A command tool wrapping the R package "lintr" for static R code analysis
The "linter" package from Jim Hester is an excellent R package that lints R code and has integrations with RStudio, Emacs, Vim, Sublime, and Atom. This project is a very simple wrapper around this that creates a tool that can be used from the command line.
In the future, this tool is planned to be in the Debian/Ubuntu repos. On the (as of yet unreleased) Debian Bullseye and Ubuntu 20.04, the package will be easily installed with
$ sudo apt install lintr
For now, the tool can be manually installed.
The command line tool requires r and the "lintr" CRAN package. Install R according to the directions provided on the R-project. The "lintr" R package can then be installed from CRAN using
$ Rscript -e 'install.packages("lintr")'
Now install the lintr script
$ sudo curl https://raw.github.com/dashaub/lintr/master/lintr -Lo /usr/local/bin/lintr && sudo chmod a+x /usr/local/bin/lintr
We can lint the provided example.R file.
$ lintr example.R
example.R:2:3: style: Use <-, not =, for assignment.
a = 3
^
example.R:3:7: style: Put spaces around all infix operators.
a <- a*2
~^~
example.R:6:12: style: Only use double-quotes.
message <- 'hello world'
^~~~~~~~~~~~~
example.R:7:1: style: Variable and function names should be all lowercase.
myFunc <- function()
^~~~~~
example.R:8:1: style: Opening curly braces should never go on their own line and should always be followed by a new line.
{
^
example.R:9:12: warning: no visible binding for global variable ‘missing_var’
return(missing_var)
^~~~~~~~~~~
example.R:12:1: style: Variable and function names should be all lowercase.
myFunc()
^~~~~~
example.R:13:1: style: Trailing blank lines are superfluous.
^