devtool::test Package independant "Error: invalid version specification ‘0.68’"

[stela2502]

Hi devtools developer, recently I upgraded to Ubuntu 22.04 and R 4.2.1 and out of the blue my package fails tests with an
Error: invalid version specification ‘0.68’
I know this error does not come from my package as the test script run line by line in the R interface does not produce such an error.
The traceback of the Error:

traceback()
32: stop(gettextf("invalid version specification %s", paste(sQuote(unique(x[!ok])), 
        collapse = ", ")), call. = FALSE, domain = NA)
31: .make_numeric_version(x, strict, .standard_regexps()$valid_R_system_version, 
        c("R_system_version", "package_version"))
30: R_system_version(paste(x[c("major", "minor")], collapse = "."))
29: package_version(list(major = 0, minor = VTE_VERSION))
28: ansi_has_hyperlink_support()
27: cli::style_hyperlink(paste0(basename(filename), ":", x$srcref[1], 
        ":", x$srcref[2]), paste0("file://", file.path(srcfile$wd, 
        filename)), params = c(line = x$srcref[1], col = x$srcref[2]))
26: expectation_location(x)
25: issue_header(x)
24: lapply(list(...), as.character)
23: mypaste(...)
22: crayon::bold(issue_header(x))
21: FUN(X[[i]], ...)
20: vapply(issues, issue_summary, FUN.VALUE = character(1))
19: self$report_issues(self$ctxt_issues)
18: x$end_context(...)
17: FUN(X[[i]], ...)
16: lapply(objects, f)
15: o_apply(self$reporters, "end_context", context)
14: self$end_context(self$.context)
13: reporter$end_context_if_started()
12: FUN(X[[i]], ...)
11: lapply(test_paths, test_one_file, env = env, wrap = wrap)
10: doTryCatch(return(expr), name, parentenv, handler)
9: tryCatchOne(expr, names, parentenv, handlers[[1L]])
8: tryCatchList(expr, classes, parentenv, handlers)
7: tryCatch(code, testthat_abort_reporter = function(cnd) {
       cat(conditionMessage(cnd), "\n")
       NULL
   })
6: with_reporter(reporters$multi, lapply(test_paths, test_one_file, 
       env = env, wrap = wrap))
5: test_files(test_dir = test_dir, test_package = test_package, 
       test_paths = test_paths, load_helpers = load_helpers, reporter = reporter, 
       env = env, stop_on_failure = stop_on_failure, stop_on_warning = stop_on_warning, 
       wrap = wrap, load_package = load_package)
4: test_files(test_dir = path, test_paths = test_paths, test_package = package, 
       reporter = reporter, load_helpers = load_helpers, env = env, 
       stop_on_failure = stop_on_failure, stop_on_warning = stop_on_warning, 
       wrap = wrap, load_package = load_package, parallel = parallel)
3: test_dir(test_path, package = package, reporter = reporter, ..., 
       load_package = if (package != "testthat") "source" else "none")
2: testthat::test_local(pkg$path, filter = filter, stop_on_failure = stop_on_failure, 
       ...)
1: devtools::test(filter = "Wil")

and the package that is not testing is my FastWilcoxTest package: https://github.com/sonejilab/FastWilcoxTest.
This is horribly inconvenient as I need to find a bug in that package. Please help me! Thank you!

sessionInfo()
R version 4.2.1 (2022-06-23)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 22.04 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.10.0
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.10.0

Random number generation:
 RNG:     Mersenne-Twister 
 Normal:  Inversion 
 Sample:  Rounding 
 
locale:
 [1] LC_CTYPE=en_AU.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=sv_SE.UTF-8        LC_COLLATE=en_AU.UTF-8    
 [5] LC_MONETARY=sv_SE.UTF-8    LC_MESSAGES=en_AU.UTF-8   
 [7] LC_PAPER=sv_SE.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=sv_SE.UTF-8 LC_IDENTIFICATION=C       

