Escape invalid code points

base/json/string_escape.cc

Index: base/json/string_escape.cc
diff --git a/base/json/string_escape.cc b/base/json/string_escape.cc
index f67fa93bf2648191564eb1a77d46664036b1f41d..1ba1210ad4708a6b948e5fc8b3745669b346fe21 100644
--- a/base/json/string_escape.cc
+++ b/base/json/string_escape.cc
@@ -91,7 +91,8 @@ bool EscapeJSONStringImpl(const S& str, bool put_in_quotes, std::string* dest) {
 
   for (int32_t i = 0; i < length; ++i) {
     uint32_t code_point;
-    if (!ReadUnicodeCharacter(str.data(), length, &i, &code_point)) {
+    if (!ReadUnicodeCharacter(str.data(), length, &i, &code_point) ||
+        code_point < 0 || !IsValidCharacter(code_point)) {

[Mark Mentovai, 2017/05/24 19:02:48]

How can code_point, a uint32_t, ever be < 0?

[Mark Mentovai, 2017/05/24 19:02:48]

More importantly, you’re confusing characters with codepoints. The documentation
says “Invalid code units will be replaced.” It doesn’t say that invalid
characters will be.

UTF-8 f3 bf bf bf = U+fffff may well be in invalid as a Unicode character, but
it’s valid and unambiguous as a code point.

[sabbakumov, 2017/05/25 03:26:14]

I've updated the documentation.
IsValidCharacter() check is still in the CL because I'm sure we don't want to
write a JSON that we're unable to read using JSONParser.

[sabbakumov, 2017/05/25 03:26:14]

Thank you. I've mistakenly copy-pasted it from here:
https://cs.chromium.org/chromium/src/base/json/json_parser.cc?sq&l=487

[sabbakumov, 2017/05/25 03:26:14]

Done.

[sabbakumov, 2017/05/25 03:26:14]

Done.

       code_point = kReplacementCodePoint;
       did_replacement = true;
     }