Merge branch 'master' into contributing-restore

Author: MichaelChirico

…vcontainer/r-devel-clang-ubsan/.Rprofile .devcontainer/.Rprofile.devcontainer/r-devel-clang-ubsan/.Rprofile renamed to .devcontainer/.Rprofile .devcontainer/r-devel-clang-ubsan/.Rprofile renamed to .devcontainer/.Rprofile

@@ -1,5 +1,5 @@
 {
-  "build": { "dockerfile": "Dockerfile", "context": "../.." },
+  "build": { "dockerfile": "r-devel-gcc/Dockerfile", "context": ".." },
   "customizations": { "vscode": {
      "extensions": [
       "REditorSupport.r",

…vcontainer/r-devel-gcc/devcontainer.json .devcontainer/devcontainer.json.devcontainer/r-devel-gcc/devcontainer.json renamed to .devcontainer/devcontainer.json .devcontainer/r-devel-gcc/devcontainer.json renamed to .devcontainer/devcontainer.json

@@ -16,7 +16,7 @@ read.dcf("DESCRIPTION", c("Imports", "Suggests")) |>      \
 
 # setup cc()
 WORKDIR /root
-COPY .devcontainer/r-devel-clang-ubsan/.Rprofile .
+COPY .devcontainer/.Rprofile .
 
 # set ubsan flags
 WORKDIR .R

.devcontainer/r-devel-clang-ubsan/Dockerfile

@@ -15,4 +15,4 @@ read.dcf("DESCRIPTION", c("Imports", "Suggests")) |>      \
 
 # setup cc()
 WORKDIR /root
-COPY .devcontainer/r-devel-gcc/.Rprofile .
+COPY .devcontainer/.Rprofile .

.devcontainer/r-devel-gcc/.Rprofile

@@ -202,7 +202,7 @@
     #  v1.14.4  0.4826  0.5586  0.6586  0.6329  0.7348  1.318   100
     ```
 
-30. `rbind()` and `rbindlist()` now support `fill=TRUE` with `use.names=FALSE` instead of issuing the warning `use.names= cannot be FALSE when fill is TRUE. Setting use.names=TRUE.`, [#5444](https://github.com/Rdatatable/data.table/issues/5444). Thanks to @sindribaldur for testing dev and filing a bug report which was fixed before release.
+30. `rbind()` and `rbindlist()` now support `fill=TRUE` with `use.names=FALSE` instead of issuing the warning `use.names= cannot be FALSE when fill is TRUE. Setting use.names=TRUE.`, [#5444](https://github.com/Rdatatable/data.table/issues/5444). Thanks to @sindribaldur, @dcaseykc, @fox34, @adrian-quintario and @berg-michael for testing dev and filing a bug report which was fixed before release.
 
     ```R
     DT1

.devcontainer/r-devel-gcc/Dockerfile

@@ -180,7 +180,7 @@ 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) 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=" "))
   }
   bynull = !missingby && is.null(by) #3530
   byjoin = !is.null(by) && is.symbol(bysub) && bysub==".EACHI"
@@ -244,7 +244,7 @@ replace_dot_alias = function(e) {
         substitute2(.j, env),
         list(.j = substitute(j))
       ))
-      if (missing(jsub)) {j = substitute(); jsub=NULL} else if (verbose) 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=" "))
     }
   }
   if (!missing(j)) {
@@ -332,7 +332,7 @@ replace_dot_alias = function(e) {
         substitute2(.i, env),
         list(.i = substitute(i))
       ))
-      if (missing(isub)) {i = substitute(); isub=NULL} else if (verbose) 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=" "))
     }
   }
   if (!missing(i)) {

