Unable to read a file when setting vroom(col_names = FALSE, col_select = c(1, 2)) (Linux/Windows)

[arisp99]

Hello!

When using Github Actions and setting vroom(col_names = FALSE), I am unable to read a file on Linux and Windows, but can on macOS. Unfortunately, I am using a mac, so I have been unable to reproduce this error locally.

I am not sure if this is the place to report this or in https://github.com/r-lib/actions, but I suspect that {vroom} is the culprit. I have created an example package (https://github.com/arisp99/testactionspkg) which defines the following function:

read <- function(.file,
                 .col_names = FALSE,
                 .col_select = everything(),
                 .name_repair = "unique") {
  vroom::vroom(
    file = .file,
    col_names = .col_names,
    col_select = .col_select,
    show_col_types = FALSE,
    .name_repair = .name_repair
  )
}

When I attempt to use this function in a vignette: testactionspkg::read(path, .col_names = FALSE), I get the following error on Github Actions for Windows and Ubuntu when running an R CMD check workflow derived from https://github.com/r-lib/actions:

* creating vignettes ... ERROR
Error: --- re-building 'read-file.Rmd' using rmarkdown
Quitting from lines 24-25 (read-file.Rmd) 
Error: Error: processing vignette 'read-file.Rmd' failed with diagnostics:
Names repair functions can't return `NA` values.
--- failed re-building 'read-file.Rmd'
SUMMARY: processing the following file failed:
  'read-file.Rmd'
Error: Error: Vignette re-building failed.
Execution halted
Error: Error in proc$get_built_file() : Build process failed
Calls: <Anonymous> ... build_package -> with_envvar -> force -> <Anonymous>
Execution halted
Error: Process completed with exit code 1.

Here is a link to the GHA: https://github.com/arisp99/testactionspkg/runs/3969422512?check_suite_focus=true run.

A reprex for those working on Linux or Windows is below (this should fail, but I cannot test it...):

data <- tibble::tribble(
              ~X1,                ~X2,                ~X3,
        "Gene ID",    "PF3D7_0106300",    "PF3D7_0523000",
           "Gene",             "atp6",             "mdr1",
  "Mutation Name",   "atp6-Ala623Glu",    "mdr1-Asn86Tyr",
     "ExonicFunc", "missense_variant", "missense_variant",
      "AA Change",        "Ala623Glu",         "Asn86Tyr",
       "Targeted",              "Yes",              "Yes",
     "D10-JJJ-23",                "0",               "13",
     "D10-JJJ-43",                "0",                "0"
  )

vroom::vroom_write(data, "~/Desktop/file.csv")

vroom::vroom(
    file = "~/Desktop/file.csv",
    col_names = FALSE,
    show_col_types = FALSE,
  )

Any insight would be appreciated! Thanks!

[maelle]

The reprex does not fail for me though (Ubuntu)

data <- tibble::tribble(
              ~X1,                ~X2,                ~X3,
        "Gene ID",    "PF3D7_0106300",    "PF3D7_0523000",
           "Gene",             "atp6",             "mdr1",
  "Mutation Name",   "atp6-Ala623Glu",    "mdr1-Asn86Tyr",
     "ExonicFunc", "missense_variant", "missense_variant",
      "AA Change",        "Ala623Glu",         "Asn86Tyr",
       "Targeted",              "Yes",              "Yes",
     "D10-JJJ-23",                "0",               "13",
     "D10-JJJ-43",                "0",                "0"
  )

local_file <- withr::local_tempfile(fileext = ".csv")
#> Setting deferred event(s) on global environment.
#>   * Execute (and clear) with `withr::deferred_run()`.
#>   * Clear (without executing) with `withr::deferred_clear()`.
vroom::vroom_write(data, local_file)

vroom::vroom(
    file = local_file,
    col_names = FALSE,
    show_col_types = FALSE,
  )
#> # A tibble: 9 × 3
#>   X1            X2               X3              
#>   <chr>         <chr>            <chr>           
#> 1 X1            X2               X3              
#> 2 Gene ID       PF3D7_0106300    PF3D7_0523000   
#> 3 Gene          atp6             mdr1            
#> 4 Mutation Name atp6-Ala623Glu   mdr1-Asn86Tyr   
#> 5 ExonicFunc    missense_variant missense_variant
#> 6 AA Change     Ala623Glu        Asn86Tyr        
#> 7 Targeted      Yes              Yes             
#> 8 D10-JJJ-23    0                13              
#> 9 D10-JJJ-43    0                0

^{Created on 2021-10-22 by the reprex package (v2.0.0)}

[arisp99]

I just checked the above reprex using GHA and also found that it passed... However, when I run the following it fails:

data <- tibble::tribble(
              ~X1,                ~X2,                ~X3,
        "Gene ID",    "PF3D7_0106300",    "PF3D7_0523000",
           "Gene",             "atp6",             "mdr1",
  "Mutation Name",   "atp6-Ala623Glu",    "mdr1-Asn86Tyr",
     "ExonicFunc", "missense_variant", "missense_variant",
      "AA Change",        "Ala623Glu",         "Asn86Tyr",
       "Targeted",              "Yes",              "Yes",
     "D10-JJJ-23",                "0",               "13",
     "D10-JJJ-43",                "0",                "0"
  )

local_file <- withr::local_tempfile(fileext = ".csv")
vroom::vroom_write(data, local_file)

vroom::vroom(
    file = local_file,
    col_names = FALSE,
    col_select = c(1, 2),
    show_col_types = FALSE
  )

