An approach and its specification to data reformatting in data conversion

摘要

Those data conversion cases which require both data restructuring and data reformatting at the same time have been proven to be the most complex and troublesome conversion cases. In this paper we address this issue. Our approach contains those factors in terms of which conversion process can be simplified in complex data conversion cases.We concentrate on data reformatting aspects. Data reformatting is a process which changes or reformates the representations of stored data so that they can be processed correctly in a new hardware/software environment. We show that the load and unload utilities of file and database systems provide a meaningful interface between the actual conversion process and the software available in the source and target environments. The complexity of data reformatting can be reduced considerably by utilizing load and unload utilities effectively. In spite of using the load and unload utilities, stored data contain some data base management system (dbms) and file system oriented characteristics which must be reformatted. To reformat these characteristics we define exact encoding functions in terms of which bit string representations of flat files and hierarchical databases can be built up.In this paper we also connect the data reformatting aspects to our previous data restructuring specification for flat files and hierarchical databases. In other words, after this extension we have a precise specification for complex data conversion cases. In this paper we show that the attribute method (attribute grammars) provides a precise and convenient tool for this kind of software specification.