Rule-based data processing
Not only can the RapidRep Reporting Suite create reports but it can also implement requirements for the back-end. The range of possible fields of application results from the scope of functions that a report definition can provide. The ability to include sets of rules in the calculation is a great help in the implementation of requirements in data processing. The RapidRep rule engine can process sets of rules in the form “If then” when implementing calculation specifications.
Typical fields of application
- Implementation of ETL processes
- Rule-based implementation of data interfaces
- Implementation of data cleansing
- Implementation of specifications for the calculation of business figures based on sets of rules
- Production of expected results within the framework of a rule-based test execution (for the validation of rules)
Enterprises of all industries are constantly changing. Global competition and ever shorter innovation cycles require a high degree of adaptability. In addition, there are new legal regulations which affect entire industrial sectors.
An example is the financial sector, which is from experience particularly strongly regulated, for example with Basel III or withholding tax.
These changes do not only effect business processes. The impacts on the IT and especially the information systems are often very high as well. Whoever can implement these changes quickly and reliably has clearly an advantage over his competitors.
The solution introduced here shows how companies can ensure that necessary changes to the company data can not only be realised quickly but above all quality assured.
The IT processes can be different from company to company and vary over time. In practice, the V-Model is widely used but agile software methods are more and more put to use as well.
All methods share the fact that business requirements are implemented step by step by IT tools and the principal reviews in the end whether the program matches these requirements.
The illustration shows the most common case of data processing. The program processes input values from N different source systems and produces M results or interim results.
Programs of data processing transform input data according to a deterministic logic into results, which are stored to one or more output tables. Call parameters are used for the control of the program logic, which are accessed in the program code.
The results in the M target objects are derived functionally from the input data and the parameters used. Otherwise the results of the data processing would not be deterministic and depend on chance.
RapidRep makes a valuable contribution in the implementation of programs which process data.
Sets of rules play an essential part in the specification for implementation (DP concept) because they unmistakably depict the functional aspects, which are shown in the figure above as group of functions f1-fM.
The rules in these sets are declarative and precise. The principals, usually the business department of a company, can read and understand these rules without IT knowledge and even help with their creation.
Development is faster and less prone to errors
Rules in the RapidRep rulebooks are unambiguous and do not leave room for interpretation. Programmers can faster and with significantly less errors implement unambiguous specifications than it is the case with textual input.
Automated test evaluation with the RapidRep Test Suite
With help of the same rules that the IT also uses for implementing the programs, RapidRep is able to determine the target result.
The RapidRep Test Suite contrasts the target results with the actual results produced by the IT. Verification then determines whether the program implements all rules correctly or not.
The costs that are saved when implementing programms in the data processing in the way presented here are considerable. Our project experiences have confirmed this impressively several times.