I have a class design that looks like this one:
When Main receives an item, properties of other items can change based on some rules. For example, if Item1 has a red color, item2.level must be 3. Sometimes a value can change properties of multiple items. Implementing all rules in Main class is possible but it's a mess code.
So, I have implemented the Aggregate class that include all items. When Main.setItem(item1) is executed, I update Aggregate with Aggregate.setItem1 and run Rules.updateAggregate that updates items of Aggregate according to all rules.
It works, but Rules class is very inefficient. Because it doesn't know which item has updated, it apply all rules all times. Also, updateAggregate method is very large and difficult to test.
Is there any better design?
To reduce the coupling you could use a design based on events.
Main and Item classes should be publishers of events and Item classes should be also subscribers of event types as they want to react to events.
You should define some event types. For example :
event type : "new Item with ..."
event type : "Item 1 changed the foo property"
and so for...
According to question and comments, here is one example of a design that could work. Gist link to the example code
The example is made using C# but should not have any problems to implement other languages. (Could be easier with dynamic languages)
The design probably doesn’t fit directly to your use case but here are a few points what I try to demonstrate with it and maybe you can find better ideas with this way.
I made a few iterations and this was the first version that I could work on a similar use case without any third party libraries.
Main points of design
In the example, I try to keep Item -classes very clean from rules.
Rules -class is disappeared because, in the example, rules are in Aggregate class. However, I don’t see any problem to implement the same rules in separated Rules class, if needed. I don’t know exactly how complex your rules are and tried to keep the example simple.
The aggregate has methods to raise “notification” about Item changes.
I didn’t test my example but the implementation should be easier to test. For example, you should be put Items on specific state and try how rule behavior works. This can be validated to use unit tests but I didn't have any real use cases.
I tried to kept classes decoupled as possible.
Downsides
Depending on your needs, classes can be too coupled. Specially Aggregate can contain a lot of code so the more decoupled solution should be possible.
Implementation can be varied by programming languages and not always so clean. Originally, I used a similar pattern with event sourcing and modified to this example.
If a number of Items are huge, then the design is not very flexible.
How to improve
As said in comments and other answers, event-based implementation is probably what you want. I used Observer pattern as a starting point but it was going towards event-based implementation.
You should also look at some event-driven libraries that can have a good example of how to implement an application that reacts to changes. Also, those libraries usually help to wire up objects and give more decoupled implementation. Of course, then you are coupled with those libraries but probably not a bad case. No need to reinvent this mechanism anyway.
Check following links that can help with new ideas. Observer vs pub-sub pattern
I completelly agree with #davidxxx that events based design is better approach in your case. You can have a look at the observer design pattern. In case your business logic is so complex that you are not able to provide reqired parameters , you can always inject the whole list/tree/whatever of items and perform find operations on it.
Related
I haven't found any clear articles on this, but I was wondering about why polymorphism is the recommended design pattern over exhaustive switch case / pattern matching. I ask this because I've gotten a lot of heat from experienced developers for not using polymorphic classes, and it's been troubling me. I've personally had a terrible time with polymorphism and a wonderful time with switch cases, the reduction in abstractions and indirection makes readability of the code so much easier in my opinion. This is in direct contrast with books like "clean code" which are typically seen as industry standards.
Note: I use TypeScript, so the following examples may not apply in other languages, but I think the principle generally applies as long as you have exhaustive pattern matching / switch cases.
List the options
If you want to know what the possible values of an action, with an enum, switch case, this is trivial. For classes this requires some reflection magic
// definitely two actions here, I could even loop over them programmatically with basic primitives
enum Action {
A = 'a',
B = 'b',
}
Following the code
Dependency injection and abstract classes mean that jump to definition will never go where you want
function doLetterThing(myEnum: Action) {
switch (myEnum) {
case Action.A:
return;
case Action.B;
return;
default:
exhaustiveCheck(myEnum);
}
}
versus
function doLetterThing(action: BaseAction) {
action.doAction();
}
If I jump to definition for BaseAction or doAction I will end up on the abstract class, which doesn't help me debug the function or the implementation. If you have a dependency injection pattern with only a single class, this means that you can "guess" by going to the main class / function and looking for how "BaseAction" is instantiated and following that type to the place and scrolling to find the implementation. This seems generally like a bad UX for a developer though.
