Merge branch 'master' into memcpy-ro

Author: MichaelChirico

.ci/atime/tests.R

@@ -117,6 +117,11 @@ test.list <- atime::atime_test_list(
       file.path("src", "init.c"),
       paste0("R_init_", Package_regex),
       paste0("R_init_", gsub("[.]", "_", new.Package_)))
+    # allow compilation on new R versions where 'Calloc' is not defined
+    pkg_find_replace(
+      file.path("src", "*.c"),
+      "\\b(Calloc|Free|Realloc)\\b",
+      "R_\\1")
     pkg_find_replace(
       "NAMESPACE",
       sprintf('useDynLib\\("?%s"?', Package_regex),

.ci/linters/c/alloc_linter.R

@@ -6,7 +6,7 @@ alloc_linter = function(c_obj) {
   lines = c_obj$lines
   # Be a bit more precise to avoid mentions in comments, and allow
   #   malloc(0) to be used for convenience (e.g. #6757)
-  alloc_lines = grep(R"{=\s*([(]\w+\s*[*][)])?[mc]alloc[(][^0]}", lines)
+  alloc_lines = grep(R"{=\s*([(]\s*\w+\s*[*]\s*[)])?\s*[mc]alloc[(][^0]}", lines)
   if (!length(alloc_lines)) return()
   # int *tmp=(int*)malloc(...); or just int tmp=malloc(...);
   alloc_keys = lines[alloc_lines] |>
@@ -31,7 +31,7 @@ alloc_linter = function(c_obj) {
       cat("FILE: ", c_obj$path, "; LINES: ", head(bad_lines_idx, 1L), "-", tail(bad_lines_idx, 1L), "\n", sep="")
       writeLines(lines[bad_lines_idx])
       cat(strrep("-", max(nchar(lines[bad_lines_idx]))), "\n", sep="")
-      stop("Expected the malloc()/calloc() usage above to be followed immediately by error checking.", call.=FALSE)
+      stop("Expected the malloc()/calloc() usage above to be followed immediately by error checking (using '!', not '==NULL').", call.=FALSE)
     }
   })
 }

.ci/linters/cocci/malloc_return_value_cast.cocci

@@ -0,0 +1,6 @@
+@malloc_return_value_cast expression@
+type T;
+expression E;
+@@
+- (T)
+  malloc(E)

.github/workflows/R-CMD-check.yaml

@@ -27,8 +27,9 @@ jobs:
           # GHA does run these jobs concurrently but even so reducing the load seems like a good idea.
           - {os: windows-latest, r: 'devel'}
           # - {os: macOS-latest, r: 'release'}    # test-coverage.yaml uses macOS
-          - {os: ubuntu-24.04, r: 'release', rspm: "https://packagemanager.rstudio.com/cran/__linux__/focal/latest"}
-          # - {os: ubuntu-24.04,   r: 'devel', rspm: "https://packagemanager.rstudio.com/cran/__linux__/focal/latest", http-user-agent: "R/4.1.0 (ubuntu-24.04) R (4.1.0 x86_64-pc-linux-gnu x86_64 linux-gnu) on GitHub Actions" }
+          # TODO(remotes>2.5.0): Use 24.04[noble?]
+          - {os: ubuntu-22.04, r: 'release', rspm: "https://packagemanager.rstudio.com/cran/__linux__/jammy/latest"}
+          # - {os: ubuntu-22.04,   r: 'devel', rspm: "https://packagemanager.rstudio.com/cran/__linux__/jammy/latest", http-user-agent: "R/4.1.0 (ubuntu-22.04) R (4.1.0 x86_64-pc-linux-gnu x86_64 linux-gnu) on GitHub Actions" }
           #   GLCI covers R-devel; no need to delay contributors in dev due to changes in R-devel in recent days
 
     env:
@@ -64,7 +65,7 @@ jobs:
           while read -r cmd
           do
             eval sudo $cmd
-          done < <(Rscript -e 'writeLines(remotes::system_requirements("ubuntu", "24.04"))')
+          done < <(Rscript -e 'writeLines(remotes::system_requirements("ubuntu", "22.04"))')
 
