Simulate the "Per Graph" lifestyle in Castle Windsor - castle-windsor

I'm trying to simulate the "Per Graph" lifestyle in Windsor in order to resolve my objects in the correct way. I have tried the Bound lifestyle but without any success, probably due to a lack of knowledge on the subject.
What I need is expressed in the following example code:
class Program
{
static void Main( string[] args )
{
var container = new WindsorContainer();
container.Register(
Component.For<ClassA>().LifestyleBoundTo<ClassC1>(),
//Component.For<ClassA>().LifestyleBoundTo<ClassC2>(),
Component.For<ClassB1>().LifeStyle.Transient,
Component.For<ClassB2>().LifeStyle.Transient,
Component.For<ClassC1>().LifeStyle.Transient,
Component.For<ClassC2>().LifeStyle.Transient );
var c1 = container.Resolve<ClassC1>();
Assert.AreSame( c1.B1.A, c1.A );
var c2 = container.Resolve<ClassC2>();
Assert.AreSame( c2.B2.A, c2.A );
Assert.AreNotSame( c1.A, c2.A );
}
}
class ClassA
{
}
class ClassB1
{
public ClassB1( ClassA a)
{
A = a;
}
public ClassA A { get; set; }
}
class ClassC1
{
public ClassC1( ClassA a, ClassB1 b1 )
{
A = a;
B1 = b1;
}
public ClassA A { get; set; }
public ClassB1 B1 { get; set; }
}
class ClassB2
{
public ClassB2( ClassA a )
{
A = a;
}
public ClassA A { get; set; }
}
class ClassC2
{
public ClassC2( ClassA a, ClassB2 b2 )
{
A = a;
B2 = b2;
}
public ClassA A { get; set; }
public ClassB2 B2 { get; set; }
}
}

Ok so it all works as expected.
Your first assert passes as expected. The second one fails as expected, since you're scoping ClassA per ClassC1 and there's no instance of that class in the graph for the second resolve.

Related

Is it possible to tell Newtonsoft to map a declared $type to a different type when deserialising? [duplicate]