(small note about whether dependency injection is good, traits seem to do a good enough job in cases where they are necessary (though either done prematurely as a rule rather than as a necessity seems to lead to more difficult to follow code))
Write less code
This depends, but if have to define an extra abstract class for your base type, plus override all the function types, how is that less code than single line switch cases? With good types here if you add an option to the enum, your type checker will flag all the places you need to handle this which will usually involve adding 1 line each for the case and 1+ line for implementation. Compare this with polymorphic classes which you need to define a new class, which needs the new function syntax with the correct params and the opening and closing parens. In most cases, switch cases have less code and less lines.
Colocation
Everything for a type is in one place which is nice, but generally whenever I implement a function like this is I look for a similarly implemented function. With a switch case, it's extremely adjacent, with a derived class I would need to find and locate in another file or directory.
If I implemented a feature change such as trimming spaces off the ends of a string for one type, I would need to open all the class files to make sure if they implement something similar that it is implemented correctly in all of them. And if I forget, I might have different behaviour for different types without knowing. With a switch the co location makes this extremely obvious (though not foolproof)
Conclusion
Am I missing something? It doesn't make sense that we have these clear design principles that I basically can only find affirmative articles about but don't see any clear benefits, and serious downsides compared to some basic pattern matching style development
Consider the solid-principles, in particular OCP and DI.
To extend a switch case or enum and add new functionality in the future, you must modify the existing code. Modifying legacy code is risky and expensive. Risky because you may inadvertently introduce regression. Expensive because you have to learn (or re-learn) implementation details, and then re-test the legacy code (which presumably was working before you modified it).
Dependency on concrete implementations creates tight coupling and inhibits modularity. This makes code rigid and fragile, because a change in one place affects many dependents.
In addition, consider scalability. An abstraction supports any number of implementations, many of which are potentially unknown at the time the abstraction is created. A developer needn't understand or care about additional implementations. How many cases can a developer juggle in one switch, 10? 100?
Note this does not mean polymorphism (or OOP) is suitable for every class or application. For example, there are counterpoints in, Should every class implement an interface? When considering extensibility and scalability, there is an assumption that a code base will grow over time. If you're working with a few thousand lines of code, "enterprise-level" standards are going to feel very heavy. Likewise, coupling a few classes together when you only have a few classes won't be very noticeable.
Benefits of good design are realized years down the road when code is able to evolve in new directions.
I think you are missing the point. The main purpose of having a clean code is not to make your life easier while implementing the current feature, rather it makes your life easier in future when you are extending or maintaining the code.
In your example, you may feel implementing your two actions using switch case. But what happens if you need to add more actions in future? Using the abstract class, you can easily create a new action type and the caller doesn't need to be modified. But if you keep using switch case it will be lot more messier, especially for complex cases.
Also, following a better design pattern (DI in this case) will make the code easier to test. When you consider only easy cases, you may not find the usefulness of using proper design patterns. But if you think broader aspect, it really pays off.
"Base class" is against the Clean Code. There should not be a "Base class", not just for bad naming, also for composition over inheritance rule. So from now on, I will assume it is an interface in which other classes implement it, not extend (which is important for my example). First of all, I would like to see your concerns:
Answer for Concerns
This depends, but if have to define an extra abstract class for your
base type, plus override all the function types, how is that less code
than single line switch cases
I think "write less code" should not be character count. Then Ruby or GoLang or even Python beats the Java, obviously does not it? So I would not count the lines, parenthesis etc. instead code that you should test/maintain.
