fread respects encoding on header cols as well, closes #1680.

NEWS.md

@@ -191,6 +191,8 @@
 
   54. `chmatch()` handles `nomatch = integer(0)` properly, [#1672](https://github.com/Rdatatable/data.table/issues/1672).
 
+  55. `fread()` did not respect encoding on header column. Now fixed, [#1680](https://github.com/Rdatatable/data.table/issues/1680). Thanks @nachti.
+
 #### NOTES
 
   1. Updated error message on invalid joins to reflect the new `on=` syntax, [#1368](https://github.com/Rdatatable/data.table/issues/1368). Thanks @MichaelChirico.

inst/tests/tests.Rraw

@@ -8824,6 +8824,12 @@ x = data.table(Date = as.Date(c("2015-12-29", "2015-12-29", "2015-12-29", "2015-
 y = CJ(Date = as.Date(c("2015-12-31", "2016-01-31", "2016-02-29", "2016-03-31")), ID = unique(x$ID))
 test(1669, x[y, on=c("ID", "Date"), roll=TRUE, which=TRUE], 1:16)
 
+# 1680 fix, fread header encoding issue
+x = "Stra\xdfe"
+Encoding(x) = "latin1"
+nm = names(fread("1680-fread-header-encoding.csv", encoding="Latin-1"))
+test(1670, nm[2], x)
+
 ##########################
 
 # TODO: Tests involving GForce functions needs to be run with optimisation level 1 and 2, so that both functions are tested all the time.

src/fread.c

@@ -929,7 +929,7 @@ SEXP readfile(SEXP input, SEXP separg, SEXP nrowsarg, SEXP headerarg, SEXP nastr
             // Also Fiedl() it takes care of leading spaces. Easier to understand the logic.
             skip_spaces(); Field();
             if (fieldLen) {
-                SET_STRING_ELT(names, i, mkCharLen(fieldStart, fieldLen));
+                SET_STRING_ELT(names, i, mkCharLenCE(fieldStart, fieldLen, ienc)); // #1680 fix, respect encoding on header col
             } else {
                 sprintf(buff,"V%d",i+1);
                 SET_STRING_ELT(names, i, mkChar(buff));