From e8a23cf8bc04f3b865063dab0f5624a2a50e697c Mon Sep 17 00:00:00 2001
From: Wilson Snyder <wsnyder@wsnyder.org>
Date: Tue, 26 Jun 2018 07:11:56 -0400
Subject: [PATCH] Fix to ignore Unicode UTF-8 BOM sequences, msg2576.

---
 Changes                     |  2 ++
 src/V3PreLex.l              |  3 +++
 test_regress/t/t_unicode.pl | 50 +++++++++++++++++++++++++++++++++++++
 3 files changed, 55 insertions(+)
 create mode 100755 test_regress/t/t_unicode.pl

diff --git a/Changes b/Changes
index 6200b2026..8a65773c1 100644
--- a/Changes
+++ b/Changes
@@ -27,6 +27,8 @@ The contributors that suggested a given feature are shown in []. Thanks!
 
 ****  Fix define argument stringification (`"), broke since 3.914. [Joe DErrico]
 
+****  Fix to ignore Unicode UTF-8 BOM sequences, msg2576. [HyungKi Jeong]
+
 
 * Verilator 3.924 2018-06-12
 
diff --git a/src/V3PreLex.l b/src/V3PreLex.l
index c763de013..25e2e4b16 100644
--- a/src/V3PreLex.l
+++ b/src/V3PreLex.l
@@ -76,10 +76,13 @@ symb		([a-zA-Z_][a-zA-Z0-9_$]*|\\[^ \t\f\r\n]+)
 symbdef		([a-zA-Z_][a-zA-Z0-9_$]*|\\[^ \t\f\r\n`]+)
 word		[a-zA-Z0-9_]+
 drop		[\032]
+bom		[\357\273\277]
+
 
 	/**************************************************************/
 %%
 
+<INITIAL>{bom}		{ }
 <INITIAL,STRIFY>^{ws}*"`line"{ws}+.*{crnl}	{ LEXP->lineDirective(yytext);
 						  return(VP_LINE); }
 
diff --git a/test_regress/t/t_unicode.pl b/test_regress/t/t_unicode.pl
new file mode 100755
index 000000000..ecc4c14c2
--- /dev/null
+++ b/test_regress/t/t_unicode.pl
@@ -0,0 +1,50 @@
+#!/usr/bin/perl
+if (!$::Driver) { use FindBin; exec("$FindBin::Bin/bootstrap.pl", @ARGV, $0); die; }
+# DESCRIPTION: Verilator: Verilog Test driver/expect definition
+#
+# Copyright 2003 by Wilson Snyder. This program is free software; you can
+# redistribute it and/or modify it under the terms of either the GNU
+# Lesser General Public License Version 3 or the Perl Artistic License
+# Version 2.0.
+use IO::File;
+#use Data::Dumper;
+use strict;
+use vars qw($Self);
+
+scenarios(simulator => 1);
+
+sub gen {
+    my $filename = shift;
+
+    my $fh = IO::File->new(">$filename");
+    $fh->print(chr(0xEF).chr(0xBB).chr(0xBF));  # BOM
+    $fh->print("// Bom\n");
+    $fh->print("// Generated by t_unicode.pl\n");
+    $fh->print("module t;\n");
+    $fh->print("   // Chinese "
+               .chr(0xe8).chr(0xaf).chr(0x84).chr(0xe8).chr(0xae).chr(0xba)  # Comment
+               ."\n");
+    $fh->print("   initial begin\n");
+    $fh->print("      \$write(\"Hello "
+               .chr(0xe4).chr(0xb8).chr(0x96).chr(0xe7).chr(0x95).chr(0x8c)  # World
+               ."\\n\");\n");
+    $fh->print("      \$write(\"*-* All Finished *-*\\n\");\n");
+    $fh->print("      \$finish;\n");
+    $fh->print("   end\n");
+    $fh->print("endmodule\n");
+}
+
+top_filename("$Self->{obj_dir}/t_unicode.v");
+
+gen($Self->{top_filename});
+
+compile(
+    );
+
+execute(
+    check_finished => 1,
+    expect => q{Hello \344\270\226\347\225\214.*},
+    );
+
+ok(1);
+1;