Everything for a type is in one place which is nice, but generally
whenever I implement a function like this is I look for a similarly
implemented function.
If "look for a similarly" means, having implementation together makes copy some parts from the similar function then we also have some clue here for refactoring. Having Implementation class differently has its own reason; their implementation is completely different. They may follow some pattern, lets see from Communication perspective; If we have Letter and Phone implementations, we should not need to look their implementation to implement one of them. So your assumption is wrong here, if you look to their code to implement new feature then your interface does not guide you for the new feature. Let's be more specific;
interface Communication {
sendMessage()
}
Letter implements Communication {
sendMessage() {
// get receiver
// get sender
// set message
// send message
}
}
Now we need Phone, so if we go to Letter implementation to get and idea to how to implement Phone then our interface does not enough for us to guide our implementation. Technically Phone and Letter is different to send a message. Then we need a Design pattern here, maybe Template Pattern? Let's see;
interface Communication {
default sendMessage() {
getMessageFactory().sendMessage(getSender(), getReceiver(), getBody())
}
getSender()
getReceiver()
getBody()
}
Letter implements Communication {
getSender() { returns sender }
getReceiver() {returns receiver }
getBody() {returns body}
getMessageFactory {returns LetterMessageFactory}
}
Now when we need to implement Phone we don't need to look the details of other implementations. We exactly now what we need to return and also our Communication interface's default method handles how to send the message.
If I implemented a feature change such as trimming spaces off the ends
of a string for one type, I would need to open all the class files to
make sure if they implement something similar that it is implemented
correctly in all of them...
So if there is a "feature change" it should be only its implemented class, not in all classes. You should not change all of the implementations. Or if it is same implementation in all of them, then why each implements it differently? It should be kept as the default method in their interface. Then if feature change required, only default method is changed and you should update your implementation and test in one place.
These are the main points that I wanted to answer your concerns. But I think the main point is you don't get the benefit. I was also struggling before I work on a big project that other teams need to extend my features. I will divide benefits to topics with extreme examples which may be more helpful to understand:
Easy to read
Normally when you see a function, you should not feel to go its implementation to understand what is happening there. It should be self-explanatory. Based on this fact; action.doAction(); -> or lets say communication.sendMessage() if they implement Communicate interface. I don't need to go for its base class, search for implementations etc. for debugging. Even implementing class is "Letter" or "Phone" I know that they send message, I don't need their implementation details. So I don't want to see all implemented classes like in your example "switch Letter; Phone.." etc. In your example doLetterThing responsible for one thing (doAction), since all of them do same thing, then why you are showing your developer all these cases?. They are just making the code harder to read.
Easy to extend
Imagine that you are extending a big project where you don't have an access to their source(I want to give extreme example to show its benefit easier). In the java world, I can say you are implementing SPI (Service Provider Interface). I can show you 2 example for this, https://github.com/apereo/cas and https://github.com/keycloak/keycloak where you can see that interface and implementations are separated and you just implement new behavior when it is required, no need to touch the original source. Why this is important? Imagine the following scenario again;
Let's suppose that Keycloak calls communication.sendMessage(). They don't know implementations in build time. If you extend Keycloak in this case, you can have your own class that implements Communication interface, let's say "Computer". Know if you have your SPI in the classpath, Keycloak reads it and calls your computer.sendMessage(). We did not touch the source code but extended the capabilities of Message Handler class. We can't achieve this if we coded against switch cases without touching the source.
I have problem in understanding Single Responsibility Principle . Should SRP be applied at class level or at method level.
Lets say i have Student Class ,i need to create student , update student and delete student.
If I create a service class that has methods for these three actions does this break SRP principle.
SRP is at both at class and method level.So if you ar talking about student class then only responsibility it has in this case to do CRUD on student entity.At the same time when you talk about methods the you should not have an InsertStudent method and do both Update and Insert in it based on ID .That breaks SRP.But if you have InsertStudent which inserts and UpdateStudent which updates it follows SRP
I'd say you have a service class which is responsible for CRUD operations on objects of type Student. I don't see this design to violate SRP at all.
