When checking the encoding of an 8-bit string literal, don't just check the

first codepoint! Also, don't reject empty raw string literals for spurious
"encoding" issues. Also, don't rely on undefined behavior in ConvertUTF.c.

llvm-svn: 152344

2 files changed, 21 insertions, 2 deletions

diff --git a/clang/test/Lexer/cxx0x_raw_string_delim_length.cpp b/clang/test/Lexer/cxx0x_raw_string_delim_length.cpp
index e7d5c6f8cd2..b9f6d13ab74 100644
--- a/clang/test/Lexer/cxx0x_raw_string_delim_length.cpp
+++ b/clang/test/Lexer/cxx0x_raw_string_delim_length.cpp
@@ -1,3 +1,7 @@
-// RUN: %clang_cc1 -std=c++11 -E %s 2>&1 | grep 'error: raw string delimiter longer than 16 characters'
+// RUN: %clang_cc1 -std=c++11 -verify %s
 
-const char *str = R"abcdefghijkmnopqrstuvwxyz(abcdef)abcdefghijkmnopqrstuvwxyz";
+const char *str1 = R"(abcdef)"; // ok
+const char *str2 = R"foo()foo"; // ok
+const char *str3 = R"()"; // ok
+// FIXME: recover better than this.
+const char *str4 = R"abcdefghijkmnopqrstuvwxyz(abcdef)abcdefghijkmnopqrstuvwxyz"; // expected-error {{raw string delimiter longer than 16 characters}} expected-error {{expected expression}}
diff --git a/clang/test/Lexer/string-literal-encoding.c b/clang/test/Lexer/string-literal-encoding.c
index aa7cb73f626..57097dca88a 100644
--- a/clang/test/Lexer/string-literal-encoding.c
+++ b/clang/test/Lexer/string-literal-encoding.c
@@ -15,4 +15,19 @@ void f() {
 
     char const *g = "Àéîõü"; // expected-warning {{illegal character encoding in string literal}}
     char const *h = u8"Àéîõü"; // expected-error {{illegal character encoding in string literal}}
+    char const *i = R"(Àéîõü)"; // expected-warning {{illegal character encoding in string literal}}
+}
+
+void g() {
+    wchar_t const *a = L"foo Àéîõü"; // expected-error {{illegal character encoding in string literal}}
+
+    char16_t const *b = u"foo Àéîõü"; // expected-error {{illegal character encoding in string literal}}
+    char32_t const *c = U"foo Àéîõü"; // expected-error {{illegal character encoding in string literal}}
+    wchar_t const *d = LR"(foo Àéîõü)"; // expected-error {{illegal character encoding in string literal}}
+    char16_t const *e = uR"(foo Àéîõü)"; // expected-error {{illegal character encoding in string literal}}
+    char32_t const *f = UR"(foo Àéîõü)"; // expected-error {{illegal character encoding in string literal}}
+
+    char const *g = "foo Àéîõü"; // expected-warning {{illegal character encoding in string literal}}
+    char const *h = u8"foo Àéîõü"; // expected-error {{illegal character encoding in string literal}}
+    char const *i = R"(foo Àéîõü)"; // expected-warning {{illegal character encoding in string literal}}
 }