NEWS.md

@@ -96,8 +96,20 @@ merge.data.table = function(x, y, by = NULL, by.x = NULL, by.y = NULL, all = FAL
   if (all.y && nrow(y)) {  # If y does not have any rows, no need to proceed
     # Perhaps not very commonly used, so not a huge deal that the join is redone here.
     missingyidx = y[!x, which=TRUE, on=by, allow.cartesian=allow.cartesian]
+    # TO DO: replace by following once #5446 is merged
+    # if (length(missingyidx)) dt = rbind(dt, y[missingyidx], use.names=FALSE, fill=TRUE, ignore.attr=TRUE)
     if (length(missingyidx)) {
-      dt = rbind(dt, y[missingyidx], use.names=FALSE, fill=TRUE)
+      yy = y[missingyidx]
+      othercolsx = setdiff(nm_x, by)
+      if (length(othercolsx)) {
+        # create NA rectangle with correct types and attributes of x to cbind to y
+        tmp = rep.int(NA_integer_, length(missingyidx))
+        # TO DO: use set() here instead..
+        yy = cbind(yy, x[tmp, othercolsx, with = FALSE])
+      }
+      # empty data.tables (nrow =0, ncol>0) doesn't skip names anymore in new rbindlist
+      # takes care of #24 without having to save names. This is how it should be, IMHO.
+      dt = rbind(dt, yy, use.names=FALSE)
     }
   }
   # X[Y] syntax puts JIS i columns at the end, merge likes them alongside i.

R/data.table.R

@@ -1919,6 +1919,11 @@ test(631.5, DT3, data.table(a=c(2), total=c(5), key="a"))
 # .. nrow(y)=1, i subset y with 1 and match with x
 test(631.6, merge(DT1,DT4,all.y=TRUE), data.table(a=c(3),total.x=c(1),total.y=c(1),key="a"))
 test(631.7, DT4, data.table(a=c(3), total=c(1), key="a"))
+# merge columns with different attributes #5309
+x = data.table(a=1L, b=as.IDate(16801))
+y = data.table(a=2L, b=NA)
+test(631.8, merge(x,y,by="a",all=TRUE), data.table(a=c(1L,2L), b.x=as.IDate(c(16801,NA)), b.y=NA, key="a"))
+test(631.9, merge(y,x,by="a",all=TRUE), data.table(a=c(1L,2L), b.x=NA, b.y=as.IDate(c(16801,NA)), key="a"))
 
 test(632, merge(DT1,DT2,all=TRUE), data.table(a=c(1,2,3,4,5),total.x=c(2,NA,1,3,1),total.y=c(NA,5,1,NA,2),key="a"))
 test(632.1, merge(DT1,DT2,all=TRUE), setkey(adt(merge(adf(DT1),adf(DT2),by="a",all=TRUE)),a))
@@ -14335,6 +14340,14 @@ test(2003.5, rbindlist(list(data.table(a=1:2), data.table(b=3:4, c=5:6)), fill=T
 # rbindlist segfault with fill=TRUE and usenames=FALSE #5444
 test(2003.6, rbindlist(list(list(1), list(2,3)), fill=TRUE, use.names=FALSE), data.table(c(1,2), c(NA, 3)))
 test(2003.7, rbindlist(list(list(1), list(2,factor(3))), fill=TRUE, use.names=FALSE), data.table(c(1,2), factor(c(NA, 3))))
+# rbind with different attributes #5309
+x=data.table(a=as.Date(NA))
+y=data.table(a=as.Date('2021-10-05'), b=as.POSIXct("2021-10-06 13:58:00 UTC"))
+ans=data.table(a=as.Date(c(NA_character_, '2021-10-05')), b=as.POSIXct(c(NA_character_, "2021-10-06 13:58:00 UTC")))
+test(2003.81, rbind(x, y, fill=TRUE, use.names=TRUE), ans)
+test(2003.82, rbind(y, x, fill=TRUE, use.names=TRUE), ans[2:1,])
+test(2003.83, rbind(x, y, fill=TRUE, use.names=FALSE), ans)
+test(2003.84, rbind(y, x, fill=TRUE, use.names=FALSE), ans[2:1,])
 
 # chmatch coverage for two different non-ascii encodings matching; issues mentioned in comments in chmatch.c #69 #2538 #111
 x1 = "fa\xE7ile"