Merge branch 'master' into resizable-API

Author: aitap

.gitattributes

@@ -8,3 +8,16 @@
 #   want to preserve the line endings robustly
 inst/tests/** -text
 inst/tests/*.Rraw text eol=lf linguist-language=R
+# Ensure GitHub  linguist only considers source code files
+#   for language statistics
+docs/** linguist-documentation
+man/** linguist-documentation
+tests/** linguist-vendored
+vignettes/** linguist-vendored
+po/** linguist-vendored
+.github/** linguist-vendored
+.ci/** linguist-vendored
+.dev/** linguist-vendored
+.devcontainer/** linguist-vendored
+.graphics/** linguist-vendored
+.Rproj.user/** linguist-vendored

NEWS.md

@@ -375,6 +375,8 @@ See [#2611](https://github.com/Rdatatable/data.table/issues/2611) for details. T
 
 7. In rare situations a data.table object may lose its internal attribute that holds a self-reference. New helper function `.selfref.ok()` tests just that. It is only intended for technical use cases. See manual for examples.
 
+8. Retain important information in the error message about the source of the error when `i=` fails, e.g. pointing to `charToDate()` failing in `DT[date_col == "20250101"]`, [#7444](https://github.com/Rdatatable/data.table/issues/7444). Thanks @jan-swissre for the report and @MichaelChirico for the fix.
+
 ## data.table [v1.17.8](https://github.com/Rdatatable/data.table/milestone/41) (6 July 2025)
 
 1. Internal functions used to signal errors are now marked as non-returning, silencing a compiler warning about potentially unchecked allocation failure. Thanks to Prof. Brian D. Ripley for the report and @aitap for the fix, [#7070](https://github.com/Rdatatable/data.table/pull/7070).

R/data.table.R

@@ -111,7 +111,7 @@ replace_dot_alias = function(e) {
   )
   idx = regexpr(missing_obj_regex, err_str, perl=TRUE)
   if (idx == -1L)
-    stopf("%s", err_str, domain=NA) # Don't use stopf() directly, since err_str might have '%', #6588
+    stop(err) # Pass 'err' to retain call site data (#7444); beware also #6588
   start = attr(idx, "capture.start", exact=TRUE)[ , "obj_name"]
   used = substr(
     err_str,

inst/tests/programming.Rraw

@@ -616,3 +616,10 @@ test(202.2, f(dt, params=jpar), data.table(y=2:1))
 "." = function(...) list(.j=list("x"))
 test(202.3, dt[, .j, env=.(.j=list("y"))], data.table(x=1:2))
 rm(list=".")
+x = 2L ## no NSE, tests explicitly : ! -
+y = 2L
+test(202.4, dt[, .j, env = list(.j = x:y)], data.table(y=2:1))
+col = "x"
+test(202.5, dt[, .j, env = list(.j = !col)], error = "invalid argument type")
+test(202.6, dt[, .j, env = list(.j = -col)], error = "invalid argument to unary operator")
+rm(x, y, col)

inst/tests/tests.Rraw

@@ -20800,7 +20800,8 @@ baz(x)
 test(2295.7, is.data.table(x))
 
 # #6588: .checkTypos used to give arbitrary strings to stopf as the first argument
-test(2296, d2[x %no such operator% 1], error = '%no such operator%')
+DT = data.table(a = 1)
+test(2296, DT[x %no such operator% 1], error = '%no such operator%')
 
 # fix coercing integer/double for joins on multiple columns, #6602
 x = data.table(a=1L)
@@ -21870,3 +21871,8 @@ test(2346, data.table::fread(text = text), data.table(mary = rep("mary", 99), ha
 DT = data.table(a="a", grp=1L)
 i = c(1, 1, 1, NA, NA)
 test(2347, DT[i, .(result = all(is.na(grp) == is.na(a))), by = grp][,all(result)], options = list(datatable.optimize = 0))
+
+# .checkTypos internal implementation not exposed for non-typo bug, #7444
+DT = data.table(a = as.Date("2010-01-01"), b = 1L)
+test(2348.1, tryCatch(DT[a == as.Date("20100101")], error=conditionCall)[[1L]], quote(charToDate))
+test(2348.2, tryCatch(DT[a == as.Date("20100101") | b == 2L], error=conditionCall)[[1L]], quote(charToDate))

src/dogroups.c

@@ -511,21 +511,6 @@ SEXP dogroups(SEXP dt, SEXP dtcols, SEXP groups, SEXP grpcols, SEXP jiscols, SEX
   return(ans);
 }
 
-SEXP keepattr(SEXP to, SEXP from)
-{
-  // Same as R_copyDFattr in src/main/attrib.c, but that seems not exposed in R's api
-  // Only difference is that we reverse from and to in the prototype, for easier calling above
-  SET_ATTRIB(to, ATTRIB(from));
-  if (isS4(from)) {
-    to = PROTECT(asS4(to, TRUE, 1));
-    SET_OBJECT(to, isObject(from));
-    UNPROTECT(1);
-  } else {
-    SET_OBJECT(to, isObject(from));
-  }
-  return to;
-}
-
 SEXP growVector(SEXP x, const R_len_t newlen)
 {
   // Similar to EnlargeVector in src/main/subassign.c, with the following changes :
@@ -538,7 +523,7 @@ SEXP growVector(SEXP x, const R_len_t newlen)
   PROTECT(newx = R_allocResizableVector(TYPEOF(x), newlen));   // TO DO: R_realloc(?) here?
   if (newlen < len) len=newlen;   // i.e. shrink
   if (!len) { // cannot memcpy invalid pointer, #6819
-    keepattr(newx, x);
+    SHALLOW_DUPLICATE_ATTRIB(newx, x);
     UNPROTECT(1);
     return newx;
   }
@@ -563,7 +548,7 @@ SEXP growVector(SEXP x, const R_len_t newlen)
   }
   // if (verbose) Rprintf(_("Growing vector from %d to %d items of type '%s'\n"), len, newlen, type2char(TYPEOF(x)));
   // Would print for every column if here. Now just up in dogroups (one msg for each table grow).
-  keepattr(newx,x);
+  SHALLOW_DUPLICATE_ATTRIB(newx, x);
   UNPROTECT(1);
   return newx;
 }

tests/autoprint.R

@@ -1,4 +1,5 @@
-require(data.table)
+# TODO(#7453): re-enable warn.conflicts
+require(data.table, warn.conflicts=FALSE)
 # Tests the suppression of := output
 # Since this tests autoprinting at the console, it needs to use the .Rout.save mechanism in R CMD check
 DT = data.table(a=1:2)                # Should print at console?

tests/autoprint.Rout.save

@@ -15,8 +15,8 @@ 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.
 
-> require(data.table)
-Loading required package: data.table
+> # TODO(#7453): re-enable warn.conflicts
+> require(data.table, warn.conflicts=FALSE)
 > # Tests the suppression of := output
 > # Since this tests autoprinting at the console, it needs to use the .Rout.save mechanism in R CMD check
 > DT = data.table(a=1:2)                # Should print at console?