Clustered Code Warning Consolidation via Overlapped Function Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Static analysis of large-scale software applications generates a high number of warnings due to imprecise analysis and conservative inter-cluster communication, leading to increased manual review efforts, with existing solutions failing to efficiently address redundancy and scalability issues in clustered code analysis.
Innovation Solution
A computer-implemented method and system that identifies common point warnings and unique warnings in clustered software code, groups them based on the top-most overlapped function, and assigns this function as a constraint for review, eliminating redundant warnings by applying the review output across associated clusters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If static analysis is performed on large-scale software applications, then code coverage is improved, but the number of warnings increases due to imprecise analysis and conservative inter-cluster communication
Solution Approach 1:
The patent merges multiple cluster analysis results by identifying common point warnings that appear across multiple clusters and consolidating them into single warning entries. This combining approach reduces the total number of warnings while preserving the precision benefits of analyzing each cluster separately.
2Measurement precision
If all warnings are reviewed manually to determine actual defects, then analysis precision is improved, but manual review time and effort increase significantly
Solution Approach 1:
The patent combines multiple warnings that point to the same program location into a single consolidated warning entry. This merging reduces the total number of warnings that require manual review while maintaining the ability to identify actual defects through the consolidated information.
Solution Approach 2:
The patent creates a universal review mechanism where a single warning entry can represent multiple cluster contexts. This multi-functional warning system allows reviewers to assess one consolidated warning that encompasses information from multiple clusters, reducing review effort while maintaining comprehensive defect detection.
3Productivity
If code is divided into multiple clusters for scalable analysis, then analysis scalability is improved, but the number of warnings increases due to conservative inter-cluster communication
Solution Approach 1:
The patent merges warnings from multiple clusters by identifying common point warnings that appear across cluster boundaries. This consolidation approach maintains the scalability benefits of cluster-based analysis while reducing the warning proliferation caused by conservative inter-cluster communication analysis.
Data Source
AI summary
A system and method for reviewing of warning generated during static analysis of a clustered software code by identifying, common point warnings and unique warnings from warnings associated with a software code, and further identifying, a top of must overlapped function for each of the common point warnings. Generating, one or more groups of the common point warnings based on the top of must overlapped function, and assigning, the top of must overlapped function as a constraint for corresponding group of common point warnings. Eliminating, warnings from each of the one or more groups using a review output wherein the review output is identified by reviewing a common point warning from the one or more group under the constraint such that the review of the common point warning under the constraint satisfy for review of all remaining for reviews of all remaining common point warnings of the group.