Bug 1478170 - Implement SourceUnits::peekCodePoint for UTF-8. r=arai

--HG--
extra : rebase_source : 7d256802f430ba813588fe9535630f67a30fbd56

js/src/frontend/TokenStream.h

@@ -185,6 +185,7 @@
 #include "mozilla/Attributes.h"
 #include "mozilla/Casting.h"
 #include "mozilla/DebugOnly.h"
+#include "mozilla/Maybe.h"
 #include "mozilla/MemoryChecking.h"
 #include "mozilla/PodOperations.h"
 #include "mozilla/TextUtils.h"
@@ -1089,6 +1090,27 @@ PeekCodePoint(const char16_t* const ptr, const char16_t* const end)
     return PeekedCodePoint<char16_t>(c, len);
 }
 
+inline PeekedCodePoint<mozilla::Utf8Unit>
+PeekCodePoint(const mozilla::Utf8Unit* const ptr, const mozilla::Utf8Unit* const end)
+{
+    if (MOZ_UNLIKELY(ptr >= end))
+        return PeekedCodePoint<mozilla::Utf8Unit>::none();
+
+    const mozilla::Utf8Unit lead = ptr[0];
+    if (mozilla::IsAscii(lead))
+        return PeekedCodePoint<mozilla::Utf8Unit>(lead.toUint8(), 1);
+
+    const mozilla::Utf8Unit* afterLead = ptr + 1;
+    mozilla::Maybe<char32_t> codePoint = mozilla::DecodeOneUtf8CodePoint(lead, &afterLead, end);
+    if (codePoint.isNothing())
+        return PeekedCodePoint<mozilla::Utf8Unit>::none();
+
+    auto len = mozilla::AssertedCast<uint8_t>(mozilla::PointerRangeSize(ptr, afterLead));
+    MOZ_ASSERT(len <= 4);
+
+    return PeekedCodePoint<mozilla::Utf8Unit>(codePoint.value(), len);
+}
+
 // This is the low-level interface to the JS source code buffer.  It just gets
 // raw Unicode code units -- 16-bit char16_t units of source text that are not
 // (always) full code points, and 8-bit units of UTF-8 source text soon.

mfbt/Utf8.h

@@ -198,6 +198,13 @@ union Utf8Unit
   // that other code.
 };
 
+/** Returns true iff |aUnit| is an ASCII value. */
+inline bool
+IsAscii(Utf8Unit aUnit)
+{
+  return IsAscii(aUnit.toUint8());
+}
+
 /**
  * Returns true if the given length-delimited memory consists of a valid UTF-8
  * string, false otherwise.