Quoting from http://www.developerfusion.com/article/137636/taking-the-single-responsibility-principle-seriously/
Two methods of the same class (or even different classes) should focus on different aspects. However, two methods in the same class, e.g. a repository, likely will both have to be concerned with the same higher level responsibility, e.g. persistence.
I see CRUD as well-known operations within a single context unless you have some business associated with it. For example you might want to allow some classes to only be able to read data and deny them from making any changes to it. That's when you can make use of another SOLID principle Interface segregation.
You can define an interface with only read method defined to be used in those classes. Or if it makes sense (performance-wise for example), create a separate concrete class that just implements read operation.
Not to criticize because I believe in the principal, but don't follow the advice that says it fits if you can summarize the functionality without using "and". With this kind of logic you could still have an enormous one file application and say its responsibility without using "and". A web browser is a complicated piece of software but you can still describe it in one short statement. And it makes total sense because the thing is like a pyramid and you should always be able to describe the top level regardless of the parts being split or not.
That is precisely what we do everyday with functions. You pick a very simple function name which hides the complexity like "connect" for a socket. You actually don't know from this point of view if it is split afterwards. It could be a giant function.
I am afraid it is still subjective. You don't want to judge your design based on your ability to summarize a functionality with words. You always should be because this is how you pick method names and we all know naming is hard.
My advice is to see the SOLID principals as a whole instead of individual rules and build separation around what you think is going to change, and what is less likely to change. The obvious candidate being dependency. It is still going to be subjective, there is no way around that, but it'll help you.
I personally find it very difficult to do at times, but it is worth it. I don't know if you know Ecto which is an elixir project, but I had a "Voilà" moment when I've discovered it. It is not perfect is many ways, but the thing with Ecto and separation of concerns in general is that it seems a lot of indirections at first, but then the things are separated make sense. In its best blissful moments, it feels like a lot of small parts that you can trust.
I used to be in the position where it made sense to me that a model should be so smart that it knows how to save itself to the database, how to validates itself, how to do all sorts of things. But the reality is that as soon as you decide you want to work with another database, or validates differently depending on cases, etc, then it becomes hard to get your way out of this. I am sure some developer never felt this way and it is then fine. But for me it is a challenge.
Lots of simple cases, but you want each class to know as less as possible. You don't want your Mail class to know that the css colour for "urgent" is "#FF0000". And then harder ones like sometimes you don't even want it to know it is "urgent" because it depends on use case facts.
This is not easy. In you specific case, I personally would not bother mixing "create" and "delete" for example, but I would make sure interacting with the database is all it does. It does not know if the thing is valid, if it has callbacks, etc. Pretty much the Repository pattern. Again Ecto is a good example, or at least I find it helpful.
I am looking for the correct design pattern to use in the following situation:
I have a process which runs and during this process I need to attach several properties to an object in the system. The object is of the same type but at runtime it might exhibit slightly different behaviour and therefore the way the properties are set can be different depending on the type.
Regardless of the type and behaviour of these objects I want to set the same properties on each.
I then need an object to parse these properties at another point in the process
What is the best way to approach this?
I would suggest you not try to pick a design pattern before coding. First, write the code. Then, start abstracting any redundant code, or code that varies. To understand abstracting code that varies, read Head First Design Patterns. At the very beginning of that book is an example of abstracting what varies, using the strategy pattern. The SimUDuck example there is one of the best-explained examples I've ever seen of the strategy pattern. It sounds like that's what you're asking about. However, your question doesn't have a concrete example of what you're trying to do, so giving a concrete example is difficult here.
Having said that, it sounds like you need good, ol' fashioned polymorphism here: you need to treat all objects the same way, and set the same properties, just with different values. To do this, create an interface and have all of your different types implement that interface. Then, in the calling/consuming code, deal with each of those concrete types as the interface.