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

other attached packages:
[1] FastWilcoxTest_0.1.16 testthat_3.1.4       

loaded via a namespace (and not attached):
 [1] Biobase_2.56.0      pkgload_1.2.4       splines_4.2.1      
 [4] qqconf_1.2.3        brio_1.1.3          sn_2.0.2           
 [7] Rdpack_2.3.1        stats4_4.2.1        TFisher_0.2.0      
[10] remotes_2.4.2       sessioninfo_1.2.2   numDeriv_2016.8-1.1
[13] pillar_1.7.0        lattice_0.20-45     glue_1.6.2         
[16] rbibutils_2.2.8     sandwich_3.0-1      Matrix_1.4-0       
[19] plyr_1.8.7          pkgconfig_2.0.3     devtools_2.4.3     
[22] purrr_0.3.4         mvtnorm_1.1-3       processx_3.5.3     
[25] tibble_3.1.7        usethis_2.1.6       ellipsis_0.3.2     
[28] TH.data_1.1-1       cachem_1.0.6        withr_2.5.0        
[31] BiocGenerics_0.42.0 cli_3.3.0           mnormt_2.1.0       
[34] survival_3.2-13     magrittr_2.0.3      crayon_1.5.1       
[37] memoise_2.0.1       ps_1.7.0            fs_1.5.2           
[40] fansi_1.0.3         MASS_7.3-55         pkgbuild_1.3.1     
[43] tools_4.2.1         prettyunits_1.1.1   lifecycle_1.0.1    
[46] multcomp_1.4-19     mutoss_0.1-12       stringr_1.4.0      
[49] plotrix_3.8-2       callr_3.7.0         compiler_4.2.1     
[52] rlang_1.0.2         grid_4.2.1          rstudioapi_0.13    
[55] waldo_0.4.0         codetools_0.2-18    multtest_2.52.0    
[58] reshape2_1.4.4      R6_2.5.1            zoo_1.8-10         
[61] fastmap_1.1.0       utf8_1.2.2          mathjaxr_1.6-0     
[64] rprojroot_2.0.3     metap_1.8           desc_1.4.1         
[67] stringi_1.7.6       Rcpp_1.0.8.3        vctrs_0.4.1  

[jennybc]

This is unlikely to be a problem with devtools. I suspect it's a problem with your machine, e.g. perhaps some corruption in your package library.

[stela2502]

Thank you for this info. I will try with a new library and get back to you.

[stela2502]

It is quite strange, as I actually had installed the system the day before. I had only started to use R the day this issue occurred.
But I'll nuke my lib and restart ;-)

[stela2502]

THANK YOU!
This error reproducibly occured when dirt-upgrading from Ubuntu 20.04 to 22.04. And yes removing my whole R library and re-installing everything did do the trick!
A whole work day down the drain ;-)
But thanks to you not more days!

[stela2502]

Crap - the error is back. Do you have any idea on how to debug this?
I have removed my personal library and re-installed devtools. Seams I do not have the right compiler? Or am I wrong?

[stela2502]

I really think that is some error in devtools:

> source('tests/testthat/test-entropy.R')
Fehler: Reason: useless and depricated
> devtools::test(filter='entropy')
ℹ Loading FastWilcoxTest
ℹ Testing FastWilcoxTest
✔ | F W S  OK | Context
✔ |     1   0 | entropy                                                                                      
─────────────────────────────────────────────────────────────────────────────────────────────────────────────
Fehler: ungültige Versionsspezifikation ‘0.68’

And the error (Fehler) seams to be some C stuff I do not understand. All - really all - functions work is not started with devtools::test but source. I have even checked the two last functions of the backtrace:

gettextf("some test: %s",sQuote( 2131))
[1] "some test: ‘2131’"

I do not understand this error - can you please help me?

[stela2502]

Oh and the two relevant lines in the test script are:

context( 'entropy')
skip('useless and depricated')

[stela2502]

ARGH - now I have hit the bottom of this error:

> R_system_version("4.2")
Fehler: ungültige Versionsspezifikation ‘4.2’
> R_system_version("4.2.3")
[1] ‘4.2.3’

So it is some issue with devtools after all.
Could you please check this?
It seams what one should do here is a

>numeric_version("0.68")
[1] ‘0.68’
> R_system_version("0.68")
Fehler: ungültige Versionsspezifikation ‘0.68’

Thank you!

[stela2502]

Hi again. It was the cli package. Installing that one from github instead of cran fixed my problem.

[krassowski]

I also experience this on Ubuntu 22.04, triggered by devtools::test() after tests run through.
After installing cli from GitHub and restarting R session the problem went away (but there might be another explanation).

[stela2502]

I wonder how you get to you statement "but there might be another explanation". The error was thrown by the cli package and installing an update of the package solved the problem. Can you please come up with another explanation if you assume there is one? Or is that your general strategy?

[bbolker]

I think this is fixed in r-lib/cli#452 . This was unfortunately merged one day after (26 April 2022) the last cli version on CRAN. The underlying problem is with the environment variable VTE_VERSION, which is incorrectly handled by cli.

[stela2502]

I was going to link this here, too. bbolker - Thank you!