On the fly synthesis of edit suggestions

Miltner, Anders; Gulwani, Sumit; Le, Vu; Leung, Alan; Radhakrishna, Arjun; Soares, Gustavo; Tiwari, Ashish; Udupa, Abhishek

doi:10.1145/3360569

Cited by 52 publications

(38 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Prior to PROSE, authors of inductive synthesizers such as FlashFill [19], FlashExtract [20] had to implement their synthesizers from scratch. Newer synthesizers such as Refazer [21], BluePencil [15] leverage PROSE to speed up the development. We follow the same strategy.…”

Section: Synthesizing Programs From Examplesmentioning

confidence: 99%

“…With this motivation, we conduct a large-scale empirical study of textual merge conflicts that occur in Microsoft Edge (a fork of the Chromium browser), and apply program synthesis (in particular, programming by example (PBE) [14], [15]) to illustrate the feasibility of learning patterns for a large class of resolutions. We chose Microsoft Edge as a subject due to (a) the collaborative nature of the project with 500+ developers and (b) the large number of merge conflicts that manifest as a result of merging changes from the Chromium repository to this fork.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Can Program Synthesis be Used to Learn Merge Conflict Resolutions? An Empirical Analysis

Pan

Nagappan

et al. 2021

2021 IEEE/ACM 43rd International Conference on Software Engineering (ICSE)

Self Cite

View full text Add to dashboard Cite

Forking structure is widespread in the open-source repositories and that causes a significant number of merge conflicts. In this paper, we study the problem of textual merge conflicts from the perspective of Microsoft Edge, a large, highly collaborative fork of the main Chromium branch with significant merge conflicts. Broadly, this study is divided into two sections. First, we empirically evaluate textual merge conflicts in Microsoft Edge and classify them based on the type of files, location of conflicts in a file, and the size of conflicts. We found that ~28% of the merge conflicts are 1-2 line changes, and many resolutions have frequent patterns. Second, driven by these findings, we explore Program Synthesis (for the first time) to learn patterns and resolve structural merge conflicts. We propose a novel domain-specific language (DSL) that captures many of the repetitive merge conflict resolution patterns and learn resolution strategies as programs in this DSL from example resolutions. We found that the learned strategies can resolve 11.4% of the conflicts (~41% of 1-2 line changes) that arise in the C++ files with 93.2% accuracy.Index Terms-merge conflict, program synthesis, automated fixing This work has been done when Nachiappan Nagappan was an employee at Microsoft and Rangeet Pan was an intern at Microsoft.

show abstract

Section: Synthesizing Programs From Examplesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Can Program Synthesis be Used to Learn Merge Conflict Resolutions? An Empirical Analysis

Pan

Nagappan

et al. 2021

2021 IEEE/ACM 43rd International Conference on Software Engineering (ICSE)

Self Cite

View full text Add to dashboard Cite

show abstract

“…-The work [26] identifies repetitive edits to text documents by keeping track of a graph of edits and suggests automation rules for SW robots. While this work focuses on supporting expert users in the manual development of SW code, SmartRPA is targeted to automatically generate executable scripts for SW robots.…”

Section: Related Workmentioning

confidence: 99%

Automated Generation of Executable RPA Scripts from User Interface Logs

Agostinelli

Lupia

Marrella

et al. 2020

Lecture Notes in Business Information Processing

View full text Add to dashboard Cite

Robotic Process Automation (RPA) operates on the user interface (UI) of software applications and automates -by means of a software (SW) robot -mouse and keyboard interactions to remove intensive routine tasks (or simply routines). With the recent advances in Artificial Intelligence, the automation of routines is expected to undergo a radical transformation. Nonetheless, to date, the RPA tools available in the market are not able to automatically learn to automate such routines, thus requiring the support of skilled human experts that observe and interpret how routines are executed on the UIs of the applications. Being the current practice time-consuming and error-prone, in this paper we present SmartRPA, a cross-platform tool that tackles such issues by exploiting UI logs to automatically generate executable RPA scripts that automate the routines enactment by SW robots.

show abstract

“…repetitive refactoring by making on the fly code suggestions [Miltner et al 2019]. It looks at what the developer is doing, identifies repetitive edits and makes real time suggestions, rather than requiring the developer to explicitly provide the system with input-output examples.…”

Section: Refactoringmentioning

confidence: 99%

“…More specifically, the task of inductive synthesis or programming by example, where the specification is given by (partial) examples of the desired output on given input, allows for the automation of repetitive tasks in a variety of domains. Examples of domains in which robust synthesizers have been rapidly adapted in industrial tools are IntelliCode suggestions for code refactoring in Visual Studio [Gao et al 2020;Miltner et al 2019;Rolim et al 2017], extracting tabular data in PowerQuery [Le and Gulwani 2014] Current approaches in inductive synthesis are limited to writing programs that perform only syntactic transformations of the input. All information required to perform such a syntactic transformation is either available from the specification or has to be explicitly encoded in the domain specific language used by the synthesizer.…”

Section: Introductionmentioning

confidence: 99%

Semantic programming by example with pre-trained models

Verbruggen

Gulwani

2021

Proc. ACM Program. Lang.

Self Cite

View full text Add to dashboard Cite

The ability to learn programs from few examples is a powerful technology with disruptive applications in many domains, as it allows users to automate repetitive tasks in an intuitive way. Existing frameworks on inductive synthesis only perform syntactic manipulations, where they rely on the syntactic structure of the given examples and not their meaning. Any semantic manipulations, such as transforming dates, have to be manually encoded by the designer of the inductive programming framework. Recent advances in large language models have shown these models to be very adept at performing semantic transformations of its input by simply providing a few examples of the task at hand. When it comes to syntactic transformations, however, these models are limited in their expressive power. In this paper, we propose a novel framework for integrating inductive synthesis with few-shot learning language models to combine the strength of these two popular technologies. In particular, the inductive synthesis is tasked with breaking down the problem in smaller subproblems, among which those that cannot be solved syntactically are passed to the language model. We formalize three semantic operators that can be integrated with inductive synthesizers. To minimize invoking expensive semantic operators during learning, we introduce a novel deferred query execution algorithm that considers the operators to be oracles during learning. We evaluate our approach in the domain of string transformations: the combination methodology can automate tasks that cannot be handled using either technologies by themselves. Finally, we demonstrate the generality of our approach via a case study in the domain of string profiling.

show abstract

On the fly synthesis of edit suggestions

Cited by 52 publications

References 32 publications

Can Program Synthesis be Used to Learn Merge Conflict Resolutions? An Empirical Analysis

Can Program Synthesis be Used to Learn Merge Conflict Resolutions? An Empirical Analysis

Automated Generation of Executable RPA Scripts from User Interface Logs

Semantic programming by example with pre-trained models

Contact Info

Product

Resources

About