Rystem.RepositoryFramework allows you to use correctly concepts like repository pattern, CQRS and DDD. You have interfaces for your domains, auto-generated api, auto-generated HttpClient to simplify connection "api to front-end", a functionality for auto-population in memory of your models, a functionality to simulate exceptions and waiting time from external sources to improve your implementation/business test and load test.
With this library you can add in memory integration with the chance to create random data with random values, random based on regular expressions and delegated methods
public static RepositoryInMemoryBuilder<T, TKey> AddRepositoryInMemoryStorage<T, TKey>(
this IServiceCollection services,
Action<RepositoryBehaviorSettings<T, TKey>>? settings = default)
where TKey : notnull
Populate with specific key (example with Guid)
public RepositoryInMemoryBuilder<TNext, Guid> AddRepositoryInMemoryStorageWithGuidKey<TNext>(Action<RepositoryBehaviorSettings<TNext, Guid>>? settings = default)
How to populate with random data?
Simple random (example)
Populate your in memory storage with 120 users
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddRepositoryInMemoryStorageWithStringKey<User>()
.PopulateWithRandomData(x => x.Email!, 120);
and in app after build during startup of your application
var app = builder.Build();
app.Services.Populate();
Simple random with regex (example)
Populate your in memory storage with 100 users and property Email with a random regex @"[a-z]{4,10}@gmail.com"
You can use the pattern in Class, IEnumerable, IDictionary, or Array, and in everything that extends IEnumerable or IDictionary
Important!! You can override regex service in your DI
public static IServiceCollection AddRegexService<T>(
this IServiceCollection services)
where T : class, IRegexService
IEnumerable or Array one-dimension (example)
You have your model x (User) that has a property Groups as IEnumerable or something that extends IEnumerable, Groups is a class with a property Id as string.
In the code below you are creating a list of class Groups with 8 elements in each 100 User instances, in each element of Groups you randomize based on this regex "[a-z]{4,5}".
You may take care of use First() linq method to set correctly the Id property.
and in app after build during startup of your application
app.Services.Populate()
IDictionary (example)
Similar to IEnumerable population you may populate your Claims property (a dictionary) with random key but with values based on regular expression "[a-z]{4,5}". As well as IEnumerable implementation you will have 6 elements (because I choose to create 6 elements in Populate method)
and in app after build during startup of your application
app.Services.Populate()
Populate with Implementation
If you have an interface or abstraction in your model, you can specify an implementation type for population.
You have two different methods, with typeof
If you want to test with possible exceptions (for your reliability tests) and waiting time (for your load tests) you may do it with this library and in memory behavior settings.
Add random exceptions
You can set different custom exceptions and different percentage for each operation: Delete, Get, Insert, Update, Query.
In the code below I'm adding three exceptions with a percentage of throwing them, they are the same for each operation.
I have a 0.45% for normal Exception, 0.1% for "Big Exception" and 0.548% for "Great Exception"
.AddRepositoryInMemoryStorage<User, string>(options =>
{
var customExceptions = new List<ExceptionOdds>
{
new ExceptionOdds()
{
Exception = new Exception(),
Percentage = 0.45
},
new ExceptionOdds()
{
Exception = new Exception("Big Exception"),
Percentage = 0.1
},
new ExceptionOdds()
{
Exception = new Exception("Great Exception"),
Percentage = 0.548
}
};
options.ExceptionOddsForDelete.AddRange(customExceptions);
options.ExceptionOddsForGet.AddRange(customExceptions);
options.ExceptionOddsForInsert.AddRange(customExceptions);
options.ExceptionOddsForUpdate.AddRange(customExceptions);
options.ExceptionOddsForQuery.AddRange(customExceptions);
})
Add random waiting time
You can set different range in milliseconds for each operation.
In the code below I'm adding a same custom range for Delete, Insert, Update, Get between 1000ms and 2000ms, and a unique custom range for Query between 3000ms and 7000ms.
.AddRepositoryInMemoryStorage<User, string>(options =>
{
var customRange = new Range(1000, 2000);
options.MillisecondsOfWaitForDelete = customRange;
options.MillisecondsOfWaitForInsert = customRange;
options.MillisecondsOfWaitForUpdate = customRange;
options.MillisecondsOfWaitForGet = customRange;
options.MillisecondsOfWaitForQuery = new Range(3000, 7000);
})
Add random waiting time before an exception
You can set different range in milliseconds for each operation before to throw an exception.
In the code below I'm adding a same custom range for Delete, Insert, Update, Get between 1000ms and 2000ms, and a unique custom range for Query between 3000ms and 7000ms in case of exception.
.AddRepositoryInMemoryStorage<User, string>(options =>
{
var customRange = new Range(1000, 2000);
options.MillisecondsOfWaitBeforeExceptionForDelete = customRange;
options.MillisecondsOfWaitBeforeExceptionForInsert = customRange;
options.MillisecondsOfWaitBeforeExceptionForUpdate = customRange;
options.MillisecondsOfWaitBeforeExceptionForGet = customRange;
options.MillisecondsOfWaitBeforeExceptionForQuery = new Range(3000, 7000);
var customExceptions = new List<ExceptionOdds>
{
new ExceptionOdds()
{
Exception = new Exception(),
Percentage = 0.45
},
new ExceptionOdds()
{
Exception = new Exception("Big Exception"),
Percentage = 0.1
},
new ExceptionOdds()
{
Exception = new Exception("Great Exception"),
Percentage = 0.548
}
};
options.ExceptionOddsForDelete.AddRange(customExceptions);
options.ExceptionOddsForGet.AddRange(customExceptions);