After the previous question "What are the important rules in Object Model Design", now I want to ask this:
Is there any way to have dynamic properties for class instances?
Suppose that we have this schematic object model:
So, each object could have lots of properties due to the set of implemented Interfaces, and then become relatively heavy object. Creating all the possible -and of course reasonable- object can be a way for solving this problem (i.e. Pipe_Designed v.s. Pipe_Designed_NeedInspection), but I have a large number of interfaces by now, that make it difficult.
I wonder if there is a way to have dynamic properties, something like the following dialog to allow the end user to select available functionalities for his/hers new object.
What you want is Properties pattern. Check out long and boring but clever article from Steve Yegge on this
I think maybe you're putting too many roles into the "Road" and "Pipe" classes, because your need for dynamic properties seems to derive from various states/phases of the artifacts in your model. I would consider making an explicit model using associations to different classes instead of putting everything in the "Road" or "Pipe" class using interfaces.
If you mean the number of public properties, use explicit interface implementation.
If you mean fields (and object space for sparse objects): you can always use a property bag for the property implementation.
For a C# example:
string IDesigned.ApprovedBy {
get {return GetValue<string>("ApprovedBy");}
set {SetValue("ApprovedBy", value);}
}
with a dictionary for the values:
readonly Dictionary<string, object> propValues =
new Dictionary<string, object>();
protected T GetValue<T>(string name)
{
object val;
if(!propValues.TryGetValue(name, out val)) return default(T);
return (T)val;
}
protected void SetValue<T>(string name, T value)
{
propValues[name] = value;
}
Note that SetValue would also be a good place for any notifications - for example, INotifyPropertyChanged in .NET to implement the observer pattern. Many other architectures have something similar. You can do the same with object keys (like how EventHandlerList works), but string keys are simpler to understand ;-p
This only then takes as much space as the properties that are actively being used.
A final option is to encapsulate the various facets;
class Foo {
public bool IsDesigned {get {return Design != null;}}
public IDesigned Design {get;set;}
// etc
}
Here Foo doesn't implement any of the interfaces, but provides access to them as properties.
Related
To externalize UI strings we use the "Messages-class" approach as supported e.g. in Eclipse and other IDEs. This approach requires that in each package where one needs some UI strings there has to be a class "Messages" that offers a static method String getString(key) via which one obtains the actual String to display to the user. The Strings are internally accessed/fetched using Java's Resources mechanism for i18n.
Esp. after some refactoring - we again and again have accidental imports from a class Messages from a different package.
Thus I would like to create an archunit rule checking whether we only access classes called "Messages" from the very same package. I.e. each import of a class x.y.z.Messages is an error if the package x.y.z is not the same package as the current class (i.e. the class that contains the import)
I got as far as this:
#ArchTest
void preventReferencesToMessagesOutsideCurrentPackage(JavaClasses classes) {
ArchRule rule;
rule = ArchRuleDefinition.noClasses()
.should().accessClassesThat().haveNameMatching("Messages")
.???
;
rule.check(classes);
}
but now I got stuck at the ???.
How can one phrase a condition "and the referenced/imported class "Messages" is not in the same package as this class"?
I somehow got lost with all these archunit methods of which none seems to fit here nor lend itself to compose said condition. Probably I just can't see the forest for the many trees.
Any suggestion or guidance anyone?
You need to operate on instances of JavaAccess to validate the dependencies. JavaAccess provides information about the caller and the target such that you can validate the access dynamically depending on the package name of both classes.
