allow UTF-8 in rule descriptions

The lexer already mostly allowed this, except that chars >127 were
being interpreted as negative indexes into the lexer table.

1 files changed, 7 insertions, 0 deletions

diff --git a/src/parsers_test.cc b/src/parsers_test.cc
index a8bf179..c5151b8 100644
--- a/src/parsers_test.cc
+++ b/src/parsers_test.cc
@@ -658,3 +658,10 @@ TEST_F(ParserTest, DefaultStatements) {
   EXPECT_EQ("b", nodes[1]->path());
   EXPECT_EQ("c", nodes[2]->path());
 }
+
+TEST_F(ParserTest, UTF8) {
+  ASSERT_NO_FATAL_FAILURE(AssertParse(
+"rule utf8\n"
+"  command = true\n"
+"  description = compilaci\xC3\xB3\n"));
+}