Synthesizing Data Wranglers
Abstract
Maintaining software systems that manage evolving representations of data is tedious and error prone. This study developed algorithms for automatically synthesizing lenses, which are software adaptors that convert between two different representations of the same information. These synthesis algorithms take as input (i) the type of each source as a regular expression, and (ii) a small collection of representative examples of the desired translation. They produce as output different varieties of string lenses. The products of this investigation include several published papers describing and evaluating the algorithms, as well as open-source code that implements those algorithms.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 2019
- Accession Number
- AD1071633
Entities
People
- Anders Miltner
- Benjamin Pierce
- David Walker
- Kathleen Fisher
- Solomon Maina
- Steve Zdancewic
Organizations
- Princeton University