CodeGenModule::GetAddrOfConstantCFString():

- Finish up support for converting UTF8->UTF16 to support ObjC @"string" constants.

Remove warning from CheckObjCString.

As the FIXME in the test case indicates, I still have a bug to work out (apparently with \u handling).

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@68245 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/clang/Basic/DiagnosticSemaKinds.td b/include/clang/Basic/DiagnosticSemaKinds.td
index 74a13478c162c4ed05d6e218ad6abd5392e2fa91..44317c3c6c1d50100f35c453928e15dbb3c506f1 100644 (file)
--- a/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/include/clang/Basic/DiagnosticSemaKinds.td
@@ -1441,8 +1441,6 @@ def err_ret_local_block : Error<
 // CFString checking
 def err_cfstring_literal_not_string_constant : Error<
   "CFString literal is not a string constant">;
-def warn_cfstring_literal_contains_non_ascii_character : Warning<
-  "CFString literal contains non-ASCII character">;
 def warn_cfstring_literal_contains_nul_character : Warning<
   "CFString literal contains NUL character">;
 

diff --git a/lib/CodeGen/CodeGenModule.cpp b/lib/CodeGen/CodeGenModule.cpp
index 2e84c6049d5f72df0a2af9a11a92d93e41aea58a..44a045ef8a51a9272d71da7333126d84cb50c231 100644 (file)
--- a/lib/CodeGen/CodeGenModule.cpp
+++ b/lib/CodeGen/CodeGenModule.cpp
@@ -1000,10 +1000,11 @@ static void appendFieldAndPadding(CodeGenModule &CGM,
   }
 }
 
-// We still need to work out the details of handling UTF-16. 
-// See: <rdr://2996215>
 llvm::Constant *CodeGenModule::
 GetAddrOfConstantCFString(const StringLiteral *Literal) {
+  std::string str;
+  unsigned StringLength;
+  
   bool isUTF16 = false;
   if (Literal->containsNonAsciiOrNull()) {
     // Convert from UTF-8 to UTF-16.
@@ -1016,10 +1017,14 @@ GetAddrOfConstantCFString(const StringLiteral *Literal) {
                                 &ToPtr, ToPtr+Literal->getByteLength(),
                                 strictConversion);
     assert(Result == conversionOK && "UTF-8 to UTF-16 conversion failed");
+    
+    StringLength = ToPtr-&ToBuf[0];
+    str.assign((char *)&ToBuf[0], StringLength*2); // Twice as many UTF8 chars.
     isUTF16 = true;
-    // FIXME: Do something with the converted value!
+  } else {
+    str.assign(Literal->getStrData(), Literal->getByteLength());
+    StringLength = str.length();
   }
-  std::string str(Literal->getStrData(), Literal->getByteLength());
   llvm::StringMapEntry<llvm::Constant *> &Entry = 
     CFConstantStringMap.GetOrCreateValue(&str[0], &str[str.length()]);
   
@@ -1093,7 +1098,7 @@ GetAddrOfConstantCFString(const StringLiteral *Literal) {
   NextField = 0;
   Ty = getTypes().ConvertType(getContext().LongTy);
   appendFieldAndPadding(*this, Fields, CurField, NextField,
-                        llvm::ConstantInt::get(Ty, str.length()), CFRD, STy);
+                        llvm::ConstantInt::get(Ty, StringLength), CFRD, STy);
   
   // The struct.
   C = llvm::ConstantStruct::get(STy, Fields);

diff --git a/lib/Sema/SemaChecking.cpp b/lib/Sema/SemaChecking.cpp
index 4bf8fc9dce1801132ed61c761b75d37ae92503da..1802b6e942541b584d1dd3438da23f9e4f03707a 100644 (file)
--- a/lib/Sema/SemaChecking.cpp
+++ b/lib/Sema/SemaChecking.cpp
@@ -173,13 +173,6 @@ bool Sema::CheckObjCString(Expr *Arg) {
   unsigned Length = Literal->getByteLength();
   
   for (unsigned i = 0; i < Length; ++i) {
-    if (!isascii(Data[i])) {
-      Diag(getLocationOfStringLiteralByte(Literal, i),
-           diag::warn_cfstring_literal_contains_non_ascii_character)
-        << Arg->getSourceRange();
-      break;
-    }
-    
     if (!Data[i]) {
       Diag(getLocationOfStringLiteralByte(Literal, i),
            diag::warn_cfstring_literal_contains_nul_character)

diff --git a/test/SemaObjC/ucn-objc-string.m b/test/SemaObjC/ucn-objc-string.m
new file mode 100644 (file)
index 0000000..918fbc2
--- /dev/null
+++ b/test/SemaObjC/ucn-objc-string.m
@@ -0,0 +1,13 @@
+// RUN: clang %s -verify -fsyntax-only
+@class NSString;
+extern void NSLog(NSString *format, ...) __attribute__((format(__NSString__, 1, 2)));
+
+int main() {
+  NSLog(@"Hi…");
+  NSLog(@"Exposé");
+  // FIXME: the following 2 are still not working (will investigate).
+  //NSLog(@"hello \u2192 \u2603 \u2190 world");
+  //NSLog(@"\U00010400\U0001D12B");
+  return 0;
+}
+