Suppose I have the following class hierarchy:
public abstract class Organization
{
/* properties related to all organizations */
}
public sealed class Company : Organization
{
/* properties related to companies */
}
public sealed class NonProfitOrganization : Organization
{
/* properties related to non profit organizations */
}
Is it possible to have json.net use property (say "type" or "discriminator") to determine which type the object when it deserializes the organization? For example, the following should deserialize an instance of Company.
{
"type": "company"
/* other properties related to companies */
}
And the following should deserialize an instance of NonProfitOrganization.
{
"type": "non-profit"
/* other properties related to non profit */
}
When I call the following:
Organization organization = JsonConvert.DeserializeObject<Organization>(payload);
where payload is the above JSON snippets. I had a look at setting the "TypeNameHandling" on properties or classes but it serializes the whole .NET type, which isn't "portable" between the client and server when the classes are defined in different namespaces and assemblies.
I'd rather define the type is a neutral manner which clients written in any language can use to determine the actual type of the object type being serialized.
In case you are still looking, here is an example: http://james.newtonking.com/archive/2011/11/19/json-net-4-0-release-4-bug-fixes.aspx
This will allow you to create a table based mapping:
public class TypeNameSerializationBinder : SerializationBinder
{
public TypeNameSerializationBinder(Dictionary<Type, string> typeNames = null)
{
if (typeNames != null)
{
foreach (var typeName in typeNames)
{
Map(typeName.Key, typeName.Value);
}
}
}
readonly Dictionary<Type, string> typeToName = new Dictionary<Type, string>();
readonly Dictionary<string, Type> nameToType = new Dictionary<string, Type>(StringComparer.OrdinalIgnoreCase);
public void Map(Type type, string name)
{
this.typeToName.Add(type, name);
this.nameToType.Add(name, type);
}
public override void BindToName(Type serializedType, out string assemblyName, out string typeName)
{
var name = typeToName.Get(serializedType);
if (name != null)
{
assemblyName = null;
typeName = name;
}
else
{
assemblyName = serializedType.Assembly.FullName;
typeName = serializedType.FullName;
}
}
public override Type BindToType(string assemblyName, string typeName)
{
if (assemblyName == null)
{
var type = this.nameToType.Get(typeName);
if (type != null)
{
return type;
}
}
return Type.GetType(string.Format("{0}, {1}", typeName, assemblyName), true);
}
}
The code has a slight defect in that if a type name mapping is attempted where the type is unique but the name is already used, the Map method will throw an exception after the type-to-name mapping is already added leaving the table in an inconsistent state.
To take eulerfx's answer further; I wanted to apply DisplayName attribute to a class and have that automatically become the type name used; to that end:
public class DisplayNameSerializationBinder : DefaultSerializationBinder
{
private Dictionary<string, Type> _nameToType;
private Dictionary<Type, string> _typeToName;
public DisplayNameSerializationBinder()
{
var customDisplayNameTypes =
this.GetType()
.Assembly
//concat with references if desired
.GetTypes()
.Where(x => x
.GetCustomAttributes(false)
.Any(y => y is DisplayNameAttribute));
_nameToType = customDisplayNameTypes.ToDictionary(
t => t.GetCustomAttributes(false).OfType<DisplayNameAttribute>().First().DisplayName,
t => t);
_typeToName = _nameToType.ToDictionary(
t => t.Value,
t => t.Key);
}
public override void BindToName(Type serializedType, out string assemblyName, out string typeName)
{
if (false == _typeToName.ContainsKey(serializedType))
{
base.BindToName(serializedType, out assemblyName, out typeName);
return;
}
var name = _typeToName[serializedType];
assemblyName = null;
typeName = name;
}
public override Type BindToType(string assemblyName, string typeName)
{
if (_nameToType.ContainsKey(typeName))
return _nameToType[typeName];
return base.BindToType(assemblyName, typeName);
}
}
and usage example:
public class Parameter
{
public string Name { get; set; }
};
[DisplayName("bool")]
public class BooleanParameter : Parameter
{
}
[DisplayName("string")]
public class StringParameter : Parameter
{
public int MinLength { get; set; }
public int MaxLength { get; set; }
}
[DisplayName("number")]
public class NumberParameter : Parameter
{
public double Min { get; set; }
public double Max { get; set; }
public string Unit { get; set; }
}
