LSP diagnostics shows `lintr` error although `lintr::lint_package()` does not

[jsr-p]

Hi

I am developing a package where I use .data from rlang together with dplyr. The code snippet is

#' @export
#' @importFrom rlang .data
my_fun <- function(df) {
  df |> dplyr::mutate(
    x = .data[[col]]
  )
}

with NAMESPACE file:

# Generated by roxygen2: do not edit by hand

export(my_fun)
importFrom(rlang,.data)

When running lintr::lint_package() inside RStudio I do not get any errors. However, using the rlanguageserver inside Neovim gives the following error:

My sessionInfo() is:

> sessionInfo()
R version 4.3.2 (2023-10-31)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Arch Linux

Matrix products: default
BLAS:   /usr/lib/libblas.so.3.12.0 
LAPACK: /usr/lib/liblapack.so.3.12.0

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

time zone: Europe/Copenhagen
tzcode source: system (glibc)

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] lintr_3.1.1           languageserver_0.3.16 testpkg_0.0.0.9000   
[4] devtools_2.4.5        usethis_2.2.2        

loaded via a namespace (and not attached):
 [1] generics_0.1.3     utf8_1.2.4         xml2_1.3.5         stringi_1.8.3     
 [5] digest_0.6.33      magrittr_2.0.3     pkgload_1.3.3      fastmap_1.1.1     
 [9] rprojroot_2.0.4    processx_3.8.3     pkgbuild_1.4.3     sessioninfo_1.2.2 
[13] brio_1.1.4         backports_1.4.1    ps_1.7.5           urlchecker_1.0.1  
[17] promises_1.2.1     purrr_1.0.2        fansi_1.0.6        codetools_0.2-19  
[21] lazyeval_0.2.2     cli_3.6.2          shiny_1.8.0        crayon_1.5.2      
[25] rlang_1.1.2        ellipsis_0.3.2     remotes_2.4.2.1    withr_2.5.2       
[29] cachem_1.0.8       parallel_4.3.2     tools_4.3.2        memoise_2.0.1     
[33] dplyr_1.1.3        httpuv_1.6.13      vctrs_0.6.5        cyclocomp_1.1.1   
[37] R6_2.5.1           mime_0.12          lifecycle_1.0.4    stringr_1.5.1     
[41] fs_1.6.3           htmlwidgets_1.6.3  miniUI_0.1.1.1     callr_3.7.3       
[45] pkgconfig_2.0.3    desc_1.4.3         rex_1.2.1          pillar_1.9.0      
[49] later_1.3.2        data.table_1.14.8  glue_1.6.2         profvis_0.3.8     
[53] Rcpp_1.0.11        tidyselect_1.2.0   xfun_0.41          tibble_3.2.1      
[57] rstudioapi_0.15.0  knitr_1.45         xtable_1.8-4       htmltools_0.5.7   
[61] testthat_3.2.1     compiler_4.3.2     roxygen2_7.2.3     xmlparsedata_1.0.5

and I am using the newest r-languageserver:

The only way that I can make it go away is by writing in my .lintr file:

linters: linters_with_defaults(
    object_usage_linter = NULL
  )

but how can I do it correctly :D?

Thank you!

[pit00]

I have a similar problem at VS Code. (also did the same steps)

[pit00]

I put a .Rprofile file at project folder, with devtools::load_all(). This workaround solves my linter problems, can you do the same?

[jsr-p]

@pit00 does not work for me; still getting the diagnostic message :/

[luciorq]

This is not really a solution for your problem, but just namespace the .data object.

With:

my_fun <- function(df, col) {

  .data <- rlang::.data

  df |> dplyr::mutate(
    x = .data[[col]]
  )
}

This could potentially also bring some safety regarding namespace collisions.