System Specification Extractor for Legacy Code Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for understanding the relationship between input items and internal processing in information systems, such as those described in Japanese Patent Application Laid-Open Publication No. 1997-26897, are limited as they require source code analysis, which is not applicable when source code is unavailable during replacement development.
Innovation Solution
The technique associates input items with internal processing by analyzing differences in input and output data changes, specifically through the use of a system that classifies output data changes and relates them to corresponding processing operations, even in the absence of source code, using a device with a CPU, memory, and analysis program that includes a change analysis case generator, output data change information acquisition unit, and internal system processing specifying unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If source code analysis is used to understand system processing, then the relationship between input items and internal processing can be extracted, but this method cannot be applied when source code is unavailable
Solution Approach 1:
The patent creates a virtual model (copy) of the system's input-output behavior by executing the system with various inputs and observing output changes. This copy replaces the need for source code analysis, allowing specification extraction through behavioral observation rather than direct code examination.
Solution Approach 2:
The patent replaces the mechanical approach of source code analysis with an automated execution-based system. By substituting static code review with dynamic input-output observation and automated pattern recognition, the system can extract specifications without requiring access to the underlying code structure.
2Measurement precision
If different inputs are provided to analyze output data changes, then the relationship between input items and internal processing can be identified, but this requires multiple system executions and complex analysis
Solution Approach 1:
The patent performs preliminary classification of output data changes into categories (addition, deletion, update, no change) before associating them with input items. This preliminary organization of data reduces the complexity of subsequent analysis and enables more efficient pattern recognition across multiple execution results.
Solution Approach 2:
The patent uses feedback from observed output changes to refine the association between input items and processing operations. By iteratively analyzing how specific input changes correlate with specific output patterns, the system progressively improves its precision in identifying relationships while managing the time cost through intelligent sampling and pattern recognition.
Data Source
AI summary
The present invention has an object to analyze, regarding an existing system without a source code, the relationship between input items and internal system processing with respect to output items and present the analysis results to a user. An existing system processing specification extractor is an apparatus for making an existing system operate and analogizing input items of the existing system and internal system processing patterns without performing analysis of implementation and includes: an output data change information acquisition unit that classifies changes of output data associated with input to the existing system on the basis of types of changes of each output element; and an internal system processing specifying unit that combines the classifications obtained by making different input to the existing system and specifies a type of processing based on a usage pattern of an input item value.