Merge pull request #7293 from Vdimir/feature/isvalidjson-5910

Add function isValidJSON

dbms/src/Functions/FunctionsJSON.cpp

@@ -8,6 +8,7 @@ namespace DB
 void registerFunctionsJSON(FunctionFactory & factory)
 {
     factory.registerFunction<FunctionJSON<NameJSONHas, JSONHasImpl>>();
+    factory.registerFunction<FunctionJSON<NameIsValidJSON, IsValidJSONImpl>>();
     factory.registerFunction<FunctionJSON<NameJSONLength, JSONLengthImpl>>();
     factory.registerFunction<FunctionJSON<NameJSONKey, JSONKeyImpl>>();
     factory.registerFunction<FunctionJSON<NameJSONType, JSONTypeImpl>>();

dbms/src/Functions/FunctionsJSON.h

@@ -279,6 +279,7 @@ private:
 
 
 struct NameJSONHas { static constexpr auto name{"JSONHas"}; };
+struct NameIsValidJSON { static constexpr auto name{"isValidJSON"}; };
 struct NameJSONLength { static constexpr auto name{"JSONLength"}; };
 struct NameJSONKey { static constexpr auto name{"JSONKey"}; };
 struct NameJSONType { static constexpr auto name{"JSONType"}; };
@@ -311,6 +312,36 @@ public:
 };
 
 
+template <typename JSONParser>
+class IsValidJSONImpl
+{
+public:
+    static DataTypePtr getType(const char * function_name, const ColumnsWithTypeAndName & arguments)
+    {
+        if (arguments.size() != 1)
+        {
+            /// IsValidJSON() shouldn't get parameters other than JSON.
+            throw Exception{"Function " + String(function_name) + " needs exactly one argument",
+                            ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH};
+        }
+        return std::make_shared<DataTypeUInt8>();
+    }
+
+    using Iterator = typename JSONParser::Iterator;
+    static bool addValueToColumn(IColumn & dest, const Iterator &)
+    {
+        /// This function is called only if JSON is valid.
+        /// If JSON isn't valid then `FunctionJSON::Executor::run()` adds default value (=zero) to `dest` without calling this function.
+        ColumnVector<UInt8> & col_vec = assert_cast<ColumnVector<UInt8> &>(dest);
+        col_vec.insertValue(1);
+        return true;
+    }
+
+    static constexpr size_t num_extra_arguments = 0;
+    static void prepare(const char *, const Block &, const ColumnNumbers &, size_t) {}
+};
+
+
 template <typename JSONParser>
 class JSONLengthImpl
 {

dbms/tests/queries/0_stateless/00918_json_functions.reference

@@ -7,6 +7,10 @@
 1
 1
 0
+--isValidJSON--
+1
+0
+0
 --JSONKey--
 a
 b

dbms/tests/queries/0_stateless/00918_json_functions.sql

@@ -11,6 +11,11 @@ SELECT JSONHas('{"a": "hello", "b": [-100, 200.0, 300]}', 'a');
 SELECT JSONHas('{"a": "hello", "b": [-100, 200.0, 300]}', 'b');
 SELECT JSONHas('{"a": "hello", "b": [-100, 200.0, 300]}', 'c');
 
+SELECT '--isValidJSON--';
+SELECT isValidJSON('{"a": "hello", "b": [-100, 200.0, 300]}');
+SELECT isValidJSON('not a json');
+SELECT isValidJSON('"HX-=');
+
 SELECT '--JSONKey--';
 SELECT JSONKey('{"a": "hello", "b": [-100, 200.0, 300]}', 1);
 SELECT JSONKey('{"a": "hello", "b": [-100, 200.0, 300]}', 2);

docs/en/query_language/functions/json_functions.md

@@ -57,6 +57,17 @@ There is currently no support for code points in the format `\uXXXX\uYYYY` that
 
 The following functions are based on [simdjson](https://github.com/lemire/simdjson) designed for more complex JSON parsing requirements. The assumption 2 mentioned above still applies.
 
+## isValidJSON(json)
+
+Checks that passed string is a valid json.
+
+Examples:
+
+```sql
+SELECT isValidJSON('{"a": "hello", "b": [-100, 200.0, 300]}') = 1
+SELECT isValidJSON('not a json') = 0
+```
+
 ## JSONHas(json[, indices_or_keys]...)
 
 If the value exists in the JSON document, `1` will be returned.

docs/ru/query_language/functions/json_functions.md

@@ -57,6 +57,17 @@ visitParamExtractString('{"abc":"hello}', 'abc') = ''
 
 Следующие функции используют [simdjson](https://github.com/lemire/simdjson) который разработан по более сложны требования для разбора JSON. Упомянутое выше предположение 2 по-прежнему применимо.
 
+## isValidJSON(json)
+
+Проверяет, является ли переданная строка валидным json значением.
+
+Примеры:
+
+```sql
+SELECT isValidJSON('{"a": "hello", "b": [-100, 200.0, 300]}') = 1
+SELECT isValidJSON('not a json') = 0
+```
+
 ## JSONHas(json[, indices_or_keys]...)
 
 Если значение существует в документе JSON, то возвращается `1`.