Source Code Property Verification via Grouped Slicing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software verification methods are inefficient and costly due to the need for repeated analysis of code segments, especially when verifying properties in large-scale concurrent systems, as they lack scalable and time-efficient techniques for grouping related properties.
Innovation Solution
A system and method for verifying source code properties by employing grouping techniques such as slice subsumption, data flow dependence, and data and control flow dependence to merge properties into disjoint groups, allowing for simultaneous verification of related properties and reducing repeated analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If properties are verified individually using conventional methods, then verification completeness is achieved, but verification time and computational cost increase significantly due to repeated analysis of common code segments
Solution Approach 1:
The patent merges multiple properties into groups based on slicing criteria (same procedure, same loop, same function). By combining properties that share common code segments into single verification groups, the system avoids repeated analysis of identical code portions while maintaining complete verification coverage across all properties.
Solution Approach 2:
The patent segments the verification process by dividing properties into distinct groups based on their code location and dependency relationships. Each group is verified independently as a unit, allowing the system to process multiple properties simultaneously without redundant analysis, thus reducing overall verification time while preserving completeness.
2Adaptability or versatility
If verification is performed on large-scale concurrent systems with numerous properties, then comprehensive property coverage is achieved, but scalability and efficiency deteriorate due to the exponential increase in analysis complexity
Solution Approach 1:
The patent combines multiple properties into verification groups based on slicing criteria including procedure identity, loop identity, and function identity. This merging approach allows the system to handle large numbers of properties by processing them in consolidated groups rather than individually, maintaining scalability while achieving comprehensive property coverage in concurrent systems.
3Measurement precision
If conventional verification methods are used without property grouping, then individual property verification accuracy is maintained, but overall verification cost and computational resources increase due to repeated analysis of common code
Solution Approach 1:
The patent merges properties into groups based on slicing criteria while preserving the accuracy of individual property verification. By analyzing common code segments once and reusing results across multiple properties within a group, the system maintains verification accuracy for each property while significantly reducing computational complexity and resource requirements.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
This disclosure relates generally to method and system for verifying properties of source code. Verifying sufficient subset of properties by identifying implication relations between their verification outcomes is time consuming because of increased size of source code with large number of properties. The proposed disclosure processes the received source code for verifying properties by analyzing the source code to merge the plurality of properties into a plurality of groups based on a grouping technique. Then, slices for each group among the plurality of groups are created. Further, each slice created for each group is verified; verification for the one or more properties within each group is performed simultaneously. The system groups the properties thereby providing efficient and scalable system for verifying properties which reduces cost with increased efficiency and improved performance.