Add Chinese error message translations

.dev/CRAN_Release.cmd

@@ -1,3 +1,69 @@
+###############################################
+#  Updating translations
+###############################################
+
+# 1) Update messages for new release
+## (a) Update C template file: src/data.table.pot
+##     ideally, we are including _() wrapping in
+##     new PRs throughout dev cycle, and this step
+##     becomes about tying up loose ends
+## Appending _() char array wrapping to all messages
+##   that might be shown to the user. This step is slightly
+##   too greedy, as it includes too many msg, some of which
+##   need not be translated [more work to do here to make
+##   this less manual] some things to watch out for:
+##     * quote embedded (and escaped) within message [could be fixed with smarter regex]
+##     * multi-line implicit-concat arrays (in C, `"a" "b"` is the same as `"ab"`) should be wrapped "on the outside" not individually
+##     * `data.table` shares some of its `src` with `pydatatable`, so the requirement to `#include <R.h>` before the `#define _` macro meant we need to be careful about including this macro only in the R headers for these files (hence I created `po.h`)
+##     * Can't use `_()` _inside_ another functional macro. Only wrap the string passed to the macro later.
+for MSG in error warning DTWARN DTPRINT Rprintf STOP Error;
+  do for SRC_FILE in src/*.c;
+    # no inplace -i in default mac sed
+    do sed -E "s/$MSG[(]("[^"]*")/$MSG(_(\1)/g" $SRC_FILE > out;
+    mv out $SRC_FILE;
+  done
+done
+
+## checking for other lines calling these that didn't get _()-wrapped
+for MSG in error warning DTWARN DTPRINT Rprintf STOP Error;
+  do grep -Er "\b$MSG[(]" src --include=*.c | grep -v _ | grep -Ev "(?://|[*]).*$MSG[(]"
+
+## similar, but a bit more manual to check snprintf usage
+
+## look for char array that haven't been covered yet
+grep -Er '"[^"]+"' src --include=*.c | grep -Fv '_("' | grep -v "#include" | grep -v '//.*".*"'
+
+## look for lines starting with a char array (likely continued from prev line & can be combined)
+grep -Er '^\s*"' src/*.c
+
+## Now extract these messages with xgettext
+cd src
+xgettext --keyword=_ -o data.table.pot *.c
+cd ..
+
+## (b) Update R template file: src/R-data.table.pot
+## much easier, once the update_pkg_po bug is fixed
+R --no-save
+## a bug fix in R still hadn't made the 2019-12-12 release,
+##   so run the following to source the corrected function manually
+STEM='https://raw.githubusercontent.com/wch/r-source/trunk/src/library/tools/R'
+source(file.path(STEM, 'utils.R'))
+source(file.path(STEM, 'xgettext.R'))
+source(file.path(STEM, 'translations.R'))
+## shouldn't be any errors from this...
+update_pkg_po('.')
+q()
+
+# 2) Open a PR with the new templates & contact the translators
+#   * zh_CN:
+## Translators to submit commits with translations to this PR
+##   [or perhaps, if we get several languages, each to open
+##    its own PR and merge to main translation PR]
+
+## 3) Check validity
+##   update_pkg_po('.') to be run again for the PR
+##     [can this be done via Travis?]
+
 ###############################################
 #  Basic checks
 ###############################################

R/between.R

@@ -1,6 +1,6 @@
 # is x[i] in between lower[i] and upper[i] ?
 between = function(x, lower, upper, incbounds=TRUE, NAbounds=TRUE, check=FALSE) {
-  if (is.logical(x)) stop("between has been x of type logical")
+  if (is.logical(x)) stop("between has been passed an argument x of type logical")
   if (is.logical(lower)) lower = as.integer(lower)   # typically NA (which is logical type)
   if (is.logical(upper)) upper = as.integer(upper)   # typically NA (which is logical type)
   is.px = function(x) inherits(x, "POSIXct")

