Fix non-API calls for closures

[shikokuchuo]

Fixes #144.

[spgarbet]

Thank you for your contribution. Having a PR makes moving forward a lot faster.

Couple housekeeping things, do you want to add your name to the DESCRIPTION contributor? If you do that can you bump the version number for this change, it's in VERSION and DESCRIPTION.

Will a new version proposed, the package has strong dependencies and I need to reach out to a couple other groups to get them to run their tests suites and get sign off.

[spgarbet]

Tests pass. garbetsp:~/Projects/cran/yaml$ make test
R CMD INSTALL -l build/lib build/yaml

• installing source package ‘yaml’ ...
  ** using staged installation
  ** libs
  using C compiler: ‘gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0’
  make[1]: Entering directory '/home/garbetsp/Projects/cran/yaml/build/yaml/src'
  make[1]: Nothing to be done for 'all'.
  make[1]: Leaving directory '/home/garbetsp/Projects/cran/yaml/build/yaml/src'
  installing to /home/garbetsp/Projects/cran/yaml/build/lib/00LOCK-yaml/00new/yaml/libs
  ** R
  ** inst
  ** byte-compile and prepare package for lazy loading
  ** help
  *** installing help indices
  ** building package indices
  ** testing if installed package can be loaded from temporary location
  ** checking absolute paths in shared objects and dynamic libraries
  ** testing if installed package can be loaded from final location
  ** testing if installed package keeps a record of temporary installation path
• DONE (yaml)
  R --vanilla -e "library(RUnit); library(yaml, lib.loc = 'build/lib'); source('build/yaml/tests/RUnit.R')"

R version 4.4.0 (2024-04-24) -- "Puppy Cup"
Copyright (C) 2024 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

Natural language support but running in an English locale

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

library(RUnit); library(yaml, lib.loc = 'build/lib'); source('build/yaml/tests/RUnit.R')
RUNIT TEST PROTOCOL -- Wed Jul 3 08:09:21 2024

---

Number of test functions: 171
Number of errors: 0
Number of failures: 0

1 Test Suite :
yaml tests - 171 test functions, 0 errors, 0 failures

[spgarbet]

valgrind-test

RUNIT TEST PROTOCOL -- Wed Jul  3 08:13:45 2024 
*********************************************** 
Number of test functions: 171 
Number of errors: 0 
Number of failures: 0 

 
1 Test Suite : 
yaml tests - 171 test functions, 0 errors, 0 failures
> 
> 
==1995226== 
==1995226== HEAP SUMMARY:
==1995226==     in use at exit: 57,659,298 bytes in 10,668 blocks
==1995226==   total heap usage: 35,001 allocs, 24,333 frees, 101,535,306 bytes allocated
==1995226== 
==1995226== LEAK SUMMARY:
==1995226==    definitely lost: 0 bytes in 0 blocks
==1995226==    indirectly lost: 0 bytes in 0 blocks
==1995226==      possibly lost: 0 bytes in 0 blocks
==1995226==    still reachable: 57,659,298 bytes in 10,668 blocks
==1995226==                       of which reachable via heuristic:
==1995226==                         newarray           : 4,264 bytes in 1 blocks
==1995226==         suppressed: 0 bytes in 0 blocks
==1995226== Reachable blocks (those to which a pointer was found) are not shown.
==1995226== To see them, rerun with: --leak-check=full --show-leak-kinds=all
==1995226== 
==1995226== For lists of detected and suppressed errors, rerun with: -s
==1995226== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

[shikokuchuo]

Couple housekeeping things, do you want to add your name to the DESCRIPTION contributor? If you do that can you bump the version number for this change, it's in VERSION and DESCRIPTION.

Sure, I've done that. Separately, I see the GitHub release hasn't caught up to the current CRAN version - that confused me at first.

Will a new version proposed, the package has strong dependencies and I need to reach out to a couple other groups to get them to run their tests suites and get sign off.

Sure, although this fix shouldn't result in any change in behaviour. Btw. great to see you run Valgrind.

[spgarbet]

That's odd about CRAN. There were some required fixes in that version and it's been out a long while.

Edit: I remember now, the version bump mismatch was for documentation changes.

[spgarbet]

data.table test with package installed

R version 4.4.0 (2024-04-24) -- "Puppy Cup"
Copyright (C) 2024 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

  Natural language support but running in an English locale

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> source("~/Projects/cran/data.table/tests/main.R")
getDTthreads(verbose=TRUE):
  OpenMP version (_OPENMP)       201511
  omp_get_num_procs()            12
  R_DATATABLE_NUM_PROCS_PERCENT  unset (default 50)
  R_DATATABLE_NUM_THREADS        unset
  R_DATATABLE_THROTTLE           unset (default 1024)
  omp_get_thread_limit()         2147483647
  omp_get_max_threads()          12
  OMP_THREAD_LIMIT               unset
  OMP_NUM_THREADS                unset
  RestoreAfterFork               true
  data.table is using 6 threads with throttle==1024. See ?setDTthreads.
test.data.table() running: /home/garbetsp/R/x86_64-pc-linux-gnu-library/4.4/data.table/tests/tests.Rraw.bz2
Running test id 2243.56            
Wed Jul  3 09:36:58 2024  endian==little, sizeof(long double)==16, longdouble.digits==64, sizeof(pointer)==8, TZ==unset, Sys.timezone()=='America/Menominee', Sys.getlocale()=='LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=en_US.UTF-8;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C', l10n_info()=='MBCS=TRUE; UTF-8=TRUE; Latin-1=FALSE; codeset=UTF-8', getDTthreads()=='OpenMP version (_OPENMP)==201511; omp_get_num_procs()==12; R_DATATABLE_NUM_PROCS_PERCENT==unset (default 50); R_DATATABLE_NUM_THREADS==unset; R_DATATABLE_THROTTLE==unset (default 1024); omp_get_thread_limit()==2147483647; omp_get_max_threads()==12; OMP_THREAD_LIMIT==unset; OMP_NUM_THREADS==unset; RestoreAfterFork==true; data.table is using 6 threads with throttle==1024. See ?setDTthreads.', zlibVersion()==1.2.11 ZLIB_VERSION==1.2.11
10 longest running tests took 13s (48% of 27s)
      ID  time nTest
 1: 2155 3.817     5
 2: 1438 1.829   738
 3: 1437 1.471    36
 4: 1652 1.068    91
 5: 1648 1.063    91
 6: 1650 1.015    91
 7: 1223 0.932   728
 8: 2233 0.890    55
 9: 1644 0.733    91
10: 1642 0.680    91
All 11070 tests (last 2243.56) in tests/tests.Rraw.bz2 completed ok in 31.3s elapsed (36.4s cpu)

[spgarbet]

@hadley @MichaelChirico Seeking your approval for release of this patch to CRAN.

[hadley]

@spgarbet looks fine to me, but I'd feel more confident if you added some GitHub actions to check all relevant versions of R (e.g. with usethis::use_github_action("check-full"))

[spgarbet]

Our institutions account doesn't cover doing that on github. We do have full gitlab automations. I'm looking into seeing if they have something similar that can be done using gitlab. (Also "looks fine to me" have been words I've said just before serious breakage).

[hadley]

@spgarbet I don't think that should matter for publicly available repos? IIRC it's only private repos that you need to pay for.

[spgarbet]

I set up the github action on the main branch and it's passing.

[spgarbet]

Code change looks good. Hadley has approved if it passes github checks. data.table has not responded, but I ran their test suite and got no errors.

[MichaelChirico]

LGTM, thanks for the ping, and thanks @shikokuchuo for the PR :)