       - name: Install dependencies
         run: |

CODEOWNERS

@@ -23,6 +23,10 @@
 /src/programming.c @jangorecki
 /vignettes/datatable-programming.Rmd @jangorecki
 
+# roll-up & setops
+/R/groupingsets.R @jangorecki
+/R/setops.R @jangorecki
+
 # GForce groupby
 /src/gsumm.c @ben-schwen
 # datetime classes

NEWS.md

@@ -8,6 +8,10 @@
 
 2. `melt()` now supports using `patterns()` with `id.vars`, [#6867](https://github.com/Rdatatable/data.table/issues/6867). Thanks to Toby Dylan Hocking for the suggestion and PR.
 
+3. `print.data.table()` now shows column classes at the bottom of large tables when `class=TRUE` and `col.names="auto"` (default) for tables with more than 20 rows, [#6902](https://github.com/Rdatatable/data.table/issues/6902). This follows the same behavior as column names at the bottom, making it easier to see column types for large tables without scrolling back to the top. Thanks to @TimTaylor for the suggestion and @Mukulyadav2004 for the PR.
+
+4. `as.Date()` method for `IDate` no longer coerces to `double` [#6922](https://github.com/Rdatatable/data.table/issues/6922). Thanks @MichaelChirico for the report and PR. The only effect should be on overly-strict tests that assert `Date` objects have `double` storage, which is not in general true, especially from R 4.5.0.
+
 ## BUG FIXES
 
 1. Custom binary operators from the `lubridate` package now work with objects of class `IDate` as with a `Date` subclass, [#6839](https://github.com/Rdatatable/data.table/issues/6839). Thanks @emallickhossain for the report and @aitap for the fix.

R/IDateTime.R

@@ -46,7 +46,6 @@ as.IDate.POSIXct = function(x, tz = attr(x, "tzone", exact=TRUE), ...) {
 as.IDate.IDate = function(x, ...) x
 
 as.Date.IDate = function(x, ...) {
-  x = as.numeric(x)
   class(x) = "Date"
   x
 }

R/data.table.R

@@ -174,7 +174,7 @@ replace_dot_alias = function(e) {
         stopf("When by and keyby are both provided, keyby must be TRUE or FALSE")
     }
     if (missing(by)) { missingby=TRUE; by=bysub=NULL }  # possible when env is used, PR#4304
-    else if (verbose && !is.null(env)) catf("Argument '%s' after substitute: %s\n", "by", paste(deparse(bysub, width.cutoff=500L), collapse=" "))
+    else if (verbose && !is.null(env)) catf("Argument '%s' after substitute: %s\n", "by", paste(deparse(bysub, width.cutoff=500L), collapse="\n"))
   }
   bynull = !missingby && is.null(by) #3530
   byjoin = !is.null(by) && is.symbol(bysub) && bysub==".EACHI"
@@ -239,7 +239,7 @@ replace_dot_alias = function(e) {
         substitute2(.j, env),
         list(.j = substitute(j))
       ))
-      if (missing(jsub)) {j = substitute(); jsub=NULL} else if (verbose && !is.null(env)) catf("Argument '%s' after substitute: %s\n", "j", paste(deparse(jsub, width.cutoff=500L), collapse=" "))
+      if (missing(jsub)) {j = substitute(); jsub=NULL} else if (verbose && !is.null(env)) catf("Argument '%s' after substitute: %s\n", "j", paste(deparse(jsub, width.cutoff=500L), collapse="\n"))
     }
   }
   if (!missing(j)) {
@@ -328,7 +328,7 @@ replace_dot_alias = function(e) {
         substitute2(.i, env),
         list(.i = substitute(i))
       ))
-      if (missing(isub)) {i = substitute(); isub=NULL} else if (verbose && !is.null(env)) catf("Argument '%s' after substitute: %s\n", "i", paste(deparse(isub, width.cutoff=500L), collapse=" "))
+      if (missing(isub)) {i = substitute(); isub=NULL} else if (verbose && !is.null(env)) catf("Argument '%s' after substitute: %s\n", "i", paste(deparse(isub, width.cutoff=500L), collapse="\n"))
     }
   }
   if (!missing(i)) {
@@ -904,7 +904,7 @@ replace_dot_alias = function(e) {
             if (length(byvars) == 1L) tt = byvars
             else {
               # take the first variable that is (1) not eval (#3758) and (2) starts with a character that can't start a variable name
-              tt = grep("^eval$|^[^[:alpha:]. ]", byvars, invert=TRUE, value=TRUE)
+              tt = grepv("^eval$|^[^[:alpha:]. ]", byvars, invert=TRUE)
               # byvars but exclude functions or `0`+`1` becomes `+`
               tt = if (length(tt)) tt[1L] else all.vars(bysubl[[jj+1L]])[1L]
             }

