Relational database reverse engineering: Algorithms to extract cardinality constraints

摘要

This paper presents algorithms which build SQL queries to improve the reverse engineering of relational databases. Our process extracts the current cardinality constraints of n-ary relationships through a combination of data dictionary, data schema and data instance analysis. Our results can be applied to semantic data models with cardinalities constraints based either on the ER model or to semantic data models with cardinality constraints on participation constraints (MERISE, ECR, ERC +, OMT, ODMG). The process we propose can also refine conceptual diagrams of commercial tools with reverse-engineering options (AMC∗Designer™, ORACLETM Designer 2000, etc.). From our algorithms a PRO*C program using dynamic SQL has been implemented.