Relational data model is the focus of a continues standardization process, and therefore we have extracted the standard physical db schema meta-models according to the specific SQL standard. Some of the generic db schema meta-models describe conceptual db schemas, like ER or EER db schema MM, while others describe logical db schemas, like relational or OR db schema MM. They may be seen as the models at the M1 level if the MOF stack begins with a data model meta-model (Table 1, MOF stack – Example 1), or as the models (meta-models) at the M2 level (Table 1, MOF stack – Example 2). The remaining three classes of meta-models are at the same abstraction level. Therefore, as may be seen in Table 1 (MOF stack – Example 1), at M2 level appear corresponding meta-models. In a generic approach we can assume that besides well-known data models, like relational, ER, EER or OR data models, we may specify specific data types, data structure types, constraint types for some SQL standard or vendor-specific data model. They are specific for relational or ER data model, respectively. For example, one may specify relational dm meta-model or ER dm meta-model, and they are containing constructs like data types, data structure types, constraint types, etc.
Data model meta-models stands at the M2 level of MOF stack. Hereof, we are distinguishing: i) data model (dm) meta-models ii) generic db schema meta-models iii) standard physical db schema meta-models and iv) vendor-specific physical db schema meta-models. We identify different kinds of db meta-models that describe db models at certain abstraction level. The work we describe in this paper unifies two main research areas: database design and implementation and meta-modeling in the context of MDSE. An example aimed at better explanation of the concepts that are introduced in Section 3, is presented in Section 4. Section 3 is devoted to a relational database schema meta-model organized in several packages. In Section 2 a classification of db meta-models is presented. Apart from the Introduction and Conclusion, the paper has four sections. The EMF meta-modeling language is based on the Ecore meta-meta-model which is closely aligned with the Essential MOF (EMOF) specification. In the purpose of specifying and managing our meta-model we decided to use the Eclipse Modeling Framework (EMF), a current MOF-like modeling environment.
In this paper we propose a meta-model of relational database schema. One of the solutions is meta-modeling of db models. Therefore, even the mapping between SQL db schemas extracted from RDBMS data repository of different vendors may be a serious problem.
DBSCHEMA PORTABLE FULL
However, issues of SQL code portability between major RDBMS products still exist due to a lack of full compliance with the standard and proprietary vendor extensions. It is also the focus of a continuous standardization process, resulting in SQL standards (the current revision is: SQL:2011, ISO/IEC 9075:2011). Structured Query Language (SQL) is currently available in most commercial and open-source DBMSs. Starting from physical database schema, recorded into RDBMS data repository, the conceptual db schema (ER or EER db schema) or logical db schema (based on the relational data model) could be extracted. Relational or OR databases are a common source of reverse engineering. It generally includes some form of reverse engineering followed by some form of forward engineering or restructuring.
DBSCHEMA PORTABLE SOFTWARE
One of the key concepts in software maintenance is re-engineering. Additionally, the growth of eXtended Markup Language (XML) technologies has led to the need to have object-oriented (OO) wrappers for XML data and the translation from nested XML documents into flat relational databases and vice versa. and relational schemas to generate object wrappers iii) mapping between data sources and a mediated schema to drive heterogeneous data integration iv) mapping between a database schema and its next release to guide data migration or view evolution, etc.
0 kommentar(er)