[DisplayName("enum")]
public class EnumParameter : Parameter
{
public string[] Values { get; set; }
}
internal class Program
{
private static void Main(string[] args)
{
var parameters = new Parameter[]
{
new BooleanParameter() {Name = "alive"},
new StringParameter() {Name = "name", MinLength = 0, MaxLength = 10},
new NumberParameter() {Name = "age", Min = 0, Max = 120},
new EnumParameter() {Name = "status", Values = new[] {"Single", "Married"}}
};
JsonConvert.DefaultSettings = () => new JsonSerializerSettings
{
Binder = new DisplayNameSerializationBinder(),
TypeNameHandling = TypeNameHandling.Auto,
NullValueHandling = NullValueHandling.Ignore,
DefaultValueHandling = DefaultValueHandling.Ignore,
Formatting = Formatting.Indented,
ContractResolver = new CamelCasePropertyNamesContractResolver()
};
var json = JsonConvert.SerializeObject(parameters);
var loadedParams = JsonConvert.DeserializeObject<Parameter[]>(json);
Console.WriteLine(JsonConvert.SerializeObject(loadedParams));
}
}
output:
[
{
"$type": "bool",
"name": "alive"
},
{
"$type": "string",
"maxLength": 10,
"name": "name"
},
{
"$type": "number",
"max": 120.0,
"name": "age"
},
{
"$type": "enum",
"values": [
"Single",
"Married"
],
"name": "status"
}
]
I've written purely declarative solution with ability to specify custom discriminator field, and provide scoped name handling per base class (as opposed to usecure global JsonSerializationSettings, especially on different Web-Api when we do not have ability to specify custom JsonSerializationSettings).
using System;
using Newtonsoft.Json;
using Newtonsoft.Json.Linq;
using System.Reflection;
using System.Linq;
using System.Collections.Generic;
// Discriminated Json Converter (JsonSubtypes) implementation for .NET
//
// MIT License
//
// Copyright (c) 2016 Anatoly Ressin
// 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.
////////////////////// USAGE ////////////////////////////////////////////////////////////////////////////////
[JsonConverter(typeof(JsonSubtypes))] // Discriminated base class SHOULD NOT be abstract
public class ShapeBase {
[JsonTag, JsonProperty("#type")] // it SHOULD contain a property marked with [JsonTag]
public string Type {get;set;} // only one [JsonTag] annotation allowed per discriminated class
// it COULD contain other properties, however this is NOT RECOMMENDED
// Rationale: instances of this class will be created at deserialization
// only for tag sniffing, and then thrown away.
}
public abstract class Shape: ShapeBase { // If you want abstract parent - extend the root
public abstract double GetArea(); // with needed abstract stuff, then use this class everywhere (see DEMO below)
}
[JsonSubtype("circle")] // Every final class-case SHOULD be marked with [JsonSubtype(tagValue)]
public class Circle: Shape { // Two disctinct variant classes MUST have distinct tagValues
[JsonProperty("super-radius")] // You CAN use any Json-related annotation as well
public double Radius { get; set; }
public override double GetArea() {
return Radius * Radius * Math.PI;
}
}
[JsonSubtype("rectangle")]
public class Rectangle: Shape {
public double Height { get; set; }
public double Width { get; set; }
public override double GetArea() {
return Width * Height;
}
}
[JsonSubtype("group")]
public class Group: Shape {
[JsonProperty("shapes")]
public List<Shape> Items { get; set; }
public override double GetArea() {
return Items.Select(item => item.GetArea()).Sum();
}
}
// Every final class-case SHOULD be registered with JsonSubtypes.register(typeof(YourConcreteClass))
// either manually or with auto-register capability:
// You can auto-register all classes marked with [JsonSubtype(tag)] in given Assembly
// using JsonSubtypes.autoRegister(yourAssembly)
////////////////// DEMO /////////////////////////////////////////////////////////////////////////////////
public class Program
{
public static void Main()
{
JsonSubtypes.autoRegister(Assembly.GetExecutingAssembly());
Shape original = new Group() {
Items = new List<Shape> {
new Circle() { Radius = 5 },
new Rectangle() { Height = 10, Width = 20 }
}
};
string str = JsonConvert.SerializeObject(original);
Console.WriteLine(str);
var copy = JsonConvert.DeserializeObject(str,typeof(Shape)) as Shape;
// Note: we can deserialize object using any class from the hierarchy.
// Under the hood, anyway, it will be deserialized using the top-most
// base class annotated with [JsonConverter(typeof(JsonSubtypes))].
// Thus, only soft-casts ("as"-style) are safe here.
Console.WriteLine("original.area = {0}, copy.area = {1}", original.GetArea(), copy.GetArea());
}
}
