if we read an utf-8 BOM, enable the utf8-hack flag (we can test for that on

mirbsd even because the main.c enabling is only run for interactive shells)

check.t

@@ -1,4 +1,4 @@
-# $MirOS: src/bin/mksh/check.t,v 1.97 2007/04/15 12:09:56 tg Exp $
+# $MirOS: src/bin/mksh/check.t,v 1.98 2007/04/15 12:28:37 tg Exp $
 # $OpenBSD: bksl-nl.t,v 1.2 2001/01/28 23:04:56 niklas Exp $
 # $OpenBSD: history.t,v 1.5 2001/01/28 23:04:56 niklas Exp $
 # $OpenBSD: read.t,v 1.3 2003/03/10 03:48:16 david Exp $
@@ -3950,3 +3950,14 @@ expected-stdout:
 	a=BAR
 	a=BAR
 ---
+name: utf8bom-3
+description:
+	Reading the UTF-8 BOM should enable the utf8-hack flag
+category: pdksh
+stdin:
+	$0 -c ':; x=$(set +o); if [[ $x = *utf8* ]]; then print on; else print off; fi'
+	$0 -c ':; x=$(set +o); if [[ $x = *utf8* ]]; then print on; else print off; fi'
+expected-stdout:
+	off
+	on
+---

lex.c

@@ -2,7 +2,7 @@
 
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/lex.c,v 1.27 2007/04/15 10:45:58 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/lex.c,v 1.28 2007/04/15 12:28:38 tg Exp $");
 
 /* Structure to keep track of the lexing state and the various pieces of info
  * needed for each particular state. */
@@ -957,6 +957,7 @@ getsc__(void)
 		    (((const unsigned char *)(s->str))[0] == 0xBB) &&
 		    (((const unsigned char *)(s->str))[1] == 0xBF)) {
 			s->str += 2;
+			Flag(FUTFHACK) = 1;
 			goto getsc_again;
 		}
 	}

mksh.1

@@ -1,7 +1,7 @@
-.\" $MirOS: src/bin/mksh/mksh.1,v 1.81 2007/03/10 21:14:07 tg Exp $
+.\" $MirOS: src/bin/mksh/mksh.1,v 1.82 2007/04/15 12:28:38 tg Exp $
 .\" $OpenBSD: ksh.1,v 1.118 2006/11/30 08:47:58 jmc Exp $
 .\"
-.Dd March 10, 2007
+.Dd April 15, 2007
 .Dt MKSH 1
 .Os MirBSD
 .Sh NAME
@@ -3269,7 +3269,8 @@ and optionally
 and at least one of these returns something that matches
 .Dq UTF-8
 or
-.Dq utf8 .
+.Dq utf8 ,
+or if the input begins with a Byte Order Mark.
 .It Fl u \*(Ba Ic nounset
 Referencing of an unset parameter is treated as an error, unless one of the
 .Ql - ,