R/data.table.R

@@ -183,7 +183,7 @@ replace_dot_alias = function(e) {
     }
     return(x)
   }
-  if (!mult %chin% c("first","last","all")) stop("mult argument can only be 'first','last' or 'all'")
+  if (!mult %chin% c("first","last","all")) stop("mult argument can only be 'first', 'last' or 'all'")
   missingroll = missing(roll)
   if (length(roll)!=1L || is.na(roll)) stop("roll must be a single TRUE, FALSE, positive/negative integer/double including +Inf and -Inf or 'nearest'")
   if (is.character(roll)) {
@@ -820,7 +820,7 @@ replace_dot_alias = function(e) {
           if (!typeof(byval[[jj]]) %chin% ORDERING_TYPES) stop("column or expression ",jj," of 'by' or 'keyby' is type ",typeof(byval[[jj]]),". Do not quote column names. Usage: DT[,sum(colC),by=list(colA,month(colB))]")
         }
         tt = vapply_1i(byval,length)
-        if (any(tt!=xnrow)) stop("The items in the 'by' or 'keyby' list are length (",paste(tt,collapse=","),"). Each must be length ", xnrow, "; the same length as there are rows in x (after subsetting if i is provided).")
+        if (any(tt!=xnrow)) stop(gettextf("The items in the 'by' or 'keyby' list are length(s) (%s). Each must be length %d; the same length as there are rows in x (after subsetting if i is provided).", paste(tt, collapse=","), xnrow, domain='R-data.table'))
         if (is.null(bynames)) bynames = rep.int("",length(byval))
         if (length(idx <- which(!nzchar(bynames))) && !bynull) {
           # TODO: improve this and unify auto-naming of jsub and bysub
@@ -874,7 +874,7 @@ replace_dot_alias = function(e) {
             # attempt to auto-name unnamed columns
             for (jj in which(nm=="")) {
               thisq = q[[jj + 1L]]
-              if (missing(thisq)) stop("Item ", jj, " of the .() or list() passed to j is missing") #3507
+              if (missing(thisq)) stop(gettextf("Item %d of the .() or list() passed to j is missing", jj, domain="R-data.table")) #3507
               if (is.name(thisq)) nm[jj] = drop_dot(thisq)
               # TO DO: if call to a[1] for example, then call it 'a' too
             }

R/devel.R

@@ -13,7 +13,7 @@ dcf.repo = function(pkg, repo, field, type){
   idx = file(file.path(contrib.url(repo, type=type),"PACKAGES"))
   on.exit(close(idx))
   dcf = read.dcf(idx, fields=c("Package",field))
-  if (!pkg %in% dcf[,"Package"]) stop("There is no ", pkg, " package in provided repository.")
+  if (!pkg %in% dcf[,"Package"]) stop(gettextf("There is no package %s in provided repository.", pkg, domain='R-data.table'))
   dcf[dcf[,"Package"]==pkg, field][[1L]]
 }
 

R/groupingsets.R

@@ -60,10 +60,10 @@ groupingsets.data.table = function(x, j, by, sets, .SDcols, id = FALSE, jj, ...)
     stop("All columns used in 'sets' argument must be in 'by' too. Columns used in 'sets' but not present in 'by': ", brackify(setdiff(sets.all.by, by)))
   if (id && "grouping" %chin% names(x))
     stop("When using `id=TRUE` the 'x' data.table must not have a column named 'grouping'.")
-  if (!all(sapply(sets, function(x) length(x)==uniqueN(x))))
+  if (any(sapply(sets, anyDuplicated)))
     stop("Character vectors in 'sets' list must not have duplicated column names within a single grouping set.")
-  if (!identical(lapply(sets, sort), unique(lapply(sets, sort))))
-    warning("Double counting is going to happen. Argument 'sets' should be unique without taking order into account, unless you really want double counting, then get used to that warning. Otherwise `sets=unique(lapply(sets, sort))` will do the trick.")
+  if (length(sets) > 1L && (idx<-anyDuplicated(lapply(sets, sort))))
+    warning("'sets' contains a duplicate (i.e., equivalent up to sorting) element at index ", idx, "; as such, there will be duplicate rows in the output -- note that grouping by A,B and B,A will produce the same aggregations. Use `sets=unique(lapply(sets, sort))` to eliminate duplicates.")
   # input arguments handling
   jj = if (!missing(jj)) jj else substitute(j)
   av = all.vars(jj, TRUE)

