Context2Name: A Deep Learning-Based Approach to Infer Natural Variable Names from Usage Contexts

Abstract

Most of the JavaScript code deployed in the wild has been minified, a process in which identifier names are replaced with short, arbitrary and meaningless names. Minified code occupies less space, but also makes the code extremely difficult to manually inspect and understand. This paper presents Context2Name, a deep learning-based technique that partially reverses the effect of minification by predicting natural identifier names for minified names. The core idea is to predict from the usage context of a variable a name that captures the meaning of the variable. The approach combines a lightweight, token-based static analysis with an auto-encoder neural network that summarizes usage contexts and a recurrent neural network that predict natural names for a given usage context. We evaluate Context2Name with a large corpus of real-world JavaScript code and show that it successfully predicts 47.5% of all minified identifiers while taking only 2.9 milliseconds on average to predict a name. A comparison with the state-of-the-art tools JSNice and JS-Naughty shows that our approach performs comparably in terms of accuracy while improving in terms of efficiency. Moreover, Context2Name complements the state-of-theart by predicting 5.3% additional identifiers that are missed by both existing tools.

Variable Name Recovery in Decompiled Binary Code using Constrained Masked Language Modeling

“…In a recent work, CONTEXT2NAME [4] attempted to assign meaningful names to the identifiers based on the context of minified JavaScript codes. They were able to successfully predict 47.5% of meaningful identifiers on 15,000 minified codes using recurrent neural networks.…”

“…While researchers have taken steps in predicting variable names in high-level programming languages [1,3,4,26,41,45], it is worth noting that inferring variable names in decompiled binary code poses a unique set of challenges. High-level programming languages like Java, Python, and JavaScript are syntactically rich: Variable types are preserved in these languages, while they are usually eliminated in binaries.…”

Modeling Vocabulary for Big Code Machine Learning

“…Vasilescu et al [70] describe an approach to recover original names from minified JavaScript programs based on statistical machine translation (SMT). Bavishi et al [11] accomplish this using a deep learning-based technique. Jaffe et al [37] generate meaningful variable names for decompiled code by combining a translation model trained on a parallel corpus with a language model trained on unmodified C code.…”

“…We find that adding words found in strings and comments appears to have little impact on BPE 5K and 10K, both of which slightly increase the size of the corpus by 1-2%. A vocabulary of 10K words is more than 1,000 times smaller than the initial configuration (11,357,210), at the cost of increasing the number of tokens in the corpus by a factor of 1.7.…”

Nalin: Learning from Runtime Behavior to Find Name-Value Inconsistencies in Jupyter Notebooks

“…However, we are the first to focus on name-value inconsistencies, whereas prior work targets other kinds of problems. Nalin also relates to learned models that predict missing identifier names [12,17,48]. Our work differs by analyzing code with names supposed to be meaningful, instead of targeting obfuscated or compiled code.…”