If you try to pick a design pattern first, many times you'll end up finding that things change based on the details of the implementation, and your original guess at a design pattern ends up being the wrong fit. Then you end up coding to meet a design pattern and not solving the real problem. Write the code first, even if it's ugly. Get it working. Then find areas to abstract and it will naturally evolve into a design pattern on its own.
If i properly understand, you want add behaviours in runtime???
If yes, so i think - decorator (aka wrapper) design pattern can be good.
In the SRP, a 'responsibility' is usually described as 'a reason to change', so that each class (or object?) should have only one reason someone should have to go in there and change it.
But if you take this to the extreme fine-grain you could say that an object adding two numbers together is a responsibility and a possible reason to change. Therefore the object should contain no other logic, because it would produce another reason for change.
I'm curious if there is anyone out there that has any strategies for 'scoping', the single-responsibility principle that's slightly less objective?
it comes down to the context of what you are modeling. I've done some extensive writing and presenting on the SOLID principles and I specifically address your question in my discussions of Single Responsibility.
The following first appeared in the Jan/Feb 2010 issue of Code Magazine, and is available online at "S.O.L.I.D. Software Development, One Step at a Time"
The Single Responsibility Principle
says that a class should have one, and
only one, reason to change.
This may seem counter-intuitive at
first. Wouldn’t it be easier to say
that a class should only have one
reason to exist? Actually, no-one
reason to exist could very easily be
taken to an extreme that would cause
more harm than good. If you take it to
that extreme and build classes that
have one reason to exist, you may end
up with only one method per class.
This would cause a large sprawl of
classes for even the most simple of
processes, causing the system to be
difficult to understand and difficult
to change.
The reason that a class should have
one reason to change, instead of one
reason to exist, is the business
context in which you are building the
system. Even if two concepts are
logically different, the business
context in which they are needed may
necessitate them becoming one and the
same. The key point of deciding when a
class should change is not based on a
purely logical separation of concepts,
but rather the business’s perception
of the concept. When the business
perception and context has changed,
then you have a reason to change the
class. To understand what
responsibilities a single class should
have, you need to first understand
what concept should be encapsulated by
that class and where you expect the
implementation details of that concept
to change.
Consider an engine in a car, for
example. Do you care about the inner
working of the engine? Do you care
that you have a specific size of
piston, camshaft, fuel injector, etc?
Or, do you only care that the engine
operates as expected when you get in
the car? The answer, of course,
depends entirely on the context in
which you need to use the engine.
If you are a mechanic working in an
auto shop, you probably care about the
inner workings of the engine. You need
to know the specific model, the
various part sizes, and other
specifications of the engine. If you
don’t have this information available,
you likely cannot service the engine
appropriately. However, if you are an
average everyday person that only
needs transportation from point A to
point B, you will likely not need that
level of information. The notion of
the individual pistons, spark plugs,
pulleys, belts, etc., is almost
meaningless to you. You only care that
the car you are driving has an engine
and that it performs correctly.
The engine example drives straight to
the heart of the Single Responsibility
Principle. The contexts of driving the
car vs. servicing the engine provide
two different notions of what should
and should not be a single concept-a
reason for change. In the context of
servicing the engine, every individual
part needs to be separate. You need to
code them as single classes and ensure
they are all up to their individual
specifications. In the context of
driving a car, though, the engine is a
single concept that does not need to
be broken down any further. You would
likely have a single class called
Engine, in this case. In either case,
the context has determined what the
appropriate separation of
responsibilities is.
I tend to think in term of "velocity of change" of the business requirements rather than "reason to change" .
The question is indeed how likely stuffs will change together, not whether they could change or not.