DescribedPredicate<JavaAccess<?>> isForeignMessageClassPredicate =
new DescribedPredicate<JavaAccess<?>>("target is a foreign message class") {
#Override
public boolean apply(JavaAccess<?> access) {
JavaClass targetClass = access.getTarget().getOwner();
if ("Message".equals(targetClass.getSimpleName())) {
JavaClass callerClass = access.getOwner().getOwner();
return !targetClass.getPackageName().equals(callerClass.getPackageName());
}
return false;
}
};
ArchRule rule =
noClasses().should().accessTargetWhere(isForeignMessageClassPredicate);
In my plugin code, I use early bound entities (generated via the crmsvcutil). Within my code, I am using MemberExpression to retrieve the name of the property. For instance, if I want the full name of the user who initiated the plugin I do the following
SystemUser pluginExecutedBy = new SystemUser();
pluginExecutedBy = Common.RetrieveEntity(service
, SystemUser.EntityLogicalName
, new ColumnSet(new string[] {Common.GetPropertyName(() => pluginExecutedBy.FullName)})
, localContext.PluginExecutionContext.InitiatingUserId).ToEntity<SystemUser>();
The code for GetPropertyName is as follows
public static string GetPropertyName<T>(Expression<Func<T>> expression)
{
MemberExpression body = (MemberExpression)expression.Body;
return body.Member.Name.ToLower();
}
The code for RetrieveEntity is as follows
public static Entity RetrieveEntity(IOrganizationService xrmService, string entityName, ColumnSet columns, Guid entityId)
{
return (Entity)xrmService.Retrieve(entityName, entityId, columns);
}
My solution architect's comments:
Instead of writing the code like above, why not write it like this (hardcoding the name of the field - or use a struct).
SystemUser pluginExecutedBy = null;
pluginExecutedBy = Common.RetrieveEntity(service
, SystemUser.EntityLogicalName
, new ColumnSet(new string[] {"fullname"})
, localContext.PluginExecutionContext.InitiatingUserId).ToEntity<SystemUser>();
Reason:
Your code unnecessarily creates an object before it requires it (as you instantiate the object with the new keyword before the RetrieveEntity in order to use it with my GetProperty method) which is bad programming practice. In my code, I have never used the new keyword, but merely casting it and casting does not create a new object. Now, I am no expert in C# or .NET, but I like to read and try out different things. So, I looked up the Microsoft.Xrm.Sdk.dll and found that ToEntity within Sdk, actually did create a new Entity using the keyword new.
If the Common.Retrieve returns null, your code has unnecessarily allocated memory which will cause performance issues whereas mine would not?
A managed language like C# "manages the memory" for me, does it not?
Question
Is my code badly written? If so, why? If it is better - why is it? (I believe it is a lot more cleaner and even if a field name changes as long as as the early bound class file is regenerated, I do not have to re-write any code)
I agree that cast does not create a new object, but does my code unnecessarily create objects?
Is there a better way (a completely different third way) to write the code?
Note: I suggested using the GetPropertyName because, he was hard-coding attribute names all over his code and so in a different project which did not use early bound entities I used structs for attribute names - something like below. I did this 3 weeks into my new job with CRM 2011 but later on discovered the magic of MemberExpression. He was writing a massive cs file for each of the entity that he was using in his plugin and I told him he did not have to do any of this as he could just use my GetPropertyName method in his plugin and get all the fields required and that prompted this code review comments. Normally he does not do a code review.
public class ClientName
{
public struct EntityNameA
{
public const string LogicalName = "new_EntityNameA";
public struct Attributes
{
public const string Name = "new_name";
public const string Status = "new_status";
}
}
}
PS: Or is the question / time spent analyzing just not worth it?
Early Bound, Late Bound, MemberExpression, bla bla bla :)
I can understand the "philosophy", but looking at your code a giant alarm popup in my head:
public static Entity RetrieveEntity(IOrganizationService xrmService, string entityName, ColumnSet columns, Guid entityId)
{
return (Entity)xrmService.Retrieve(entityName, entityId, columns);
}
the Retrieve throws an exception if the record is not found.
About the other things, the GetPropertyName is ok, but are always choices, for example I try to use always late bound in plugins, maybe in a project I prefer to use early bound, often there is more than one way to resolve a problem.
Happy crm coding!
Although GetPropertyName is a quite a clever solution I don't like it, and that's entirely to do with readability. To me its far easier to understand what is going on with: new ColumnSet(new string[] {"fullname"}).
But that's pretty much personal preference, but its important to remember that your not just writing code for yourself you are writing it for your team, they should be able to easily understand the work you have produced.