//////////////////////// IMPLEMENTATION //////////////////////////////////////////////////////////////////
public class JsonSubtypeClashException: Exception {
public string TagValue { get; private set;}
public Type RootType { get; private set; }
public Type OldType { get; private set; }
public Type NewType { get; private set; }
public JsonSubtypeClashException(Type rootType, string tagValue, Type oldType, Type newType): base(
String.Format(
"JsonSubtype Clash for {0}[tag={1}]: oldType = {2}, newType = {3}",
rootType.FullName,
tagValue,
oldType.FullName,
newType.FullName
)
) {
TagValue = tagValue;
RootType = rootType;
OldType = oldType;
NewType = newType;
}
}
public class JsonSubtypeNoRootException: Exception {
public Type SubType { get; private set; }
public JsonSubtypeNoRootException(Type subType): base(
String.Format(
"{0} should be inherited from the class with the [JsonConverter(typeof(JsonSubtypes))] attribute",
subType.FullName
)
) {
SubType = subType;
}
}
public class JsonSubtypeNoTagException: Exception {
public Type SubType { get; private set; }
public JsonSubtypeNoTagException(Type subType): base(
String.Format(
#"{0} should have [JsonSubtype(""..."")] attribute",
subType.FullName
)
) {
SubType = subType;
}
}
public class JsonSubtypeNotRegisteredException: Exception {
public Type Root { get; private set; }
public string TagValue { get; private set; }
public JsonSubtypeNotRegisteredException(Type root, string tagValue): base(
String.Format(
#"Unknown tag={1} for class {0}",
root.FullName,
tagValue
)
) {
Root = root;
TagValue = tagValue;
}
}
[AttributeUsage(AttributeTargets.Class)]
public class JsonSubtypeAttribute: Attribute {
private string tagValue;
public JsonSubtypeAttribute(string tagValue) {
this.tagValue = tagValue;
}
public string TagValue {
get {
return tagValue;
}
}
}
public static class JsonSubtypesExtension {
public static bool TryGetAttribute<T>(this Type t, out T attribute) where T: Attribute {
attribute = t.GetCustomAttributes(typeof(T), false).Cast<T>().FirstOrDefault();
return attribute != null;
}
private static Dictionary<Type, PropertyInfo> tagProperties = new Dictionary<Type, PropertyInfo>();
public static bool TryGetTagProperty(this Type t, out PropertyInfo tagProperty) {
if (!tagProperties.TryGetValue(t, out tagProperty)) {
JsonConverterAttribute conv;
if (t.TryGetAttribute(out conv) && conv.ConverterType == typeof(JsonSubtypes)) {
var props = (from prop in t.GetProperties() where prop.GetCustomAttribute(typeof(JsonTagAttribute)) != null select prop).ToArray();
if (props.Length == 0) throw new Exception("No tag");
if (props.Length > 1) throw new Exception("Multiple tags");
tagProperty = props[0];
} else {
tagProperty = null;
}
tagProperties[t] = tagProperty;
}
return tagProperty != null;
}
public static bool TryGetTagValue(this Type t, out string tagValue) {
JsonSubtypeAttribute subtype;
if (t.TryGetAttribute(out subtype)) {
tagValue = subtype.TagValue;
return true;
} else {
tagValue = null;
return false;
}
}
public static bool TryGetJsonRoot(this Type t, out Type root, out PropertyInfo tagProperty) {
root = t;
do {
if (root.TryGetTagProperty(out tagProperty)) {
return true;
}
root = root.BaseType;
} while (t != null);
return false;
}
}
public class JsonTagAttribute: Attribute {
}
public class JsonTagInfo {
public PropertyInfo Property { get; set; }
public string Value { get; set; }
}
public class JsonRootInfo {
public PropertyInfo Property { get; set; }
public Type Root { get; set; }
}
public abstract class DefaultJsonConverter: JsonConverter {
[ThreadStatic]
private static bool silentWrite;
[ThreadStatic]
private static bool silentRead;
public sealed override bool CanWrite {
get {
var canWrite = !silentWrite;
silentWrite = false;
return canWrite;
}
}
public sealed override bool CanRead {
get {
var canRead = !silentRead;
silentRead = false;
return canRead;
}
}
protected void _WriteJson(JsonWriter writer, Object value, JsonSerializer serializer) {
silentWrite = true;
serializer.Serialize(writer, value);
}
protected Object _ReadJson(JsonReader reader, Type objectType, Object existingValue, JsonSerializer serializer) {
silentRead = true;
return serializer.Deserialize(reader, objectType);
}
}
public class JsonSubtypes: DefaultJsonConverter {
private static Dictionary<Type, Dictionary<string, Type>> implementations = new Dictionary<Type, Dictionary<string, Type>>();
private static Dictionary<Type, JsonTagInfo> tags = new Dictionary<Type, JsonTagInfo>();
private static Dictionary<Type, JsonRootInfo> roots = new Dictionary<Type, JsonRootInfo>();
public static void register(Type newType) {