R/last.R

@@ -35,7 +35,7 @@ last = function(x, n=1L, ...) {
     }
   } else {
     if (!requireNamespace("xts", quietly=TRUE))
-      stop("'xts' class passed to data.table::last function but 'xts' is not available, you should have 'xts' installed already") # nocov
+      stop(gettextf("'xts' class passed to %s function but 'xts' is not available, you should have 'xts' installed already", "data.table::last", domain="R-data.table")) # nocov
     if (verbose)
       cat("last: using xts::last: is.xts(x)\n")
     xts::last(x, n=n, ...)
@@ -76,7 +76,7 @@ first = function(x, n=1L, ...) {
     }
   } else {
     if (!requireNamespace("xts", quietly=TRUE))
-      stop("'xts' class passed to data.table::first function but 'xts' is not available, you should have 'xts' installed already") # nocov
+      stop(gettextf("'xts' class passed to %s function but 'xts' is not available, you should have 'xts' installed already", "data.table::first", domain="R-data.table")) # nocov
     if (verbose)
       cat("first: using xts::first: is.xts(x)\n")
     xts::first(x, n=n, ...)

R/onAttach.R

@@ -20,6 +20,8 @@
   if (!isTRUE(getOption("datatable.quiet"))) {   # new option in v1.12.4, #3489
     packageStartupMessage("data.table ", v, if(dev)paste0(" IN DEVELOPMENT built ",d,g),
                           " using ", getDTthreads(verbose=FALSE), " threads (see ?getDTthreads).  Latest news: r-datatable.com")
+    if (gettext("TRANSLATION CHECK", domain='R-data.table') != "TRANSLATION CHECK")
+      packageStartupMessage(gettext("**********\n", "Running data.table in English; package support is available in English only. When searching for online help, be sure to also check for the English error message. This can be obtained by looking at the po/R-<locale>.po and po/<locale>.po files in the package source, where the native language and English error messages can be found side-by-side"))
     if (dev && (Sys.Date() - as.Date(d))>28L)
       packageStartupMessage("**********\nThis development version of data.table was built more than 4 weeks ago. Please update: data.table::update.dev.pkg()\n**********")
     if (!.Call(ChasOpenMP))

R/setkey.R

@@ -251,7 +251,7 @@ setorder = function(x, ..., na.last=FALSE)
 # na.last=FALSE here, to be consistent with data.table's default
 # as opposed to DT[order(.)] where na.last=TRUE, to be consistent with base
 {
-  if (!is.data.frame(x)) stop("x must be a data.frame or data.table.")
+  if (!is.data.frame(x)) stop("x must be a data.frame or data.table")
   cols = substitute(list(...))[-1L]
   if (identical(as.character(cols),"NULL")) return(x)
   if (length(cols)) {
@@ -357,7 +357,7 @@ CJ = function(..., sorted = TRUE, unique = FALSE)
     }
   }
   nrow = prod( vapply_1i(l, length) )  # lengths(l) will work from R 3.2.0
-  if (nrow > .Machine$integer.max) stop("Cross product of elements provided to CJ() would result in ",nrow," rows which exceeds .Machine$integer.max == ",.Machine$integer.max)
+  if (nrow > .Machine$integer.max) stop(gettextf("Cross product of elements provided to CJ() would result in %.0f rows which exceeds .Machine$integer.max == %d", nrow, .Machine$integer.max, domain='R-data.table'))
   l = .Call(Ccj, l)
   setDT(l)
   l = setalloccol(l)  # a tiny bit wasteful to over-allocate a fixed join table (column slots only), doing it anyway for consistency since

R/test.data.table.R

@@ -43,7 +43,7 @@ test.data.table = function(script="tests.Rraw", verbose=FALSE, pkg=".", silent=F
     # nocov start
     fn2 = paste0(fn,".bz2")
     if (!file.exists(file.path(fulldir, fn2)))
-      stop("Neither ",fn," or ",fn2," exist in ",fulldir)
+      stop(gettextf("Neither %s nor %s exist in %s",fn, fn2, fulldir, domain="R-data.table"))
     fn = fn2
     # nocov end
     # sys.source() below accepts .bz2 directly.