The difference is subtle, but helps me. Let's consider the example on wikipedia about the reporting engine:
if the likelihood that the content and the template of the report change at the same time is high, it can be one component because they are apparently related. (It can also be two)
but if the likelihood that the content change without the template is important, then it must be two components, because they are not related. (Would be dangerous to have one)
But I know that's a personal interpretation of the SRP.
Also, a second technique that I like is: "Describe your class in one sentence". It usually helps me to identify if there is a clear responsibility or not.
I don't see performing a task like adding two numbers together as a responsibility. Responsibilities come in different shapes and sizes but they certainly should be seen as something larger than performing a single function.
To understand this better, it is probably helpful to clearly differentiate between what a class is responsible for and what a method does. A method should "do only one thing" (e.g. add two numbers, though for most purposes '+' is a method that does that already) while a class should present a single clear "responsibility" to it's consumers. It's responsibility is at a much higher level than a method.
A class like Repository has a clear and singular responsibility. It has multiple methods like Save and Load, but a clear responsibility to provide persistence support for Person entities. A class may also co-ordinate and/or abstract the responsibilities of dependent classes, again presenting this as a single responsibility to other consuming classes.
The bottom line is if the application of SRP is leading to single-method classes who's whole purpose seems to be just to wrap the functionality of that method in a class then SRP is not being applied correctly.
A simple rule of thumb I use is that: the level or grainularity of responsibility should match the level or grainularity of the "entity" in question. Obviously the purpose of a method will always be more precise than that of a class, or service, or component.
A good strategiy for evaluating the level of responsibility can be to use an appropriate metaphor. If you can relate what you are doing to something that exists in the real world it can help give you another view of the problem you're trying to solve - including being able to identify appropriate levels of abstraction and responsibility.
#Derick bailey: nice explanation
Some additions: It is totally acceptable that application of SRP is contextual base.
The question still remains: are there any objective ways to define if a given class violates SRP ?
Some design contexts are quite obvious ( like the car example by Derick ) but otherwise contexts in which a class's behaviour has to defined remains fuzzy many-a-times.
For such cases, it might well be helpful if the fuzzy class behaviour is analysed by splitting it's responsibilities into different classes and then measuring the impact of new behavioural and structural relations that has emanated because of the split.
As soon the split is done, the reasons to keep the splitted responsibilities or to back-merge them into single responsibility becomes obvious at once.
I have applied this approach and which has lead good results for me.
But my search to look for 'objective ways of defining a class responsibility' still continues.
I respectful don't agree when Chris Nicola's above says that "a class should presents a single clear "responsibility" to it's consumers
I think SRP is about having a good design inside the class, not class' customers.
To me it's not very clear what a responsability is, and the prove is the number of questions that this concept arises.
"single reason to change"
or
"if the description contains the word
"and" then it needs to be split"
leads to the question: where is the limit? At the end, any class with 2 public methods has 2 reasons to change, isn't it?
For me, the true SRP leads to the Facade pattern, where you have a class that simply delegades the calls to other classes
For example:
class Modem
send()
receive()
Refactors to ==>
class ModemSender
class ModelReceiver
+
class Modem
send() -> ModemSender.send()
receive() -> ModemReceiver.receive()
Opinions are wellcome
I know about "class having a single reason to change". Now, what is that exactly? Are there some smells/signs that could tell that class does not have a single responsibility? Or could the real answer hide in YAGNI and only refactor to a single responsibility the first time your class changes?
The Single Responsibility Principle
There are many obvious cases, e.g. CoffeeAndSoupFactory. Coffee and soup in the same appliance can lead to quite distasteful results. In this example, the appliance might be broken into a HotWaterGenerator and some kind of Stirrer. Then a new CoffeeFactory and SoupFactory can be built from those components and any accidental mixing can be avoided.
Among the more subtle cases, the tension between data access objects (DAOs) and data transfer objects (DTOs) is very common. DAOs talk to the database, DTOs are serializable for transfer between processes and machines. Usually DAOs need a reference to your database framework, therefore they are unusable on your rich clients which neither have the database drivers installed nor have the necessary privileges to access the DB.