PropertyInfo tagProperty;
Type root;
if (newType.TryGetJsonRoot(out root, out tagProperty)) {
for(var t = newType; t != root; t = t.BaseType) {
roots[t] = new JsonRootInfo() {
Property = tagProperty,
Root = root
};
}
roots[root] = new JsonRootInfo() {
Property = tagProperty,
Root = root
};
Dictionary<string, Type> implementationMap;
if (!implementations.TryGetValue(root, out implementationMap)) {
implementationMap = new Dictionary<string, Type>();
implementations[root] = implementationMap;
}
JsonSubtypeAttribute attr;
if (!newType.TryGetAttribute(out attr)) {
throw new JsonSubtypeNoTagException(newType);
}
var tagValue = attr.TagValue;
Type oldType;
if (implementationMap.TryGetValue(tagValue, out oldType)) {
throw new JsonSubtypeClashException(root, tagValue, oldType, newType);
}
implementationMap[tagValue] = newType;
tags[newType] = new JsonTagInfo() {
Property = tagProperty,
Value = tagValue
};
} else {
throw new JsonSubtypeNoRootException(newType);
}
}
public static void autoRegister(Assembly assembly) {
foreach(var type in assembly.GetTypes().Where(type => type.GetCustomAttribute<JsonSubtypeAttribute>() != null)) {
register(type);
}
}
public override bool CanConvert(Type t) {
return true;
}
public static T EnsureTag<T>(T value) {
JsonTagInfo tagInfo;
if (tags.TryGetValue(value.GetType(), out tagInfo)) {
tagInfo.Property.SetValue(value, tagInfo.Value);
}
return value;
}
public override void WriteJson(JsonWriter writer, Object value, JsonSerializer serializer) {
_WriteJson(writer, EnsureTag(value), serializer);
}
public override Object ReadJson(JsonReader reader, Type objectType, Object existingValue, JsonSerializer serializer) {
JsonTagInfo tagInfo;
if (tags.TryGetValue(objectType, out tagInfo)) {
return _ReadJson(reader, objectType, existingValue, serializer);
} else {
JsonRootInfo rootInfo;
if (roots.TryGetValue(objectType, out rootInfo)) {
JToken t = JToken.ReadFrom(reader);
var stub = _ReadJson(t.CreateReader(), rootInfo.Root, existingValue, serializer);
var tagValue = rootInfo.Property.GetValue(stub) as string;
var implementationMap = implementations[rootInfo.Root];
Type implementation;
if (implementationMap.TryGetValue(tagValue, out implementation)) {
return ReadJson(t.CreateReader(), implementation, null, serializer);
} else {
throw new JsonSubtypeNotRegisteredException(rootInfo.Root, tagValue);
}
} else {
return _ReadJson(reader, objectType, existingValue, serializer);
}
}
}
public static T Deserialize<T>(string s) where T: class {
return JsonConvert.DeserializeObject(s, typeof(T)) as T;
}
public static string Serialize<T>(T value) where T: class {
return JsonConvert.SerializeObject(value);
}
}
output:
{"shapes":[{"super-radius":5.0,"#type":"circle"},{"Height":10.0,"Width":20.0,"#type":"rectangle"}],"#type":"group"}
original.area = 278.539816339745, copy.area = 278.539816339745
You can grab it here:
https://dotnetfiddle.net/ELcvnk
With another JsonSubtypes converter implementation.
Usage:
[JsonConverter(typeof(JsonSubtypes), "Sound")]
[JsonSubtypes.KnownSubType(typeof(Dog), "Bark")]
[JsonSubtypes.KnownSubType(typeof(Cat), "Meow")]
public class Animal
{
public virtual string Sound { get; }
public string Color { get; set; }
}
public class Dog : Animal
{
public override string Sound { get; } = "Bark";
public string Breed { get; set; }
}
public class Cat : Animal
{
public override string Sound { get; } = "Meow";
public bool Declawed { get; set; }
}
[TestMethod]
public void Demo()
{
var input = #"{""Sound"":""Bark"",""Breed"":""Jack Russell Terrier""}"
var animal = JsonConvert.DeserializeObject<Animal>(input);
Assert.AreEqual("Jack Russell Terrier", (animal as Dog)?.Breed);
}
the converter implementation can be directly downloaded from the repository: JsonSubtypes.cs and is also availble as a nuget package
Use this JsonKnownTypes, it's very similar way to use, add couple of attribute:
[JsonConverter(typeof(JsonKnownTypeConverter<Organization>))]
[JsonDiscriminator(Name = "discriminator")]
[JsonKnownType(typeof(Company), "company")]
[JsonKnownType(typeof(NonProfitOrganization), "non-profit")]
public abstract class Organization
{
/* properties related to all organizations */
}
public sealed class Company : Organization
{
/* properties related to companies */
}
public sealed class NonProfitOrganization : Organization
{
/* properties related to non profit organizations */
}
And serialize:
var json = JsonConvert.SerializeObject(youObject)
Output json:
{..., "discriminator":"non-profit"} //if object was NonProfitOrganization
Deserialization:
var organization = JsonConvert.DeserializeObject<Organization>(payload);

