Realization of JSON protocol in mql4 / mql5. You can create JSON object with data of different types and run serialization and deserialization of JSON data.
This repo is fork of JAson (https://www.mql5.com/en/code/13663) with improvements, refactoring code, unit tests and translating of comments to English.
Download repo and copy JAson/Include/JAson.mqh folder to <TERMINAL DIR>/MQL(4/5)/Include
Add #include <JAson.mqh> and create CJAVal object to work with JSON data. See simple example for easy understand:
#include <JAson.mqh>
int OnInit(){
CJAVal data;
// --- simple structure ---
data["a"] = 12;
data["b"] = 3.14;
data["c"] = "foo";
data["d"] = true;
Print(data["a"].ToInt()); // 12
Print(data["b"].ToDbl()); // 3.14
Print(data["c"].ToStr()); // "foo"
Print(data["d"].ToBool()); // true
// --- array structure ---
data["e"].Add("bar");
data["e"].Add(2);
Print(data["e"][0].ToStr()); // "bar"
Print(data["e"][1].ToInt()); // 2
// --- nested structures ---
// - as part of array -
CJAVal sub_data_obj_in_list;
sub_data_obj_in_list["k1"] = 7;
sub_data_obj_in_list["k2"] = "baz";
data["e"].Add(sub_data_obj_in_list);
Print(data["e"][2]["k1"].ToInt()); // 7
Print(data["e"][2]["k2"].ToStr()); // "baz"
// - as object -
CJAVal sub_data_obj;
sub_data_obj["sub_c"] = "muz2";
sub_data_obj["sub_d"] = 44;
data["f"] = sub_data_obj;
Print(data["f"]["sub_c"].ToStr()); // "muz2"
Print(data["f"]["sub_d"].ToInt()); // 44
}To get value from Json object, you need use methods:
ToInt()forintegertype;ToDbl()fordoubletype;ToStr()forstringtype;ToBool()forbooleantype;
For example:
data["a"].ToInt(); // 12
data["b"].ToDbl(); // 3.14
data["c"].ToStr(); // "foo"
data["d"].ToBool(); // trueTo create array, use .Add()
data["e"].Add("bar");
data["e"].Add(2);To use nested json, create other CJAVal object and assign to existed CJAVal object. Nested CJAVal object can be value of key or contained in array:
CJAVal data;
// - adding as object -
CJAVal data_1;
data_1["d1_1"] = 7;
data_1["d1_2"] = "foo";
data["a"] = data_1;
// - adding as part of array -
CJAVal data_2;
data_2["d2_1"] = 1;
data_2["d2_1"] = "bar";
data["b"].Add("buz");
data["b"].Add(data_2);
data["b"].Add("muz");JAson provides the serialization and deserialization:
#include <JAson.mqh>
int OnInit(){
string data_str;
CJAVal data;
data["a"] = 3.14;
data["b"] = "foo";
data["c"].Add("bar");
data["c"].Add(2);
data["c"].Add("baz");
Print(data["b"].ToStr()); // foo
data_str = data.Serialize();
Print(data_str); // {"a":3.14000000,"b":"foo","c":["bar",2,"baz"]}
CJAVal data_2;
data_2.Deserialize(data_str);
Print(data_2["a"].ToDbl()); // 3.14
Print(data_2["b"].ToStr()); // foo
Print(data_2["c"][0].ToStr()); // bar
Print(data_2["c"][1].ToInt()); // 2
Print(data_2["c"][2].ToStr()); // baz
}It can be useful for data saving in a file or send Json object by POST request.
POST request with Json object:
#include <JAson.mqh>
#include <requests/requests.mqh> // https://github.com/vivazzi/mql_requests
int OnInit(){
CJAVal data;
data["a"] = 7;
data["b"] = "foo";
Requests requests;
Response response = requests.post("https://vivazzi.pro/test-request/", data.Serialize());
Print(response.text); // {"status": "OK", "method": "POST", "body": "{"a": 7, "b": "foo"}"}
data.Deserialize(response.text);
Print(data["status"].ToStr()); // "OK"
Print(data["method"].ToStr()); // "POST"
Print(data["body"].ToStr()); // {"a": 7, "b": "foo"}
// if Json object has value of key as serialized object, you can also deserialize this value
CJAVal data_2;
data_2.Deserialize(data["body"].ToStr());
Print(data_2["a"].ToInt()); // 7
Print(data_2["b"].ToStr()); // "foo"
// also you can join this Json objects to get full structure
data["body"] = data_2;
Print(data["body"]["a"].ToInt());
Print(data["body"]["b"].ToStr());
}In this example it has been used: mql_requests and online service Getest for testing GET and POST requests.
