Automatic program transformation tools can be
\nvaluable for programmers to help them with refactoring tasks,
\nand for Computer Science students in the form of tutoring systems
\nthat suggest repairs to programming assignments. However,
\nmanually creating catalogs of transformations is complex and
\ntime-consuming. In this paper, we present REFAZER, a technique
\nfor automatically learning program transformations. REFAZER
\nbuilds on the observation that code edits performed by developers
\ncan be used as input-output examples for learning program
\ntransformations. Example edits may share the same structure
\nbut involve different variables and subexpressions, which must
\nbe generalized in a transformation at the right level of abstraction.
\nTo learn transformations, REFAZER leverages state-of-the-art
\nprogramming-by-example methodology using the following key
\ncomponents: (a) a novel domain-specific language (DSL) for describing
\nprogram transformations, (b) domain-specific deductive
\nalgorithms for efficiently synthesizing transformations in the DSL,
\nand (c) functions for ranking the synthesized transformations.
\nWe instantiate and evaluate REFAZER in two domains. First,
\ngiven examples of code edits used by students to fix incorrect
\nprogramming assignment submissions, we learn program transformations
\nthat can fix other students\u2019 submissions with similar
\nfaults. In our evaluation conducted on 4 programming tasks
\nperformed by 720 students, our technique helped to fix incorrect
\nsubmissions for 87% of the students. In the second domain, we
\nuse repetitive code edits applied by developers to the same project
\nto synthesize a program transformation that applies these edits
\nto other locations in the code. In our evaluation conducted on 56
\nscenarios of repetitive edits taken from three large C# open-source
\nprojects, REFAZER learns the intended program transformation
\nin 84% of the cases using only 2.9 examples on average.<\/p>\n","protected":false},"excerpt":{"rendered":"
Automatic program transformation tools can be valuable for programmers to help them with refactoring tasks, and for Computer Science students in the form of tutoring systems that suggest repairs to programming assignments. However, manually creating catalogs of transformations is complex and time-consuming. In this paper, we present REFAZER, a technique for automatically learning program transformations. […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"msr-content-type":[3],"msr-research-highlight":[],"research-area":[13556,13560],"msr-publication-type":[193716],"msr-product-type":[],"msr-focus-area":[],"msr-platform":[],"msr-download-source":[],"msr-locale":[268875],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-379736","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-research-area-programming-languages-software-engineering","msr-locale-en_us"],"msr_publishername":"","msr_edition":"ICSE 2017","msr_affiliation":"","msr_published_date":"2017-04-28","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"379742","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"icse2017-refactoring","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/04\/icse2017-refactoring.pdf","id":379742,"label_id":0}],"msr_related_uploader":"","msr_attachments":[],"msr-author-ordering":[{"type":"text","value":"Reudismam Rolim","user_id":0,"rest_url":false},{"type":"text","value":"Gustavo Soares","user_id":0,"rest_url":false},{"type":"text","value":"Loris D\u2019Antoni","user_id":0,"rest_url":false},{"type":"text","value":"Oleksandr Polozov","user_id":0,"rest_url":false},{"type":"user_nicename","value":"sumitg","user_id":33755,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=sumitg"},{"type":"text","value":"Rohit Gheyi","user_id":0,"rest_url":false},{"type":"text","value":"Ryo Suzuki","user_id":0,"rest_url":false},{"type":"text","value":"Bj\u00f6rn Hartmann","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[663303],"msr_project":[777298,654579],"publication":[],"video":[],"download":[],"msr_publication_type":"inproceedings","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/379736"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":1,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/379736\/revisions"}],"predecessor-version":[{"id":412451,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/379736\/revisions\/412451"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=379736"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=379736"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=379736"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=379736"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=379736"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=379736"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=379736"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=379736"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=379736"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=379736"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=379736"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=379736"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=379736"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=379736"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=379736"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}