Code Smells
The methods in a class start to be grouped by areas of functionality ("these are the Coffee methods and these are the Soup methods").
Implementing many interfaces.
Write a brief, but accurate description of what the class does.
If the description contains the word "and" then it needs to be split.
Well, this principle is to be used with some salt... to avoid class explosion.
A single responsibility does not translate to single method classes. It means a single reason for existence... a service that the object provides for its clients.
A nice way to stay on the road... Use the object as person metaphor... If the object were a person, who would I ask to do this? Assign that responsibility to the corresponding class. However you wouldn't ask the same person to do your manage files, compute salaries, issue paychecks, and verify financial records... Why would you want a single object to do all these? (it's okay if a class takes on multiple responsibilities as long as they are all related and coherent.)
If you employ a CRC card, it's a nice subtle guideline. If you're having trouble getting all the responsibilities of that object on a CRC card, it's probably doing too much... a max of 7 would do as a good marker.
Another code smell from the refactoring book would be HUGE classes. Shotgun surgery would be another... making a change to one area in a class causes bugs in unrelated areas of the same class...
Finding that you are making changes to the same class for unrelated bug-fixes again and again is another indication that the class is doing too much.
A simple and practical method to check single responsibility (not only classes but also method of classes) is the name choice. When you design a class, if you easily find a name for the class that specify exactly what it defines, you're in the right way.
A difficulty to choose a name is near always a symptom of bad design.
the methods in your class should be cohesive...they should work together and make use of the same data structures internally. If you find you have too many methods that don't seem entirely well related, or seem to operate on different things, then quite likely you don't have a good single responsibility.
Often it's hard to initially find responsibilities, and sometimes you need to use the class in several different contexts and then refactor the class into two classes as you start to see the distinctions. Sometimes you find that it's because you are mixing an abstract and concrete concept together. They tend to be harder to see, and, again, use in different contexts will help clarify.
The obvious sign is when your class ends up looking like a Big Ball of Mud, which is really the opposite of SRP (single responsibility principle).
Basically, all the object's services should be focused on carrying out a single responsibility, meaning every time your class changes and adds a service which does not respect that, you know you're "deviating" from the "right" path ;)
The cause is usually due to some quick fixes hastily added to the class to repair some defects. So the reason why you are changing the class is usually the best criteria to detect if you are about to break the SRP.
Martin's Agile Principles, Patterns, and Practices in C# helped me a lot to grasp SRP. He defines SRP as:
A class should have only one reason to change.
So what is driving change?
Martin's answer is:
[...] each responsibility is an axis of change. (p. 116)
and further:
In the context of the SRP, we define a responsibility to be a reason for change. If you can think of more than one motive for changing a class, that class has more than one responsibility (p. 117)
In fact SRP is encapsulating change. If change happens, it should just have a local impact.
Where is YAGNI?
YAGNI can be nicely combined with SRP: When you apply YAGNI, you wait until some change is actually happening. If this happens you should be able to clearly see the responsibilities which are inferred from the reason(s) for change.
This also means that responsibilities can evolve with each new requirement and change. Thinking further SRP and YAGNI will provide you the means to think in flexible designs and architectures.
Perhaps a little more technical than other smells:
If you find you need several "friend" classes or functions, that's usually a good smell of bad SRP - because the required functionality is not actually exposed publically by your class.
If you end up with an excessively "deep" hierarchy (a long list of derived classes until you get to leaf classes) or "broad" hierarchy (many, many classes derived shallowly from a single parent class). It's usually a sign that the parent class does either too much or too little. Doing nothing is the limit of that, and yes, I have seen that in practice, with an "empty" parent class definition just to group together a bunch of unrelated classes in a single hierarchy.
I also find that refactoring to single responsibility is hard. By the time you finally get around to it, the different responsibilities of the class will have become entwined in the client code making it hard to factor one thing out without breaking the other thing. I'd rather err on the side of "too little" than "too much" myself.
