update code

json_tokener.c

@@ -223,7 +223,7 @@ struct json_object* json_tokener_parse_verbose(const char *str,
     :						\
     (((tok)->err = json_tokener_continue), 0)	\
     ) :						\
-   (((tok->flags & JSON_TOKENER_STRICT) &&   \
+   (((tok->flags & JSON_TOKENER_VALIDATE_UTF8) &&   \
     (!json_tokener_validate_utf8(*str, nBytesp)))?  \
     ((tok->err = json_tokener_error_parse_utf8_string), 0)  \
     :            \
@@ -956,7 +956,7 @@ struct json_object* json_tokener_parse_ex(struct json_tokener *tok,
   } /* while(PEEK_CHAR) */
 
  out:
-  if ((tok->flags & JSON_TOKENER_STRICT) && (nBytes != 0))
+  if ((tok->flags & JSON_TOKENER_VALIDATE_UTF8) && (nBytes != 0))
   {
     tok->err = json_tokener_error_parse_utf8_string;
   }
@@ -1004,19 +1004,14 @@ json_bool json_tokener_validate_utf8(const char c, unsigned int *nBytes)
   {
     if (chr >= 0x80)
     {
-      if(chr >= 0xFC && chr <= 0xFd)
-        *nBytes = 6;
-      else if (chr >= 0xF8)
-        *nBytes = 5;
-      else if (chr >= 0xF0)
-        *nBytes = 4;
-      else if (chr >= 0xE0)
-        *nBytes = 3;
-      else if (chr >= 0xC0)
+      if ((chr & 0xe0) == 0xc0)
+        *nBytes = 1;
+      else if ((chr & 0xf0) == 0xe0)
         *nBytes = 2;
+      else if ((chr & 0xf8) == 0xf0)
+        *nBytes = 3;
       else
         return 0;
-      (*nBytes)--;
     }
   }
   else

json_tokener.h

@@ -137,6 +137,17 @@ typedef struct json_tokener json_tokener;
  */
 #define JSON_TOKENER_STRICT  0x01
 
+/**
+ * Allow json_tokener_parse_ex() validate utf-8 char.
+ * The json_tokener_validate_utf8() validate one utf8 char
+ * after get one char, then begin to parse it.
+ *
+ * This flag is not set by default.
+ *
+ * @see json_tokener_set_flags()
+ */
+#define JSON_TOKENER_VALIDATE_UTF8  0x10
+
 /**
  * Given an error previously returned by json_tokener_get_error(),
  * return a human readable description of the error.

