Autoname j

[MichaelChirico]

Closes #2478
Closes #609

The meat is in the do_j_names function internal to [.data.table so it inherits the right parent & can edit jvnames with <<-. Also realized this would clean up the fix done in #3604 so did that. The rest of the diff (in 0b2e058) comes from using our vapply_1{i,b,c} wrappers more consistently

I imagine this one will have some revdep breaking (e.g. we broke two of our own tests)... maybe save for the next release then?

[MichaelChirico]

Also FWIW I think we can make a small change to allow chmatch to accept is.name objects so we can use jsub[[1L]] %chin% c('list', '.'), etc, I copied this from relop.c in r-source, but it was breaking things, any ideas?

    // adapted from R source src/main/relop.c for how symbol == character is handled
    if (isSymbol(x)) {
      PROTECT_INDEX xp;
      PROTECT_WITH_INDEX(x, &xp);

      SEXP tmp = allocVector(STRSXP, 1);
      PROTECT(tmp);
      SET_STRING_ELT(tmp, 0, PRINTNAME(x));
      REPROTECT(x = tmp, xp);
      UNPROTECT(1);
    }

[jangorecki]

I would say no to breaking changes 1.12.4

[codecov]

Codecov Report

Merging #3802 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff            @@
##           master   #3802      +/-   ##
=========================================
+ Coverage    99.4%   99.4%   +<.01%     
=========================================
  Files          72      72              
  Lines       13655   13673      +18     
=========================================
+ Hits        13574   13592      +18     
  Misses         81      81

Impacted Files	Coverage Δ
R/print.data.table.R	100% <100%> (ø)	⬆️
R/fread.R	100% <100%> (ø)	⬆️
R/groupingsets.R	100% <100%> (ø)	⬆️
R/data.table.R	100% <100%> (ø)	⬆️
R/setops.R	100% <100%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 412a781...86c444a. Read the comment docs.