Castle Windsor Property Injection does not work with dynamic parameters

We want to use NserviceBus Saga's and in order to do that you need parameterless constructors for your saga. The only other way to inject our concerns is to use property injection which does not appear to work. Any help and guidance would be greatly appreciated.
I have posted sample code below that show's the issue.
class Program
{
static void Main(string[] args)
{
var container = new WindsorContainer();
container.Register(Component.For<ClassWithDynamicProperty>()
.DynamicParameters((r, k) =>{ k["testString"] = "test"; }) );
container.Register(
Component.For<TestClassWithPropertyInjection>());
container.Register(
Component.For<TestClassWithConstructorInjection>());
var class1 = container.Resolve<TestClassWithConstructorInjection>();
var class2 = container.Resolve<TestClassWithPropertyInjection>();
Debug.Assert(class1.DynamicClass == null);
Debug.Assert(class2.ClassWithDynamicProperty == null);
}
}
internal class TestClassWithPropertyInjection
{
public TestClassWithPropertyInjection()
{
}
public ClassWithDynamicProperty ClassWithDynamicProperty { get; set; }
}
internal class TestClassWithConstructorInjection
{
private readonly ClassWithDynamicProperty _classWithDynamicProperty;
public TestClassWithConstructorInjection(ClassWithDynamicProperty classWithDynamicProperty)
{
_classWithDynamicProperty = classWithDynamicProperty;
}
public ClassWithDynamicProperty DynamicClass { get { return _classWithDynamicProperty; } }
}
public class ClassWithDynamicProperty
{
public string TestString { get; private set; }
public ClassWithDynamicProperty(string testString)
{
TestString = testString;
}
}

Error while storing a list of custom object in isolated storage

I have a custom class of the following type
[Table]
class MyApp
{
public MyApp()
: base()
{
}
[Column(IsPrimaryKey=true, UpdateCheck = UpdateCheck.Never)]
public string appCode { get; set; }
[Column(UpdateCheck = UpdateCheck.Never)]
public string procCode { get; set; }
}
I have another class which contains a list of MyApp objects which is as below:
[Table]
class ApplicationUser
{
public ApplicationUser()
:base()
{
}
[Column(IsPrimaryKey = true, UpdateCheck = UpdateCheck.Never)]
public string userId { get; set; }
[Column(UpdateCheck = UpdateCheck.Never)]
public List<MyApp> applicationList { get; set; }
}
While calling the CreateDatabase() method in my DataContext class i get the following error:
Unable to determine SQL type for 'System.Collections.Generic.List`1[XCell.Framework.data.MyApp]'
Please guide me on this.
As I see the problem is that the applicationList is marked with the Column attribute, though it represents a relation.
Basically you will have to correctly map the relation between those entities using EntityRef<T> and EntitySet<T> classes and the Association attribute.
This article may be helpful.
The example with the corrected mapping (for one-to-many relation) below:
Adjusted ApplicationUser class
[Table]
public class ApplicationUser
{
private EntitySet<MyApp> _userApplications = new EntitySet<MyApp>();
[Column(IsPrimaryKey = true, UpdateCheck = UpdateCheck.Never)]
public string UserId { get; set; }
[Association(Storage = "_userApplications", ThisKey = "UserId", OtherKey = "ApplicationUserId")]
public EntitySet<MyApp> ApplicationList
{
get { return _userApplications; }
set { _userApplications = value; }
}
}
And the adjusted MyApp class
[Table]
public class MyApp
{
private EntityRef<ApplicationUser> _applicationUserRef;
[Column(IsPrimaryKey = true, UpdateCheck = UpdateCheck.Never)]
public string AppCode { get; set; }
[Column(UpdateCheck = UpdateCheck.Never)]
public string ProcCode { get; set; }
[Column]
public string ApplicationUserId { get; set; }
[Association(Name = "FK_MyApp_ApplicationUser", Storage = "_applicationUserRef", ThisKey = "ApplicationUserId", OtherKey = "UserId")]
public ApplicationUser ApplicationUserReference
{
get { return _applicationUserRef.Entity; }
set { _applicationUserRef.Entity = value; }
}
}

