How to apply a general rule for remapping all property names when serializing with Json.NET?

Question

When deserializing a Json object into a .Net type, if the field names don't match up I found you can decorate your type's properties with [JsonProperty(PropertyName = "name")]

This is fine and dandy for a couple of properties that don't match up, but is there a way to set a convention or rule?

Json

{
  "Job": [
    {
      "Job #": "1",
      "Job Type": "A",
    }
  ]
}

C#

    [JsonProperty(PropertyName = "Job Type")]
    public string JobType { get; set; }

    [JsonProperty(PropertyName = "Job #")]
    public string JobNumber { get; set; }

I have many fields using similar names, what I would like to figure out, is there a way to tell to set a rule to always remove spaces (EG: Job Type -> JobType) and replace # with Number (eg: Job # -> JobNumber)?

It looks like a custom ContractResolver might be the only solution, but I can't seem to figure out how to go about using it to pluck out spaces and replace "#" with "Number". Does anyone have a reference example?

Or, I'm hoping there's a nice simple solution that I've overlooked.

P.S. Also accepting suggestions for a better title.

Answer 1

Assuming you are working with Json.NET 9.0.1 or later, this can be done with a custom NamingStrategy. For instance, here's one based on SnakeCaseNamingStrategy and StringUtils.ToSnakeCase() by James Newton-King:

public class CustomNamingStrategy : NamingStrategy
{
    public CustomNamingStrategy(bool processDictionaryKeys, bool overrideSpecifiedNames)
    {
        ProcessDictionaryKeys = processDictionaryKeys;
        OverrideSpecifiedNames = overrideSpecifiedNames;
    }

    public CustomNamingStrategy(bool processDictionaryKeys, bool overrideSpecifiedNames, bool processExtensionDataNames)
        : this(processDictionaryKeys, overrideSpecifiedNames)
    {
        ProcessExtensionDataNames = processExtensionDataNames;
    }

    public CustomNamingStrategy()
    {
    }

    protected override string ResolvePropertyName(string name)
    {
        return SpaceWords(name);
    }

    enum WordState
    {
        Start,
        Lower,
        Upper,
        NewWord
    }

    static string SpaceWords(string s)
    {
        // Adapted from StringUtils.ToSnakeCase()
        // https://github.com/JamesNK/Newtonsoft.Json/blob/master/Src/Newtonsoft.Json/Utilities/StringUtils.cs#L191
        // 
        // Copyright (c) 2007 James Newton-King
        //
        // Permission is hereby granted, free of charge, to any person
        // obtaining a copy of this software and associated documentation
        // files (the "Software"), to deal in the Software without
        // restriction, including without limitation the rights to use,
        // copy, modify, merge, publish, distribute, sublicense, and/or sell
        // copies of the Software, and to permit persons to whom the
        // Software is furnished to do so, subject to the following
        // conditions:
        //
        // The above copyright notice and this permission notice shall be
        // included in all copies or substantial portions of the Software.
        //
        // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
        // EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
        // OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
        // NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
        // HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
        // WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
        // FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
        // OTHER DEALINGS IN THE SOFTWARE.

        char wordBreakChar = ' ';

        if (string.IsNullOrEmpty(s))
        {
            return s;
        }

        StringBuilder sb = new StringBuilder();
        WordState state = WordState.Start;

        for (int i = 0; i < s.Length; i++)
        {
            if (s[i] == ' ')
            {
                if (state != WordState.Start)
                {
                    state = WordState.NewWord;
                }
            }
            else if (char.IsUpper(s[i]))
            {
                switch (state)
                {
                    case WordState.Upper:
                        bool hasNext = (i + 1 < s.Length);
                        if (i > 0 && hasNext)
                        {
                            char nextChar = s[i + 1];
                            if (!char.IsUpper(nextChar) && nextChar != ' ')
                            {
                                sb.Append(wordBreakChar);
                            }
                        }
                        break;
                    case WordState.Lower:
                    case WordState.NewWord:
                        sb.Append(wordBreakChar);
                        break;
                }

                sb.Append(s[i]);

                state = WordState.Upper;
            }
            else if (s[i] == wordBreakChar)
            {
                sb.Append(wordBreakChar);
                state = WordState.Start;
            }
            else
            {
                if (state == WordState.NewWord)
                {
                    sb.Append(wordBreakChar);
                }

                sb.Append(s[i]);
                state = WordState.Lower;
            }
        }

        sb.Replace("Number", "#");
        return sb.ToString();
    }
}

Then you can apply it to your type as follows:

[JsonObject(NamingStrategyType = typeof(CustomNamingStrategy))]
public class RootObject
{
    public string JobType { get; set; }

    public string JobNumber { get; set; }

    public int JobItemCount { get; set; }

    public string ISOCode { get; set; }

    public string SourceXML { get; set; }
}

And the JSON generated will be as follows:

{
  "Job Type": "job type",
  "Job #": "01010101",
  "Job Item Count": 3,
  "ISO Code": "ISO 9000",
  "Source XML": "c:\temp.xml"
}

Notes:

• If you want the strategy to apply to properties that already have property names specified via JsonPropertyAttribute.PropertyName, set NamingStrategy.OverrideSpecifiedNames == true.

• To apply your naming strategy to all types rather than setting it on each object, you can set the naming strategy in DefaultContractResolver.NamingStrategy, then set the contract resolver in JsonSerializerSettings.ContractResolver.

• The naming strategy maps from the c# property name to the JSON property name, not vice versa. Thus you need to insert spaces rather than "pluck them out" and replace "Number" with "#". The mapping is then cached by the contract resolver and a reverse lookup is done during deserialization.

Answer 2

Yes a ContractResolver is the way to go.

The problem is that these only seem to work going from the destination property to the source, i.e. "JobType" -> "Job Type", not the other way as you would like. This makes the solution a bit more flaky than you might want.

First we make our ContractResolver, inheriting from DefaultContractResolver, so it all works as normal apart from the bit we want to customize:

public class JobContractResolver : DefaultContractResolver
{
    protected override string ResolvePropertyName(string propertyName)
    {
        // first replace all capital letters with space then letter ("A" => " A"). This might include the first letter, so trim the result.
        string result = Regex.Replace(propertyName, "[A-Z]", x => " " + x.Value).Trim();

        // now replace Number with a hash
        result = result.Replace("Number", "#");

        return result;
    }
}

Then in our deserialization, we set the ContractResolver in the JsonSerializerSettings:

static void Main(string[] args)
{
    string input = @"{""Job #"": ""1"", ""Job Type"": ""A""}";

    var job1 = JsonConvert.DeserializeObject<Job1>(input, new JsonSerializerSettings
    {
        ContractResolver = new JobContractResolver()
    });

    Console.WriteLine("JobType: {0}", job1.JobType);
    Console.WriteLine("JobNumber: {0}", job1.JobNumber);
}