Merge branch 'main' into wasm-split-remove-unused

WebAssembly · Sep 16, 2024 · b73586a · b73586a
2 parents 9dccd15 + 9292310
commit b73586a
Show file tree

Hide file tree

Showing 8 changed files with 361 additions and 337 deletions.
diff --git a/src/parser/lexer.cpp b/src/parser/lexer.cpp
@@ -280,6 +280,13 @@ struct LexStrResult : LexResult {
   // Allocate a string only if there are escape sequences, otherwise just use
   // the original string_view.
   std::optional<std::string> str;
+
+  std::string_view getStr() {
+    if (str) {
+      return *str;
+    }
+    return span;
+  }
 };
 
 struct LexStrCtx : LexCtx {
@@ -860,6 +867,9 @@ std::optional<LexIdResult> ident(std::string_view in) {
     return {};
   }
   if (auto s = str(ctx.next())) {
+    if (!String::isUTF8(s->getStr())) {
+      return {};
+    }
     ctx.isStr = true;
     ctx.str = s->str;
     ctx.take(*s);

diff --git a/src/passes/StringLowering.cpp b/src/passes/StringLowering.cpp
@@ -153,9 +153,12 @@ struct StringGathering : public Pass {
       [[maybe_unused]] bool valid =
         String::convertWTF16ToWTF8(wtf8, string.str);
       assert(valid);
-      // TODO: Use wtf8.view() once we have C++20.
+      // Then escape it because identifiers must be valid UTF-8.
+      // TODO: Use wtf8.view() and escaped.view() once we have C++20.
+      std::stringstream escaped;
+      String::printEscaped(escaped, wtf8.str());
       auto name = Names::getValidGlobalName(
-        *module, std::string("string.const_") + std::string(wtf8.str()));
+        *module, std::string("string.const_") + std::string(escaped.str()));
       globalName = name;
       newNames.insert(name);
       auto* stringConst = builder.makeStringConst(string);