Python predictive analysis for bug detection

Abstract

Python is a popular dynamic language that allows quick software development. However, Python program analysis engines are largely lacking. In this paper, we present a Python predictive analysis. It first collects the trace of an execution, and then encodes the trace and unexecuted branches to symbolic constraints. Symbolic variables are introduced to denote input values, their dynamic types, and attribute sets, to reason about their variations. Solving the constraints identifies bugs and their triggering inputs. Our evaluation shows that the technique is highly effective in analyzing real-world complex programs with a lot of dynamic features and external library calls, due to its sophisticated encoding design based on traces. It identifies 46 bugs from 11 real-world projects, with 16 new bugs. All reported bugs are true positives. CCS Concepts •Software and its engineering → Software testing and debugging;

A comprehensive study on deep learning bug characteristics

“…The defect types include conceptual program elements which are function, check, documentation, assignment, algorithm and violation type. More recently, [23] presents a Python predictive analysis. Their tool is able to detect 46 bugs with 16 unreported before and they also classify those bugs base on their defect in the Python program.…”

A study on the changes of dynamic feature code when fixing bugs: towards the benefits and costs of Python dynamic features

“…Owing to the lack of declarations of variable types and allowing dynamic changes of class/instance structures in Python, type errors and attribute errors are common in Python programs. Xu et al [23] also verified that attribute errors, type errors, and unicode errors are very common in Python applications. isinstance, issubclass and type are often introduced to aid fixing bugs, but they also cause wrong type check bugs.…”

“…In recent years, researchers began to analyze the dynamic features in Python code. As a typical dynamic language, previous studies on Python language mainly include program slicing techniques [24,25], information flow analysis [26], type analysis [23,[27][28][29], polymorphism analysis [30], and alias analysis [31]. The first study on Python dynamic features was conducted by Holkner and Harland [2], which traced the usages of dynamic features in Python projects.…”

“…In our corpus, adding type check is the most frequent change behavior for fixing bugs. Owing to the lack of variable type declarations and the limitations of type inference techniques, Python programs tend to produce type errors [23]. Some reflection functions of Python, including isinstance, issubclass, and type, help check variable types dynamically.…”

Automated Unit Test Generation for Python

“…However, this dynamic nature has also been reported to cause reduced development productivity (Kleinschmager et al 2012), code usability (Mayer et al 2012), or code quality (Meyerovich and Rabkin 2013;Gao et al 2017). The lack of a (visible static) type system is the main reason for type errors encountered in these dynamic languages (Xu et al 2016a).…”