tryCatch for nchar() on old R

[MichaelChirico]

See e.g.

• failing on 4.1: https://github.com/Rdatatable/data.table/actions/runs/30220885395/job/89843028144?pr=7832
• passing on 4.2: https://github.com/Rdatatable/data.table/actions/runs/30220885395/job/89843028187?pr=7832 (NB: the test still fails, but tests/main.R passes)

h/t @joshhwuu for identifying the culprit R bugfix in the comment below.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.01%. Comparing base (f1edc68) to head (a0baec7).

Additional details and impacted files

@@                 Coverage Diff                 @@
##           other-attach-order    #7848   +/-   ##
===================================================
  Coverage               99.01%   99.01%           
===================================================
  Files                      88       88           
  Lines                   17286    17286           
===================================================
  Hits                    17116    17116           
  Misses                    170      170           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

• HEAD=old-nchar-try much slower for transform improved in #5493
  

Generated via commit a0baec7

Download link for the artifact containing the test results: ↓ atime-results.zip

Task	Duration
R setup and installing dependencies	2 minutes and 48 seconds
Installing different package versions	1 minutes and 9 seconds
Running and plotting the test cases	5 minutes and 35 seconds

[joshhwuu]

changes look good to me..

this may be the fix on R's side: r-devel/r-svn@a786d1e

[MichaelChirico]

Great spot! I find that on R 4.1.3:

https://github.com/wch/r-source/blob/7f6327b076889621a2e7244ef6c769f74c6012fb/src/main/character.c#L279

But not R 4.2.0:

https://github.com/wch/r-source/blob/775760668cadaf563fe771960c8a4cb1f5d8e75c/src/main/character.c#L279

[aitap]

Since nchar(allowNA = TRUE) mistakenly signals an error instead of returning NA, we could also use

diff --git a/R/print.data.table.R b/R/print.data.table.R
index e602f80d6..817e49420 100644
--- a/R/print.data.table.R
+++ b/R/print.data.table.R
@@ -256,7 +256,7 @@ char.trunc = function(x, trunc.char = getOption("datatable.prettyprint.char")) {
   if (is.null(trunc.char)) return(x)
   trunc.char = max(0L, suppressWarnings(as.integer(trunc.char[1L])), na.rm=TRUE)
   if (!is.character(x) || trunc.char <= 0L) return(x)
-  nchar_width = nchar(x, 'width', allowNA = TRUE)
+  nchar_width = tryCatch(nchar(x, 'width', allowNA = TRUE), error = function(...) NA)
   nchar_chars = nchar(x, 'char', allowNA = TRUE)
   is_full_width = nchar_width > nchar_chars
   is_full_width[is.na(is_full_width)] = FALSE

and let the rest of the code handle the NA like it currently does on R ≥ 4.2.

[MichaelChirico]

Great suggestion! Added the NEWS now