R/print.data.table.R

@@ -142,7 +142,11 @@ print.data.table = function(x, topn=getOption("datatable.print.topn"),
   if (nrow(toprint)>20L && col.names == "auto")
     # repeat colnames at the bottom if over 20 rows so you don't have to scroll up to see them
     #   option to shut this off per request of Oleg Bondar on SO, #1482
-    toprint = rbind(toprint, matrix(if (quote) old else colnames(toprint), nrow=1L)) # fixes bug #97
+    toprint = rbind(
+      toprint,
+      matrix(if (quote) old else colnames(toprint), nrow=1L), # see #97
+      if (isTRUE(class)) matrix(if (trunc.cols) abbs[cols_to_print] else abbs, nrow=1L) # #6902
+    )
   print_default(toprint)
   invisible(x)
 }
@@ -165,7 +169,7 @@ shouldPrint = function(x) {
 # for removing the head (column names) of matrix output entirely,
 #   as opposed to printing a blank line, for excluding col.names per PR #1483
 # be sure to remove colnames from any row where they exist, #4270
-cut_colnames = function(x) writeLines(grep("^\\s*(?:[0-9]+:|---)", capture.output(x), value=TRUE))
+cut_colnames = function(x) writeLines(grepv("^\\s*(?:[0-9]+:|---)", capture.output(x)))
 
 # for printing the dims for list columns #3671; used by format.data.table()
 paste_dims = function(x) {

R/test.data.table.R

@@ -458,7 +458,7 @@ test = function(num,x,y=TRUE,error=NULL,warning=NULL,message=NULL,output=NULL,no
       # if a warning containing this string occurs, ignore it. First need for #4182 where warning about 'timedatectl' only
       # occurs in R 3.4, and maybe only on docker too not for users running test.data.table().
       stopifnot(is.character(ignore.warning), !anyNA(ignore.warning), nchar(ignore.warning)>=1L)
-      for (msg in ignore.warning) observed = grep(msg, observed, value=TRUE, invert=TRUE) # allow multiple for translated messages rather than relying on '|' to always work
+      for (msg in ignore.warning) observed = grepv(msg, observed, invert=TRUE) # allow multiple for translated messages rather than relying on '|' to always work
     }
     if (length(expected) != length(observed) && (!foreign || is.null(ignore.warning))) {
       # nocov start
@@ -515,6 +515,8 @@ test = function(num,x,y=TRUE,error=NULL,warning=NULL,message=NULL,output=NULL,no
   }
   if (!fail && !length(error) && (!length(output) || !missing(y))) {   # TODO test y when output=, too
     capture.output(y <- try(y, silent=TRUE)) # y might produce verbose output, just toss it
+    if (inherits(x, c("Date", "POSIXct"))) storage.mode(x) <- "numeric"
+    if (inherits(y, c("Date", "POSIXct"))) storage.mode(y) <- "numeric"
     if (identical(x,y)) return(invisible(TRUE))
     all.equal.result = TRUE
     if (is.data.frame(x) && is.data.frame(y)) {