//see https://blogs.msdn.microsoft.com/dotnet/2016/09/29/implementing-seeding-custom-conventions-and-interceptors-in-ef-core-1-0/
//for why I call DetectChanges before ChangeTracker, and why I then turn ChangeTracker.AutoDetectChangesEnabled off/on around SaveChanges
public static async Task <ImmutableList <ValidationResult> > SaveChangesWithValidationAsync(this DataContext dataContext)
{
var result = dataContext.ExecuteValidation();
if (result.Count > 1)
{
return(result);
}
dataContext.ChangeTracker.AutoDetectChangesEnabled = false;
try
{
await dataContext.SaveChangesAsync().ConfigureAwait(false);
}
finally
{
dataContext.ChangeTracker.AutoDetectChangesEnabled = true;
}
return(result);
}
//see https://blogs.msdn.microsoft.com/dotnet/2016/09/29/implementing-seeding-custom-conventions-and-interceptors-in-ef-core-1-0/
//for why I call DetectChanges before ChangeTracker, and why I then turn ChangeTracker.AutoDetectChangesEnabled off/on around SaveChanges
public static ImmutableList <ValidationResult> SaveChangesWithValidation(this DataContext dataContext)
{
var result = dataContext.ExecuteValidation(); //#C
if (result.Count > 1)
{
return(result); //#D
}
//I leave out the AutoDetectChangesEnabled on/off from the code shown in chapter 4 as its only a performance issue
//I'ts a concept that doesn't add anything to chapter 4. However I leave it in the real code as it has a (small) improvement on performance
dataContext.ChangeTracker.AutoDetectChangesEnabled = false; //LEAVE OUT OF CHAPTER 4 -
try
{
dataContext.SaveChanges();
}
finally
{
dataContext.ChangeTracker.AutoDetectChangesEnabled = true;
}
return(result);
}