A computable Bioconductor build report

Bioconductor spends a substantial amount of effort to build its catalog of software each day. Reporting of these results is critical for developers, users, and project leaders to understand the software “health” of the project.

The Bioconductor build reports are generally available as html pages that are navigable with bookmarks and link out to detailed reports of errors, etc. However, the build reports are not readily computable, so mining the reports, automated processing by developers, and learning about failure modes automatically is challenging. The BiocPkgTools package is a small, utilitarian toolkit for working with Bioconductor packages and the Bioconductor repository. Here, I wanted to present a new function for accessing a build report as a tidy data.frame.

Install the package, if necessary.

BiocInstaller::biocLite('seandavi/BiocPkgTools')

A one-liner returns the build report as a data.frame.

library(BiocPkgTools)
b_report = biocBuildReport()
library(DT)
datatable(b_report)