From 1a9165f7fce2fe325a0381ef056eb00f8d3014f1 Mon Sep 17 00:00:00 2001
From: Jason Turner <jason@emptycrate.com>
Date: Fri, 2 Mar 2018 07:45:24 -0700
Subject: [PATCH] Normalize on C++'s standards for \u and \U

---
 include/chaiscript/language/chaiscript_parser.hpp | 14 +++++++++++++-
 unittests/string_unicode_unicode.chai             |  6 +++---
 2 files changed, 16 insertions(+), 4 deletions(-)

diff --git a/include/chaiscript/language/chaiscript_parser.hpp b/include/chaiscript/language/chaiscript_parser.hpp
index 2c957ca8..1cac8d95 100644
--- a/include/chaiscript/language/chaiscript_parser.hpp
+++ b/include/chaiscript/language/chaiscript_parser.hpp
@@ -1064,6 +1064,13 @@ namespace chaiscript
           }
         }
 
+        void finalize_unicode()
+        {
+          if (unicode_size > 0) {
+            process_unicode();
+          }
+        }
+
         void process_hex()
         {
           auto val = stoll(hex_matches, nullptr, 16);
@@ -1087,17 +1094,20 @@ namespace chaiscript
         void process_unicode()
         {
           const auto ch = static_cast<uint32_t>(std::stoi(hex_matches, nullptr, 16));
+          const auto match_size = hex_matches.size();
           hex_matches.clear();
           is_escaped = false;
           const auto u_size = unicode_size;
           unicode_size = 0;
 
           char buf[4];
+          if (u_size != match_size) {
+            throw exception::eval_error("Incomplete unicode escape sequence");
+          }
           if (u_size == 4 && ch >= 0xD800 && ch <= 0xDFFF) {
             throw exception::eval_error("Invalid 16 bit universal character");
           }
 
-          unicode_size = 0;
 
           if (ch < 0x80) {
             match += static_cast<char>(ch);
@@ -1289,6 +1299,7 @@ namespace chaiscript
               }
             }
 
+            cparser.finalize_unicode();
             return cparser.is_interpolated;
           }();
 
@@ -1347,6 +1358,7 @@ namespace chaiscript
             for (auto s = start + 1, end = m_position - 1; s != end; ++s) {
               cparser.parse(*s, start.line, start.col, *m_filename);
             }
+            cparser.finalize_unicode();
           }
 
           if (match.size() != 1) {
diff --git a/unittests/string_unicode_unicode.chai b/unittests/string_unicode_unicode.chai
index 267fbb64..d79dcaf3 100644
--- a/unittests/string_unicode_unicode.chai
+++ b/unittests/string_unicode_unicode.chai
@@ -2,15 +2,15 @@ assert_equal("\uc39c", "쎜")
 assert_equal("U for \u00dcmlauts", "U for Ümlauts")
 
 assert_equal("Thorn \u00fe sign", "Thorn þ sign")
-assert_equal("Test\u20Me", "Test Me")
+assert_equal("Test\u0020Me", "Test Me")
 assert_equal("Test\u2022Me", "Test•Me")
-//assert_equal("Test\uDD34\uD83DMe", "Test🔴Me")
 
 assert_equal("\xF0\x9F\x8D\x8C", "🍌")
 assert_equal("\U0001F34C", "🍌")
 
 assert_throws("Invalid 16 bit universal character", fun(){ parse("\"\\uD83C\""); });
+assert_throws("Incomplete unicode escape sequence", fun(){ parse("\"\\uD83\""); });
 
-assert_equal("\U24B62", "𤭢")
+assert_equal("\U00024B62", "𤭢")
 
 assert_equal("Test\U0001F534Me", "Test🔴Me")