Some strings consisting of variable names probably shouldn't be translated

[aitap]

There's quite a lot of debugging printout that consists of argument/variable names that (arguably) should not be translated:

• data.table/src/fread.c

  Line 1343 in 3b376da

  DTPRINT(_(" NAstrings = ["));

• data.table/src/fread.c

  Line 1346 in 3b376da

  DTPRINT(_("]\n"));

• data.table/src/fread.c

  Line 2083 in 3b376da

  DTPRINT(_(" =====\n"));

  
  

  data.table/src/fread.c

  Line 2090 in 3b376da

  DTPRINT(_(" =====\n"));

• data.table/src/fread.c

  Line 2706 in 3b376da

  DTPRINT(_("%10d : %-9s '%c'\n"), typeCounts[i], typeName[i], typeLetter[i]);

• data.table/src/fread.c

  Line 2742 in 3b376da

  DTPRINT(_("=============================\n"));

• data.table/src/assign.c

  Line 440 in d7e95d1

  if (verbose) Rprintf(_("RHS_list_of_columns == %s\n"), RHS_list_of_columns ? "true" : "false");

• data.table/src/bmerge.c

  Line 69 in d7e95d1

  if (iN && it!=xt) error(_("typeof x.%s (%s) != typeof i.%s (%s)"), CHAR(STRING_ELT(getAttrib(xdt,R_NamesSymbol),xcols[col]-1)), type2char(xt), CHAR(STRING_ELT(getAttrib(idt,R_NamesSymbol),icols[col]-1)), type2char(it));

• data.table/src/dogroups.c

  Line 109 in d7e95d1

  if (isNull(jiscols) && (length(bynames)!=length(groups) || length(bynames)!=length(grpcols))) error(_("!length(bynames)[%d]==length(groups)[%d]==length(grpcols)[%d]"),length(bynames),length(groups),length(grpcols));

• data.table/src/dogroups.c

  Line 138 in d7e95d1

  if (length(names) != length(SDall)) error(_("length(names)!=length(SD)"));

• data.table/src/dogroups.c

  Line 154 in d7e95d1

  if (length(xknames) != length(xSD)) error(_("length(xknames)!=length(xSD)"));

• data.table/src/dogroups.c

  Lines 162 to 163 in d7e95d1

  	if (length(iSD)!=length(jiscols)) error(_("length(iSD)[%d] != length(jiscols)[%d]"),length(iSD),length(jiscols));
  	if (length(xSD)!=length(xjiscols)) error(_("length(xSD)[%d] != length(xjiscols)[%d]"),length(xSD),length(xjiscols));

• data.table/src/dogroups.c

  Line 403 in d7e95d1

  if (length(ans) != ngrpcols + njval) error(_("dogroups: length(ans)[%d]!=ngrpcols[%d]+njval[%d]"),length(ans),ngrpcols,njval);

• data.table/src/forder.c

  Line 784 in d7e95d1

  Rprintf(_("nradix=%d\n"), nradix);

• data.table/src/fread.c

  Line 2261 in d7e95d1

  if (verbose) DTPRINT(_(" jumps=[%d..%d), chunk_size=%"PRIu64", total_size=%"PRIu64"\n"),

• data.table/src/freadR.c

  Line 158 in d7e95d1

  if (INTEGER(nThreadArg)[0]<1) error(_("nThread(%d)<1"), INTEGER(nThreadArg)[0]);

• data.table/src/freadR.c

  Line 719 in 3b376da

  error(_("%s"), msg); // include "%s" because data in msg might include '%'

  
  

  data.table/src/utils.c

  Line 391 in 3b376da

  warning(_("%s"), ret);

• data.table/src/fsort.c

  Line 122 in d7e95d1

  if (verbose) Rprintf(_("nth=%d, nBatch=%d\n"),nth,nBatch);

• data.table/src/fsort.c

  Line 178 in d7e95d1

  if (verbose) Rprintf(_("maxBit=%d; MSBNbits=%d; shift=%d; MSBsize=%zu\n"), maxBit, MSBNbits, shift, MSBsize);

• data.table/src/fsort.c

  Line 254 in 3b376da

  Rprintf(_("Top 20 MSB counts: ")); for(int i=0; i<MIN(MSBsize,20); i++) Rprintf(_("%"PRId64" "), (int64_t)msbCounts[order[i]]); Rprintf(_("\n"));

