script analysis device and an SQL script analysis system. The SQL script analysis method comprises: A, extracting SQL scripts from running logs according to the SQL script execution sequence in the running log; B, performing the morphologic, grammar and semantic analysis of each SQL script in turn and generating a semantic result collection of the SQL scripts; and C, analyzing the type of the context of the SQL scripts according to the semantic result collection, generating the context correlation information of all SQL scripts in the running log, and acquiring the data flow information of the running log according to the semantic result collection and the context correlation information. When the SQL script analysis method, the SQL script analysis device and the SQL script analysis system, which are disclosed by the invention, are used, by morphologic, grammar and semantic analysis of the SQL scripts and the analysis on the context correlation, the loss of data flow semantic information is avoided when the metadata of the SQL scripts are generated; and thus, the integrity and accuracy of the acquired metadata of the SQL scripts are improved and the acquisition quality of the metadata of the SQL scripts is guaranteed.

Description

Technical field [0001] The invention relates to a metadata management technology, in particular to a SQL script analysis method, device and system. Background technique [0002] At present, many large enterprises such as communications, finance and other industries have established metadata management systems to strengthen the effective management of the increasingly expanding data warehouse system. In these metadata management systems, most of them still manage SQL script metadata by manual sorting. Some companies have begun to try a SQL script metadata automatic acquisition program in recent years. This solution is based on compilation technology, trying to extract the SQL scripts of ETL and data processing, then perform lexical, grammatical and semantic analysis on these SQL scripts, and translate the obtained SQL script data flow information into SQL script metadata. [0003] The whole plan is divided into 3 links. In the first link, there are currently three main ways to obt...

AI Technical Summary

Problems solved by technology

Without these context information, it is difficult to ensure the quality of metadata generation when the subsequent processing links encounter context-related SQL scripts

[0012] 2. Metadata generation in link three lacks contextual analysis and processing methods

[0013] In link three, the existing solution is limited by link one on the one hand, and on the other hand, there is no comprehensive context environment to analyze the data flow semantics of SQL scripts in this link, and each SQL script is isolated and processed

When SQL scripts have context-sensitive features, metadata is often not generated correctly

[0014] To sum up, although the existing solutions can solve the problem of metadata management of SQL scripts to a certain extent, the existing solutions do not provide a mechanism to organically organize SQL scripts and related context auxiliary information, which may be lost Some data flow semantic information affects the quality of automatic acquisition of SQL script metadata

Images