Parsing JSON using Json.net
I'm trying to parse some JSON using the JSon.Net library. The documentation seems a little sparse and I'm confused as to how to accomplish what I need. Here is the format for the JSON I need to parse through.
{
"displayFieldName" : "OBJECT_NAME",
"fieldAliases" : {
"OBJECT_NAME" : "OBJECT_NAME",
"OBJECT_TYPE" : "OBJECT_TYPE"
},
"positionType" : "point",
"reference" : {
"id" : 1111
},
"objects" : [ {
"attributes" : {
"OBJECT_NAME" : "test name",
"OBJECT_TYPE" : "test type"
},
"position" : {
"x" : 5,
"y" : 7
}
} ]
}
The only data I really need from this is the stuff in the objects array. Is it possible for me to parse through that with something like the JSonTextReader and just pull out the things I want, like OBJECT_TYPE and the x and y position? I can't seem to get JSonTextReader to work the way I want it to and I find little to no examples of usage for it.
It seems like serializing first then using LINQ with my object would be ideal and every example I find discusses serializing the JSON first, but I'm not sure how I would build an object for this structure. Particularly the objects array which would need to be something like a list of Pairs of attribute and position objects. I have no idea how I would code my object so JSon.Net would know how to serialize that.
I thought I could write my own simple parser to just pull out everything I need into an attributes object that I created, but I'm having little luck.
Hopefully this all makes sense, any ideas?
Solution 1:
I don't know about JSON.NET, but it works fine with JavaScriptSerializer from System.Web.Extensions.dll (.NET 3.5 SP1):
using System.Collections.Generic;
using System.Web.Script.Serialization;
public class NameTypePair
{
public string OBJECT_NAME { get; set; }
public string OBJECT_TYPE { get; set; }
}
public enum PositionType { none, point }
public class Ref
{
public int id { get; set; }
}
public class SubObject
{
public NameTypePair attributes { get; set; }
public Position position { get; set; }
}
public class Position
{
public int x { get; set; }
public int y { get; set; }
}
public class Foo
{
public Foo() { objects = new List<SubObject>(); }
public string displayFieldName { get; set; }
public NameTypePair fieldAliases { get; set; }
public PositionType positionType { get; set; }
public Ref reference { get; set; }
public List<SubObject> objects { get; set; }
}
static class Program
{
const string json = @"{
""displayFieldName"" : ""OBJECT_NAME"",
""fieldAliases"" : {
""OBJECT_NAME"" : ""OBJECT_NAME"",
""OBJECT_TYPE"" : ""OBJECT_TYPE""
},
""positionType"" : ""point"",
""reference"" : {
""id"" : 1111
},
""objects"" : [
{
""attributes"" : {
""OBJECT_NAME"" : ""test name"",
""OBJECT_TYPE"" : ""test type""
},
""position"" :
{
""x"" : 5,
""y"" : 7
}
}
]
}";
static void Main()
{
JavaScriptSerializer ser = new JavaScriptSerializer();
Foo foo = ser.Deserialize<Foo>(json);
}
}
Edit:
Json.NET works using the same JSON and classes.
Foo foo = JsonConvert.DeserializeObject<Foo>(json);
Link: Serializing and Deserializing JSON with Json.NET
Solution 2:
Edit: Thanks Marc, read up on the struct vs class issue and you're right, thank you!
I tend to use the following method for doing what you describe, using a static method of JSon.Net:
MyObject deserializedObject = JsonConvert.DeserializeObject<MyObject>(json);
Link: Serializing and Deserializing JSON with Json.NET
For the Objects list, may I suggest using generic lists out made out of your own small class containing attributes and position class. You can use the Point struct in System.Drawing (System.Drawing.Point or System.Drawing.PointF for floating point numbers) for you X and Y.
After object creation it's much easier to get the data you're after vs. the text parsing you're otherwise looking at.
Solution 3:
(This question came up high on a search engine result, but I ended up using a different approach. Adding an answer to this old question in case other people with similar questions read this)
You can solve this with Json.Net and make an extension method to handle the items you want to loop:
public static Tuple<string, int, int> ToTuple(this JToken token)
{
var type = token["attributes"]["OBJECT_TYPE"].ToString();
var x = token["position"]["x"].Value<int>();
var y = token["position"]["y"].Value<int>();
return new Tuple<string, int, int>(type, x, y);
}
And then access the data like this: (scenario: writing to console):
var tuples = JObject.Parse(myJsonString)["objects"].Select(item => item.ToTuple()).ToList();
tuples.ForEach(t => Console.WriteLine("{0}: ({1},{2})", t.Item1, t.Item2, t.Item3));
Solution 4:
/*
* This method takes in JSON in the form returned by javascript's
* JSON.stringify(Object) and returns a string->string dictionary.
* This method may be of use when the format of the json is unknown.
* You can modify the delimiters, etc pretty easily in the source
* (sorry I didn't abstract it--I have a very specific use).
*/
public static Dictionary<string, string> jsonParse(string rawjson)
{
Dictionary<string, string> outdict = new Dictionary<string, string>();
StringBuilder keybufferbuilder = new StringBuilder();
StringBuilder valuebufferbuilder = new StringBuilder();
StringReader bufferreader = new StringReader(rawjson);
int s = 0;
bool reading = false;
bool inside_string = false;
bool reading_value = false;
//break at end (returns -1)
while (s >= 0)
{
s = bufferreader.Read();
//opening of json
if (!reading)
{
if ((char)s == '{' && !inside_string && !reading) reading = true;
continue;
}
else
{
//if we find a quote and we are not yet inside a string, advance and get inside
if (!inside_string)
{
//read past the quote
if ((char)s == '\"') inside_string = true;
continue;
}
if (inside_string)
{
//if we reached the end of the string
if ((char)s == '\"')
{
inside_string = false;
s = bufferreader.Read(); //advance pointer
if ((char)s == ':')
{
reading_value = true;
continue;
}
if (reading_value && (char)s == ',')
{
//we know we just ended the line, so put itin our dictionary
if (!outdict.ContainsKey(keybufferbuilder.ToString())) outdict.Add(keybufferbuilder.ToString(), valuebufferbuilder.ToString());
//and clear the buffers
keybufferbuilder.Clear();
valuebufferbuilder.Clear();
reading_value = false;
}
if (reading_value && (char)s == '}')
{
//we know we just ended the line, so put itin our dictionary
if (!outdict.ContainsKey(keybufferbuilder.ToString())) outdict.Add(keybufferbuilder.ToString(), valuebufferbuilder.ToString());
//and clear the buffers
keybufferbuilder.Clear();
valuebufferbuilder.Clear();
reading_value = false;
reading = false;
break;
}
}
else
{
if (reading_value)
{
valuebufferbuilder.Append((char)s);
continue;
}
else
{
keybufferbuilder.Append((char)s);
continue;
}
}
}
else
{
switch ((char)s)
{
case ':':
reading_value = true;
break;
default:
if (reading_value)
{
valuebufferbuilder.Append((char)s);
}
else
{
keybufferbuilder.Append((char)s);
}
break;
}
}
}
}
return outdict;
}