To clear CJAVal object, use method Clear():
#include <JAson.mqh>
int OnInit(){
CJAVal data;
string data_str;
data["a"] = 3.14;
data["b"] = "foo";
data_str = data.Serialize();
Print(data_str); // "{"a":3.14000000,"b":"foo"}"
data.Clear();
data["c"] = 123;
data["d"] = "bar";
data_str = data.Serialize();
Print(data_str); // "{"c":123,"d":"bar"}"
}So if you want to get values of "c" and "d" keys, you need:
Print(data["c"].ToInt()) // 123
Print(data["d"].ToStr()) // "bar"But if you check "a" and "b" keys, that no longer exist (or never used keys), you get initial values for key:
// old notexistent keys
Print(data["a"].ToDbl()) // 0.0
Print(data["b"].ToStr()) // "" (empty)
// never used keys
Print(data["e"].ToStr()) // "" (empty)And this can lead to logical errors - be careful! Always use only a specific set of keys to avoid logical error!
If you use Json, that can be cleaned up, and you use different keys, you can check type of keys. JAson library define next types:
jtUNDEF, if key is undefinedjtNULLforNULLjtBOOLforbooleanjtINTforintjtDBLfordoublejtSTRforstringjtARRAYforarrayjtOBforobj
Then in example above:
// old notexistent keys
Print(data["a"].ToDbl()); // 0.0
Print(data["a"].type); // jtUNDEF
Print(data["b"].ToStr()); // ""
Print(data["b"].type); // jtUNDEF
// current keys
Print(data["c"].ToInt()); // 123
Print(data["c"].type); // jtINT
Print(data["d"].ToStr()); // "bar"
Print(data["d"].type); // jtSTR
// never used keys
Print(data["e"].ToStr()); // ""
Print(data["e"].type); // jtUNDEFSo you can compare key type with jtUNDEF, if you want to check for the existence of a key:
#include <JAson.mqh>
int OnInit(){
CJAVal data;
data["a"] = 3.14;
data.Clear();
data["c"] = 123;
if (data["a"].type == jtUNDEF) {
// do something
} else {
// else do something
}
}CJAVal data;- CreatesCJAVal(Json) object.CJAVal data(CJAVal* a_parent, enJAType a_type);- CreatesCJAValobject with specified type.a_parent- parent for createdCJAValobject (useNULLif no parent).a_type- type ofCJAValobject, available types:jtUNDEF,jtNULL,jtBOOL,jtINT,jtDBL,jtSTR,jtARRAY,jtOBJ.CJAVal data(const double v, int precision=-100);- CreatesCJAValobject of double type of specified precision.
data[key] = some_val;- Addssome_val(int, double, string or anotherCJAVal) to data withkeykey.data[key].Add(other_data);- Addsother_data(int, double, string or other CJAVal) tokeyarray.data[key].Add(const double v, int precision=-2);- Addsvof double type with specified precision tokeyarray.
data.Serialize();- ConvertCJAValobject to string.data.Deserialize(string json, int acp=CP_ACP);- Convertjsonstring toCJAValobject.datagets result.acp- code page for deserialization.
data.Clear(enJAType jt=jtUNDEF, bool save_key=false);- ClearsCJAValobject.jt- sets specified type ofCJAValobject.save_key- iftrue, values of all keys will be keys will be cleared, else all keys and values will be cleared.data.Size();- Gets size ofCJAValobject.
- If the
dbl_precisionvalue is in the range from 0 to 16, then a string representation of the number with the specified number of decimal places will be obtained. - If the
dbl_precisionvalue is in the range from -1 to -16, then the string representation of the number in scientific format with the specified number of decimal places will be obtained. - In all other cases, the string representation of the number will contain 8 decimal places.
- Copy
JAson/Scripts/folder to<TERMINAL DIR>/MQL(4/5)/Scripts/ - Compile
TestJAson.mq4and runTestJAson.ex4in terminal in a window of any trading pair. - Look test result in
<TERMINAL DIR>/MQL(4/5)/Files/TestJAson_unit_test_log.txt
To reporting bugs or suggest improvements, please use the issue tracker.
Thank you very much, that you would like to contribute to JAson. Thanks to the present, past and future contributors.
If you think you have discovered a security issue in code, please do not create issue or raise it in any public forum until we have had a chance to deal with it. For security issues use [email protected]
- Report bugs and suggest improvements:
- https://github.com/vivazzi/JAson/issues (strongly recommended)
- https://www.mql5.com/en/code/13663 (not tracked by the repo author)
- Author's site of this repo, Artem Maltsev: https://vivazzi.pro
Copyright © 2021 Alexey Sergeev, Artem Maltsev and contributors.
MIT licensed.