Data seed through JSON files using EF or any ORM

Last time I had demonstrated a concept for Data Seed through any ORM or MongoDB. The idea is to feed the data to DB at first run or based on condition. The advantage is no one has to manage DB manually with SQL script which feels more natural with Code First approach and every developer can have their own copy with this approach.

I am going to extend the same idea by importing data from JSON file. There are few reasons to do through JSON file as we do not want to keep increasing our codes for data import which may end up with  Combined length of user strings used by the program exceeds allowed limit. Try to decrease use of string literals. Also in the case of a huge number of records it does not make sense and with newer development practice JSON feels more natural for data.

First please go to Data seed for the application with EF, MongoDB or any other ORM as I would be extending same items to allow it through JSON file.

It is always a great idea to have a base class for common items. I would be just creating a new function under BaseSeed base class to import JSON file.

It is better to read any file through streams to avoid any memory clog. So, JsonTextReader would be used through Newtonsoft.Json package.

The whole class definition is this.

  public abstract class BaseSeed<TModel>  
     where TModel : class  
   {  
     protected readonly INoSqlRepoManager NoSqlRepositoryManager;  

     protected readonly IUnitOfWork UnitOfWork;  
     protected readonly IRepository<TModel> ModelRepository;  
    
     public BaseSeed(INoSqlRepoManager noSqlRepoManager)  
     {  
       NoSqlRepoManager = noSqlRepoManager;  
     }  
     
     public BaseSeed(IUnitOfWork unitOfWork)  
     {  
       UnitOfWork = unitOfWork;  
     }  
    
     public BaseSeed(IRepository<TModel> modelRepository)  
     {  
       ModelRepository = modelRepository;  
     }  
    
     protected void Save(TModel model, IRepository<TModel> repository)  
     {  
       repository.Create(model);  
       UnitOfWork.Save();  
     }  
       
     public IEnumerable<T> JsonRead<T>(string file)  
     {  
       if (File.Exists(file)) // No JSON Schema check done.  
       {  
         using (var fs = new FileStream(file, FileMode.Open, FileAccess.Read))  
         using (var sr = new StreamReader(fs))  
         using (var reader = new JsonTextReader(sr))  
         {  
           var jsonSerializer = JsonSerializer.Create();  
           while (reader.Read())  
           {  
             if (reader.TokenType == JsonToken.StartObject)  
             {  
               yield return jsonSerializer.Deserialize<T>(reader);  
             }  
           }  
         }  
       }  
     }  
   }  

JsonRead is a function to read a file as enumerable through yield to avoid memory clog.


Mainly this is the only thing we need to read JSON file and import to DB. The few minor change need to be done in SeedDatabase function to pass the application root path to access file.

JSON shall start as an array with objects, an example from CoreLookup model:

 [  
   {  
    "LookupType":1,  
    "LookupValue":"ALABAMA",  
    "LookupIdString":"AL"  
   },  
   {  
    "LookupType":1,  
    "LookupValue":"Alaska",  
    "LookupIdString":"AK"  
   },  
   {  
    "LookupType":1,  
    "LookupValue":"AMERICAN SAMOA",  
    "LookupIdString":"AS"  
   },  
 ]  

Import of data codes

   
  public class LookupSeeder   
    : BaseSeed<CoreLookup>, ISeed   
  {   
   
    public CoreLookupSeederSql(IMyProjectUnitOfWork MyProjectUnitOfWork)   
       : base(MyProjectUnitOfWork)   
    {   
    }   
   
    public void Seed()   
    {   
       if (!MyProjectUnitOfWork.LookupRepository.Any())  
       {  
         MyProjectUnitOfWork.LookupRepository.Create(JsonRead<CoreLookup>("<JSON file location>");  
         MyProjectUnitOfWork.Save();  
       }  
    }    
  }