tests/test_parse.c

@@ -357,38 +357,36 @@ struct incremental_step {
 
   // utf-8 test
   // acsll encoding
-	{ "\x22\x31\x32\x33\x61\x73\x63\x24\x25\x26\x22",-1, -1, json_tokener_success, 3 },
+	{ "\x22\x31\x32\x33\x61\x73\x63\x24\x25\x26\x22",-1, -1, json_tokener_success, 5 },
 	{ "\x22\x31\x32\x33\x61\x73\x63\x24\x25\x26\x22",-1, -1, json_tokener_success, 1 },
   // utf-8 encoding
-	{ "\x22\xe4\xb8\x96\xe7\x95\x8c\x22",-1, -1, json_tokener_success, 3 },
-	{ "\x22\xe4\xb8",-1, -1, json_tokener_error_parse_utf8_string, 2 },
-	{ "\x96\xe7\x95\x8c\x22",-1, 0, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x22\xe4\xb8\x96\xe7\x95\x8c\x22",-1, -1, json_tokener_success, 5 },
+	{ "\x22\xe4\xb8",-1, 3, json_tokener_error_parse_utf8_string, 4 },
+	{ "\x96\xe7\x95\x8c\x22",-1, 0, json_tokener_error_parse_utf8_string, 5 },
 	{ "\x22\xe4\xb8\x96\xe7\x95\x8c\x22",-1, -1, json_tokener_success, 1 },
-	{ "\x22\xcf\x80\xcf\x86\x22",-1, -1, json_tokener_success, 3 },
-	{ "\x22\xf0\xa5\x91\x95\x22",-1, -1, json_tokener_success, 3 },
-	{ "\x22\xf8\xa5\xa5\x91\x95\x22",-1, -1, json_tokener_success, 3 },
-	{ "\x22\xfd\xa5\xa5\xa5\x91\x95\x22",-1, -1, json_tokener_success, 3 },
+	{ "\x22\xcf\x80\xcf\x86\x22",-1, -1, json_tokener_success, 5 },
+	{ "\x22\xf0\xa5\x91\x95\x22",-1, -1, json_tokener_success, 5 },
   // wrong utf-8 encoding
-	{ "\x22\xe6\x9d\x4e\x22",-1, 3, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x22\xe6\x9d\x4e\x22",-1, 3, json_tokener_error_parse_utf8_string, 5 },
 	{ "\x22\xe6\x9d\x4e\x22",-1, 5, json_tokener_success, 1 },
   // GBK encoding
-	{ "\x22\xc0\xee\xc5\xf4\x22",-1, 2, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x22\xc0\xee\xc5\xf4\x22",-1, 2, json_tokener_error_parse_utf8_string, 5 },
 	{ "\x22\xc0\xee\xc5\xf4\x22",-1, 6, json_tokener_success, 1 },
   // char after space
-	{ "\x20\x20\x22\xe4\xb8\x96\x22",-1, -1, json_tokener_success, 3 },
-	{ "\x20\x20\x81\x22\xe4\xb8\x96\x22",-1, 2, json_tokener_error_parse_utf8_string, 3 },
-	{ "\x5b\x20\x81\x31\x5d",-1, 2, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x20\x20\x22\xe4\xb8\x96\x22",-1, -1, json_tokener_success, 5 },
+	{ "\x20\x20\x81\x22\xe4\xb8\x96\x22",-1, 2, json_tokener_error_parse_utf8_string, 5 },
+	{ "\x5b\x20\x81\x31\x5d",-1, 2, json_tokener_error_parse_utf8_string, 5 },
   // char in state inf
 	{ "\x49\x6e\x66\x69\x6e\x69\x74\x79",9, 8, json_tokener_success, 1 },
-	{ "\x49\x6e\x66\x81\x6e\x69\x74\x79",-1, 3, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x49\x6e\x66\x81\x6e\x69\x74\x79",-1, 3, json_tokener_error_parse_utf8_string, 5 },
   // char in escape unicode
-	{ "\x22\x5c\x75\x64\x38\x35\x35\x5c\x75\x64\x63\x35\x35\x22",15, 14, json_tokener_success, 3 },
-	{ "\x22\x5c\x75\x64\x38\x35\x35\xc0\x75\x64\x63\x35\x35\x22",-1, 8, json_tokener_error_parse_utf8_string, 3 },
-	{ "\x22\x5c\x75\x64\x30\x30\x33\x31\xc0\x22",-1, 9, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x22\x5c\x75\x64\x38\x35\x35\x5c\x75\x64\x63\x35\x35\x22",15, 14, json_tokener_success, 5 },
+	{ "\x22\x5c\x75\x64\x38\x35\x35\xc0\x75\x64\x63\x35\x35\x22",-1, 8, json_tokener_error_parse_utf8_string, 5 },
+	{ "\x22\x5c\x75\x64\x30\x30\x33\x31\xc0\x22",-1, 9, json_tokener_error_parse_utf8_string, 5 },
   // char in number
-	{ "\x31\x31\x81\x31\x31",-1, 2, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x31\x31\x81\x31\x31",-1, 2, json_tokener_error_parse_utf8_string, 5 },
   // char in object
-	{ "\x7b\x22\x31\x81\x22\x3a\x31\x7d",-1, 3, json_tokener_error_parse_utf8_string, 3 },
+	{ "\x7b\x22\x31\x81\x22\x3a\x31\x7d",-1, 3, json_tokener_error_parse_utf8_string, 5 },
 
 	{ NULL, -1, -1, json_tokener_success, 0 },
 };
@@ -424,9 +422,19 @@ static void test_incremental_parse()
 		size_t expected_char_offset;
 
 		if (step->reset_tokener & 2)
-			json_tokener_set_flags(tok, JSON_TOKENER_STRICT);
+			{
+				if (step->reset_tokener & 4)
+					json_tokener_set_flags(tok, 3);
+				else
+					json_tokener_set_flags(tok, JSON_TOKENER_STRICT);
+			}
 		else
-			json_tokener_set_flags(tok, 0);
+			{
+				if (step->reset_tokener & 4)
+					json_tokener_set_flags(tok, JSON_TOKENER_VALIDATE_UTF8);
+				else
+					json_tokener_set_flags(tok, 0);
+			}
 
 		if (length == -1)
 			length = strlen(step->string_to_parse);

tests/test_parse.expected

@@ -191,8 +191,6 @@ json_tokener_parse_ex(tok,
 json_tokener_parse_ex(tok, "世界"    ,   8) ... OK: got object of type [string]: "世界"
 json_tokener_parse_ex(tok, "πφ"      ,   6) ... OK: got object of type [string]: "πφ"
 json_tokener_parse_ex(tok, "𥑕"      ,   6) ... OK: got object of type [string]: "𥑕"
-json_tokener_parse_ex(tok, "ø¥¥‘•"     ,   7) ... OK: got object of type [string]: "ø¥¥‘•"
-json_tokener_parse_ex(tok, "ý¥¥¥‘•"    ,   8) ... OK: got object of type [string]: "ý¥¥¥‘•"
 json_tokener_parse_ex(tok, "æ�N"       ,   5) ... OK: got correct error: invalid utf-8 string
 json_tokener_parse_ex(tok, "æ�N"       ,   5) ... OK: got object of type [string]: "æ�N"
 json_tokener_parse_ex(tok, "ÀîÅô"      ,   6) ... OK: got correct error: invalid utf-8 string
@@ -207,5 +205,5 @@ json_tokener_parse_ex(tok, "\ud855
 json_tokener_parse_ex(tok, "\ud0031À"  ,  10) ... OK: got correct error: invalid utf-8 string
 json_tokener_parse_ex(tok, 11�11       ,   5) ... OK: got correct error: invalid utf-8 string
 json_tokener_parse_ex(tok, {"1�":1}    ,   8) ... OK: got correct error: invalid utf-8 string
-End Incremental Tests OK=129 ERROR=0
+End Incremental Tests OK=127 ERROR=0
 ==================================