As a side a hardcoded string probably performs better at runtime. I usually hardcode all my values, if the entity model in CRM changes I will have to revisit to make changes in any case. There's no difference between early and late bound in that situation.
I don't understand the point of this function,
public static Entity RetrieveEntity(IOrganizationService xrmService, string entityName, ColumnSet columns, Guid entityId)
{
return (Entity)xrmService.Retrieve(entityName, entityId, columns);
}
It doesn't do anything (apart from cast something that is already of that type).
1.Your code unnecessarily creates an object before it requires it (as you instantiate the object with the new keyword before the
RetrieveEntity in order to use it with my GetProperty method) which is
bad programming practice. In my code, I have never used the new
keyword, but merely casting it and casting does not create a new
object.
I believe this refers to; SystemUser pluginExecutedBy = new SystemUser(); I can see his/her point here, in this case new SystemUser() doesn't do much, but if the object you were instantiating did something resource intensive (load files, open DB connections) you might be doing something 'wasteful'. In this case I would be surprised if changing SystemUser pluginExecutedBy = null; actually yielded any significant performance gain.
2.If the Common.Retrieve returns null, your code has unnecessarily allocated memory which will cause performance issues
I would be surprised if that caused a performance issue, and anyway as Guido points out that function wont return null in any case.
Overall there is little about this code I strongly feel needs changing - but things can be always be better and its worth discussing (e.g. the point of code review), although it can be hard not to you shouldn't be precious about your code.
Personally I would go with hardcoded attribute names and dump the Common.RetrieveEntity function as it doesn't do anything.
pluginExecutedBy = service.Retrieve(SystemUser.EntityLogicalName, localContext.PluginExecutionContext.InitiatingUserId, new ColumnSet(new String[] {"fullname"} ));
I would like to know the best practice for a class oriented DDD.
Since i am doing domain validation in custom setters named ChangeX(string x) i might be pushed to use this as property.
public virtual string example { get;
private set; }
However, that not very good since it disable me from using the object initialization feature such as :
new Object { Example = "Some example"
}
So i though why not passing the custom set into the property set ? like this
public virtual string Example { get {
return Example; } set {
ChangeExample(value); } }
Is this can lead to any problems ? it is against best practices ?
Thanks.
Real problem here is using setters as such. Why do You need them?
When You use setters, You lose isolation - You can modify state of objects from outside w/o them knowing that. That leads to procedural code.
In contrast - You should ask objects to do something (not just modify their state) that would eventually might lead to them changing their own state.
I think this solution is fine. One reason to have setters is to make sure your under laying fields never hold incorrect values.
Just wanted to get the groups thoughts on how to handle configuration details of entities.
What I'm thinking of specifically is high level settings which might be admin-changed. the sort of thing that you might store in the app or web.config ultimately, but from teh DDD perspective should be set somewhere in the objects explicitly.
For sake of argument, let's take as an example a web-based CMS or blog app.
A given blog Entry entity has any number of instance settings like Author, Content, etc.
But you also might want to set (for example) default Description or Keywords that all entries in the site should start with if they're not changed by the author. Sure, you could just make those constants in the class, but then the site owner couldn't change the defaults.
So my thoughts are as follows:
1) use class-level (static) properties to represent those settings, and then set them when the app starts up, either setting them from the DB or from the web.config.
or
2) use a separate entity for holding the settings, possibly a dictionary, either use it directly or have it be a member of the Entry class
What strikes you all as the most easy / flexible? My concerns abou the first one is that it doesn't strike me as very pluggable (if I end up wanting to add more features) as changing an entity's class methods would make me change the app itself as well (which feels like an OCP violation). The second one feels like it's more heavy, though, especially if I then have to cast or parse values out of a dictionary.
I would say that that whether a value is configurable or not is irrelevant from the Domain Model's perspective - what matters is that is is externally defined.