context.SaveChanges() not updating data in SqlServer database

When I call context.SaveChanges() to update a specific product, the update is not registered in the database. I do not get any runtime error either. All I notice is that my product catalog is not updated. I still see the same values. When I run the debugger I notice that the connection state of the database is closed.
This is the class implementing the context.SaveChanges()
namespace SportsStore.Domain.Concrete
{
public class EFProductRepository : IProductRepository
{
private EFDbContext context = new EFDbContext();
public IQueryable<Product> Products
{
get { return context.Products; }
}
public void SaveProduct(Product product)
{
if (product.ProductID == 0)
{
context.Products.Add(product);
}
context.SaveChanges();
}
}
}
namespace SportsStore.Domain.Concrete
{
public class EFDbContext : DbContext
{
public DbSet<Product> Products { get; set; }
}
}
namespace SportsStore.Domain.Entities
{
public class Product
{
[HiddenInput(DisplayValue=false)]
public int ProductID { get; set; }
public string Name { get; set; }
[DataType(DataType.MultilineText)]
public string Description { get; set; }
public string Category { get; set; }
public decimal Price { get; set; }
}
}
In the EFProductRepository class, prior to calling the context.SaveChanges() method in the SaveProduct method, you can use one of the following approaches to persist changes to the database.
public void SaveProduct(Product product)
{
if (product.ProductID == 0)
{
context.Products.Add(product);
}
//One approach to persist changes to database
//var productInDB = context.Products.Single(x => x.ProductID ==product.ProductID);
//context.Entry(productInDB).CurrentValues.SetValues(product);
//context.SaveChanges();
//Alternate Approach
if (product.ProductID != 0)
{
context.Entry(product).State = System.Data.EntityState.Modified;
}
context.SaveChanges();
}

Is there a way to skip tests in base classes?

I'm using C# 4.0, Visual Studio 2010 and am annotating my methods/classes with the attributes from the Microsoft.VisualStudio.TestTools.UnitTesting namespace.
I'd like to use inheritance in my test classes where each additional inheritance represents something changing or being created. If I could get it to not run tests from the base classes, then everything would be fine. Here's a rough example:
public class Person
{
public int Energy { get; private set; }
public int AppleCount { get; private set; }
public Person()
{
this.Energy = 10;
this.AppleCount = 5;
}
public void EatApple()
{
this.Energy += 5;
this.AppleCount--;
}
}
[TestClass]
public class PersonTest
{
protected Person _person;
[TestInitialize]
public virtual void Initialize()
{
this._person = new Person();
}
[TestMethod]
public void PersonTestEnergy()
{
Assert.AreEqual(10, this._person.Energy);
}
[TestMethod]
public void PersonTestAppleCount()
{
Assert.AreEqual(5, this._person.AppleCount);
}
}
[TestClass]
public class PersonEatAppleTest : PersonTest
{
[TestInitialize]
public override void Initialize()
{
base.Initialize();
this._person.EatApple();
}
[TestMethod]
public void PersonEatAppleTestEnergy()
{
Assert.AreEqual(15, this._person.Energy);
}
[TestMethod]
public void PersonEatAppleTestAppleCount()
{
Assert.AreEqual(4, this._person.AppleCount);
}
}
I asked a coworker and he suggested separating the initializing code from the testing. Inherit all the setup code, but then place all the tests for a particular setup in a class that inherits from said setup code. So the above would become:
public class Person
{
public int Energy { get; private set; }
public int AppleCount { get; private set; }
public Person()
{
this.Energy = 10;
this.AppleCount = 5;
}
public void EatApple()
{
this.Energy += 5;
this.AppleCount--;
}
}
[TestClass]
public class PersonSetup
{
protected Person _person;
[TestInitialize]
public virtual void Initialize()
{
this._person = new Person();
}
}
[TestClass]
public class PersonTest : PersonSetup
{
[TestMethod]
public void PersonTestEnergy()
{
Assert.AreEqual(10, this._person.Energy);
}
[TestMethod]
public void PersonTestAppleCount()
{
Assert.AreEqual(5, this._person.AppleCount);
}
}
[TestClass]
public class PersonEatAppleSetup : PersonSetup
{
[TestInitialize]
public override void Initialize()
{
base.Initialize();
this._person.EatApple();
}
}
[TestClass]
public class PersonEatAppleTest : PersonEatAppleSetup
{
[TestMethod]
public void PersonEatAppleTestEnergy()
{
Assert.AreEqual(15, this._person.Energy);
}
[TestMethod]
public void PersonEatAppleTestAppleCount()
{
Assert.AreEqual(4, this._person.AppleCount);
}
}
If someone else knows how to skip inherited methods like I originally asked, then I'll accept that. If not, then eventually I'll accept this answer.