The main purpose of current data analysis is to model the data used by one or more current systems. Data is represented in the current system both in a structured form, as in current data stores, and as data in user views, forms, reports and displays.
You begin with data stores, then move to these other views of data. The result is an implied data model that may describe one or more data stores, or the total view of data supported by one or more current systems. There may be a single current system model, but if systems are known or found to be incompatible, there may be more than one.
The models are assembled by asking a series of simple questions about each data item to ensure that it is dependent on the appropriate identifier or key. The questions are those used in the normalization technique described in the chapter "Analyzing Data."
The current system model is built up progressively. This technique is sometimes called canonical synthesis because it proceeds by progressively adding together two or more models of the data.
The models may be represented graphically or as a list. The method suggested here is to use CA Gen to develop a separate model of the data, using the diagram or entity list as convenient.
If the model is to be used to generate procedures that access current data, or procedures for data conversion or bridging, then in addition to the synthesized (implied) data model, a data store analysis model is retained.
Note: You must be careful not to remove redundant record types, identifiers, and fields from such a model.
|
Copyright © 2014 CA.
All rights reserved.
|
|