• data.table/src/fsort.c

  Line 339 in 3b376da

  Rprintf(_("%d: %.3f (%4.1f%%)\n"), i, t[i]-t[i-1], 100.*(t[i]-t[i-1])/tot);

• data.table/src/fwrite.c

  Line 632 in 3b376da

  for (int j=0; j<args.ncol; j++) DTPRINT(_("%d "), args.whichFun[j]);

  
  

  data.table/src/fwrite.c

  Line 634 in 3b376da

  for (int j=0; j<30; j++) DTPRINT(_("%d "), args.whichFun[j]);

  
  

  data.table/src/fwrite.c

  Line 636 in 3b376da

  for (int j=args.ncol-10; j<args.ncol; j++) DTPRINT(_("%d "), args.whichFun[j]);

  
  

  data.table/src/fwrite.c

  Line 635 in 3b376da

  DTPRINT(_("... "));

• data.table/src/fwrite.c

  Lines 638 to 639 in d7e95d1

  	DTPRINT(_("\nargs.doRowNames=%d args.rowNames=%p args.rowNameFun=%d doQuote=%d args.nrow=%"PRId64" args.ncol=%d eolLen=%d\n"),
  	args.doRowNames, args.rowNames, args.rowNameFun, doQuote, args.nrow, args.ncol, eolLen);

• data.table/src/fwrite.c

  Line 786 in 3b376da

  else STOP(_("%s: '%s'"), strerror(errwrite), args.filename);

• data.table/src/gsumm.c

  Line 63 in d7e95d1

  if (LENGTH(f) != ngrp) error(_("length(f)=%d != length(l)=%d"), LENGTH(f), ngrp);

• data.table/src/openmp-utils.c

  Lines 89 to 97 in d7e95d1

  	Rprintf(_(" omp_get_num_procs() %d\n"), omp_get_num_procs());
  	Rprintf(_(" R_DATATABLE_NUM_PROCS_PERCENT %s\n"), mygetenv("R_DATATABLE_NUM_PROCS_PERCENT", "unset (default 50)"));
  	Rprintf(_(" R_DATATABLE_NUM_THREADS %s\n"), mygetenv("R_DATATABLE_NUM_THREADS", "unset"));
  	Rprintf(_(" R_DATATABLE_THROTTLE %s\n"), mygetenv("R_DATATABLE_THROTTLE", "unset (default 1024)"));
  	Rprintf(_(" omp_get_thread_limit() %d\n"), omp_get_thread_limit());
  	Rprintf(_(" omp_get_max_threads() %d\n"), omp_get_max_threads());
  	Rprintf(_(" OMP_THREAD_LIMIT %s\n"), mygetenv("OMP_THREAD_LIMIT", "unset")); // CRAN sets to 2
  	Rprintf(_(" OMP_NUM_THREADS %s\n"), mygetenv("OMP_NUM_THREADS", "unset"));
  	Rprintf(_(" RestoreAfterFork %s\n"), RestoreAfterFork ? "true" : "false");

• data.table/src/reorder.c

  Line 39 in d7e95d1

  if (length(order) != nrow) error(_("nrow(x)[%d]!=length(order)[%d]"),nrow,length(order));

• data.table/src/forder.c

  Line 873 in d7e95d1

  if (stat[i]) Rprintf(_("stat[%03d]==%20"PRIu64"\n"), i, (uint64_t)stat[i]);

Which of these should have their _ removed and marked with // # notranslate for potools? Which should be translated anyway?

Topic originally started by @rikivillalba in #6503 (comment)

Edit: I have since remembered about punctuation rules, e.g. a space being required before the colon in French, and checked the corresponding checkboxes. If you can think of any reason to translate any of my examples, please check them too.

[rikivillalba]

Perhaps it should be weigthed the trade off between the number of required // # nontranslate current and new annotations versus not running the string extraction and adding the new _()'s manually?

[MichaelChirico]

Perhaps it should be weigthed the trade off between the number of required // # nontranslate current and new annotations versus not running the string extraction and adding the new _()'s manually?

There are over 400 # nocov annotations, I think the number of # notranslate tags implied here is pretty small.

[MichaelChirico]

WDYT about leaving this one: DTPRINT(_("... "));, I figure ... could be translated differently across languages, e.g. quickly googling for Japanese: https://japanese.stackexchange.com/questions/18058/ellipsis-usage-and-format