Let's say that you have a class that must have a Name. If the Name is always required, it must be encapsulated as an invariant irrespective of the source of the value. Here's a C# example:
public class MyClass
{
private string name;
public MyClass(string name)
{
if(name == null)
{
throw new ArgumentNullException("name");
}
this.name = name;
}
public string Name
{
get { return this.name; }
set
{
if(value == null)
{
throw new ArgumentNullException("name");
}
this.name = value;
}
}
}
A class like this effectively protects the invariant: Name must not be null. Domain Models must encapsulate invariants like this without any regard to which consumer will be using them - otherwise, they would not meet the goal of Supple Design.
But you asked about default values. If you have a good default value for Name, then how do you communicate that default value to MyClass.
This is where Factories come in handy. You simply separate the construction of your objects from their implementation. This is often a good idea in any case. Whether you choose an Abstract Factory or Builder implementation is less important, but Abstract Factory is a good default choice.
In the case of MyClass, we could define the IMyClassFactory interface:
public interface IMyClassFactory
{
MyClass Create();
}
Now you can define an implementation that pulls the name from a config file:
public ConfigurationBasedMyClassFactory : IMyClassFactory
{
public MyClass Create()
{
var name = ConfigurationManager.AppSettings["MyName"];
return new MyClass(name);
}
}
Make sure that code that needs instances of MyClass use IMyClassFactory to create it instead of new'ing it up manually.
If you have a class with some plain get/set properties, is there any reason to use the getters within the class methods, or should you just use the private member variables? I think there could be more of an argument over setters (validation logic?), but I'm wondering just about getters.
For example (in Java) - is there any reason to use option 2?:
public class Something
{
private int messageId;
public int getMessageId() { return this.messageId; }
public void setMessage(int messageId) { this.messageId = messageId; }
public void doSomething()
{
// Option 1:
doSomethingWithMessageId(messageId);
// Option 2:
doSomethingWithMessageId(getMessageId());
}
}
Java programmers in general tend to be very consistent about using getter methods. I program multiple languages and I'm not that consistent about it ;)
I'd say as long as you don't make a getter it's ok to use the raw variable - for private variables. When you make a getter, you should be using only that. When I make a getter for a private field, my IDE suggests that it replace raw field accesses for me automatically when I introduce a getter. Switching to using a getter is only a few keystrokes away (and without any chance of introducing errors), so I tend to delay it until I need it.
Of course, if you want to stuff like getter-injection, some types of proxying and subclassing framworks like hibernate, you have to user getters!
With getters you wont accidentally modify the variables :) Also, if you use both getters and the "raw" variable, your code can get confused.
Also, if you use inheritance and redefined the getter methods in child classes, getter-using methods will work properly, whereas those using the raw variables would not.
If you use the getter method everywhere - and in the future perform a code-search on all calls of getMessageId() you will find all of them, whereas if you had used the private ones, you may miss some.
Also if there's ever logic to be introduced in the setter method, you wont have to worry about changing more than 1 location for it.
If the value that you are assigning to the property is a known or verified value, you could safely use the private variable directly. (Except perhaps in some special situations, where it would be obvious why that would be bad.) Whether you do or not is more a matter of taste or style. It's not a performance issue either, as the getter or setter will be inlined by the compiler if it's simple enough.
If the value is unknown to the class, you should use the property to set it, so that you can protect the property from illegal values.
Here's an example (in C#):
public class Something {
private string _value;
public string Value {
get {
return _value;
}
set {
if (value == null) throw new ArgumentNullException();
_value = value;
}
}
public Something() {
// using a known value
_value = "undefined";
}
public Something(string initValue) {
// using an unknown value
Value = initValue;
}
}
If you use the getter you're ensuring you'll get the value after any logic/decisions have been applied to it. This probably isn't your typical situation but when it is, you'll thank yourself for this.
Unless I have a specific use case to use the internal field directly in the enclosing class, I've always felt that it's important to use access the field the same way it is accessed publicly. This ensures consistency in the return values across the board should there ever be any need to add some post-processing to the field via the getter method, or property. I feel like it's perfectly fine to access the raw field if you want its raw value for one reason or another.
More often than not, the getter encapsulation is plain and simple boilerplate code -- you're most likely not returning anything other than the field's value itself. However, in the case where you may want to change the way the data is presented at some point in the future, it's one less refactoring you have to make internally.