Reduced Unicode conversion code size (-122 bytes on AVR)

extras/tests/JsonDeserializer/invalid_input.cpp

@@ -7,10 +7,9 @@
 #include <catch.hpp>
 
 TEST_CASE("Invalid JSON input") {
-  const char* testCases[] = {
-      "'\\u'",     "'\\u000g'",  "'\\u000'",  "'\\u000G'", "'\\ud83d\\ud83d'",
-      "'\\udda4'", "'\\ud83d_'", "'\\u000/'", "\\x1234",   "6a9",
-      "1,",        "2]",         "3}"};
+  const char* testCases[] = {"'\\u'",     "'\\u000g'", "'\\u000'", "'\\u000G'",
+                             "'\\u000/'", "\\x1234",   "6a9",      "1,",
+                             "2]",        "3}"};
   const size_t testCount = sizeof(testCases) / sizeof(testCases[0]);
 
   DynamicJsonDocument doc(4096);
@@ -23,7 +22,14 @@ TEST_CASE("Invalid JSON input") {
 }
 
 TEST_CASE("Invalid JSON input that should pass") {
-  const char* testCases[] = {"nulL", "tru3", "fals3"};
+  const char* testCases[] = {
+      "nulL",
+      "tru3",
+      "fals3",
+      "'\\ud83d'",         // leading surrogate without a trailing surrogate
+      "'\\udda4'",         // trailing surrogate without a leading surrogate
+      "'\\ud83d\\ud83d'",  // two leading surrogates
+  };
   const size_t testCount = sizeof(testCases) / sizeof(testCases[0]);
 
   DynamicJsonDocument doc(4096);

extras/tests/Misc/CMakeLists.txt

@@ -10,6 +10,7 @@ add_executable(MiscTests
 	StringWriter.cpp
 	TypeTraits.cpp
 	unsigned_char.cpp
+	Utf8.cpp
 	version.cpp
 )
 

extras/tests/Misc/Utf8.cpp

@@ -0,0 +1,59 @@
+// ArduinoJson - arduinojson.org
+// Copyright Benoit Blanchon 2014-2019
+// MIT License
+
+#include <ArduinoJson.h>
+#include <catch.hpp>
+
+#include <string>
+
+using namespace ARDUINOJSON_NAMESPACE;
+
+static void testCodepoint(uint32_t codepoint, std::string expected) {
+  char buffer[4096];
+  MemoryPool pool(buffer, 4096);
+  StringBuilder str(&pool);
+
+  CAPTURE(codepoint);
+  Utf8::encodeCodepoint(codepoint, str);
+
+  REQUIRE(str.complete() == expected);
+}
+
+TEST_CASE("Utf8::encodeCodepoint()") {
+  SECTION("U+0000") {
+    testCodepoint(0x0000, "");
+  }
+
+  SECTION("U+0001") {
+    testCodepoint(0x0001, "\x01");
+  }
+
+  SECTION("U+007F") {
+    testCodepoint(0x007F, "\x7f");
+  }
+
+  SECTION("U+0080") {
+    testCodepoint(0x0080, "\xc2\x80");
+  }
+
+  SECTION("U+07FF") {
+    testCodepoint(0x07FF, "\xdf\xbf");
+  }
+
+  SECTION("U+0800") {
+    testCodepoint(0x0800, "\xe0\xa0\x80");
+  }
+
+  SECTION("U+FFFF") {
+    testCodepoint(0xFFFF, "\xef\xbf\xbf");
+  }
+
+  SECTION("U+10000") {
+    testCodepoint(0x10000, "\xf0\x90\x80\x80");
+  }
+
+  SECTION("U+10FFFF") {
+    testCodepoint(0x10FFFF, "\xf4\x8f\xbf\xbf");
+  }
+}