I have tested running the above code in two different repositories now. The first is the original repository I used in this issue (https://github.com/arisp99/testactionspkg), which uses an R CMD check GHA workflow. I also tested the above lines of code using a workflow that just runs the R code. This code fails on Ubuntu and Windows, but passes on macOS. The workflow is below (repo: https://github.com/arisp99/test-actions):

on:
  push:
    branches: [main, master]

name: Test Vroom

jobs:
  R-CMD-check:
    runs-on: ${{ matrix.config.os }}
    name: ${{ matrix.config.os }} (${{ matrix.config.r }})
    strategy:
      fail-fast: false
      matrix:
        config:
          - {os: macOS-latest,   r: 'release'}
          - {os: windows-latest, r: 'release'}
          - {os: ubuntu-latest,   r: 'devel', http-user-agent: 'release'}
          - {os: ubuntu-latest,   r: 'release'}
          - {os: ubuntu-latest,   r: 'oldrel-1'}

    env:
      GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
      R_KEEP_PKG_SOURCE: yes

    steps:
      - uses: actions/checkout@v2

      - uses: r-lib/actions/setup-r@v1
        with:
          r-version: ${{ matrix.config.r }}
          http-user-agent: ${{ matrix.config.http-user-agent }}
          use-public-rspm: true

      - name: Run reprex
        run: |
          install.packages(c("vroom", "tibble", "withr"))
          data <- tibble::tribble(
            ~X1,                ~X2,                ~X3,
            "Gene ID",    "PF3D7_0106300",    "PF3D7_0523000",
            "Gene",             "atp6",             "mdr1",
            "Mutation Name",   "atp6-Ala623Glu",    "mdr1-Asn86Tyr",
            "ExonicFunc", "missense_variant", "missense_variant",
            "AA Change",        "Ala623Glu",         "Asn86Tyr",
            "Targeted",              "Yes",              "Yes",
            "D10-JJJ-23",                "0",               "13",
            "D10-JJJ-43",                "0",                "0"
          )

          local_file <- withr::local_tempfile(fileext = ".csv")
          vroom::vroom_write(data, local_file)

          vroom::vroom(
            file = local_file,
            col_names = FALSE,
            col_select = c(1, 2),
            show_col_types = FALSE
          )
        shell: Rscript {0}

@maelle thanks for checking the reprex! Does the modified code work on your local machine?

[maelle]

I get

data <- tibble::tribble(
              ~X1,                ~X2,                ~X3,
        "Gene ID",    "PF3D7_0106300",    "PF3D7_0523000",
           "Gene",             "atp6",             "mdr1",
  "Mutation Name",   "atp6-Ala623Glu",    "mdr1-Asn86Tyr",
     "ExonicFunc", "missense_variant", "missense_variant",
      "AA Change",        "Ala623Glu",         "Asn86Tyr",
       "Targeted",              "Yes",              "Yes",
     "D10-JJJ-23",                "0",               "13",
     "D10-JJJ-43",                "0",                "0"
  )

local_file <- withr::local_tempfile(fileext = ".csv")
#> Setting deferred event(s) on global environment.
#>   * Execute (and clear) with `withr::deferred_run()`.
#>   * Clear (without executing) with `withr::deferred_clear()`.
vroom::vroom_write(data, local_file)

vroom::vroom(
    file = local_file,
    col_names = FALSE,
    col_select = c(1, 2),
    show_col_types = FALSE
  )
#> Error: Names repair functions can't return `NA` values.

^{Created on 2021-10-22 by the reprex package (v2.0.0)}

[arisp99]

That is the error I have been seeing! Thanks @maelle!

So then to formalize the bug report. I would expect that the following lines of code to return the first two columns of the .csv file. However, when settings both col_names = FALSE and col_select = c(1, 2) there is instead an error.

The code only works on macOS and fails on Linux/Windows. When setting either col_names or col_select individually, there is no error (see my GHA workflow here) .

data <- tibble::tribble(
              ~X1,                ~X2,                ~X3,
        "Gene ID",    "PF3D7_0106300",    "PF3D7_0523000",
           "Gene",             "atp6",             "mdr1",
  "Mutation Name",   "atp6-Ala623Glu",    "mdr1-Asn86Tyr",
     "ExonicFunc", "missense_variant", "missense_variant",
      "AA Change",        "Ala623Glu",         "Asn86Tyr",
       "Targeted",              "Yes",              "Yes",
     "D10-JJJ-23",                "0",               "13",
     "D10-JJJ-43",                "0",                "0"
  )

local_file <- withr::local_tempfile(fileext = ".csv")
#> Setting deferred event(s) on global environment.
#>   * Execute (and clear) with `withr::deferred_run()`.
#>   * Clear (without executing) with `withr::deferred_clear()`.
vroom::vroom_write(data, local_file)

vroom::vroom(
    file = local_file,
    col_names = FALSE,
    col_select = c(1, 2),
    show_col_types = FALSE
  )
#> Error: Names repair functions can't return `NA` values.

[jimhester]

A more minimal reprex is

vroom::vroom(I("foo\tbar\n1\t\2\n"), col_names=FALSE, col_select = 1)
#> Error: Names repair functions can't return `NA` values.

^{Created on 2021-10-22 by the reprex package (v2.0.1)}

[jimhester]

Thank you for opening the issue and for supplying a reproducible example, it is a big help!

This should be fixed in the next release of vroom.