Here are some things that help me figure out if my class is violating SRP:
Fill out the XML doc comments on a class. If you use words like if, and, but, except, when, etc., your classes probably is doing too much.
If your class is a domain service, it should have a verb in the name. Many times you have classes like "OrderService", which should probably be broken up into "GetOrderService", "SaveOrderService", "SubmitOrderService", etc.
If you end up with MethodA that uses MemberA and MethodB that uses MemberB and it is not part of some concurrency or versioning scheme, you might be violating SRP.
If you notice that you have a class that just delegates calls to a lot of other classes, you might be stuck in proxy class hell. This is especially true if you end up instantiating the proxy class everywhere when you could just use the specific classes directly. I have seen a lot of this. Think ProgramNameBL and ProgramNameDAL classes as a substitute for using a Repository pattern.
I've also been trying to get my head around the SOLID principles of OOD, specifically the single responsibility principle, aka SRP (as a side note the podcast with Jeff Atwood, Joel Spolsky and "Uncle Bob" is worth a listen). The big question for me is: What problems is SOLID trying to address?
OOP is all about modeling. The main purpose of modeling is to present a problem in a way that allows us to understand it and solve it. Modeling forces us to focus on the important details. At the same time we can use encapsulation to hide the "unimportant" details so that we only have to deal with them when absolutely necessary.
I guess you should ask yourself: What problem is your class trying to solve? Has the important information you need to solve this problem risen to the surface? Are the unimportant details tucked away so that you only have to think about them when absolutely necessary?
Thinking about these things results in programs that are easier to understand, maintain and extend. I think this is at the heart of OOD and the SOLID principles, including SRP.
Another rule of thumb I'd like to throw in is the following:
If you feel the need to either write some sort of cartesian product of cases in your test cases, or if you want to mock certain private methods of the class, Single Responsibility is violated.
I recently had this in the following way:
I had a cetain abstract syntax tree of a coroutine which will be generated into C later. For now, think of the nodes as Sequence, Iteration and Action. Sequence chains two coroutines, Iteration repeats a coroutine until a userdefined condition is true and Action performs a certain userdefined action. Furthermore, it is possible to annotate Actions and Iterations with codeblocks, which define the actions and conditions to evaluate as the coroutine walks ahead.
It was necessary to apply a certain transformation to all of these code blocks (for those interested: I needed to replace the conceptual user variables with actual implementation variables in order to prevent variable clashes. Those who know lisp macros can think of gensym in action :) ). Thus, the simplest thing that would work was a visitor which knows the operation internally and just calls them on the annotated code block of the Action and Iteration on visit and traverses all the syntax tree nodes. However, in this case, I'd have had to duplicate the assertion "transformation is applied" in my testcode for the visitAction-Method and the visitIteration-Method. In other words, I had to check the product test cases of the responsibilities Traversion (== {traverse iteration, traverse action, traverse sequence}) x Transformation (well, codeblock transformed, which blew up into iteration transformed and action transformed). Thus, I was tempted to use powermock to remove the transformation-Method and replace it with some 'return "I was transformed!";'-Stub.
However, according to the rule of thumb, I split the class into a class TreeModifier which contains a NodeModifier-instance, which provides methods modifyIteration, modifySequence, modifyCodeblock and so on. Thus, I could easily test the responsibility of traversing, calling the NodeModifier and reconstructing the tree and test the actual modification of the code blocks separately, thus removing the need for the product tests, because the responsibilities were separated now (into traversing and reconstructing and the concrete modification).
It also is interesting to notice that later on, I could heavily reuse the TreeModifier in various other transformations. :)
If you're finding troubles extending the functionality of the class without being afraid that you might end up breaking something else, or you cannot use class without modifying tons of its options which modify its behavior smells like your class doing too much.
Once I was working with the legacy class which had method "ZipAndClean", which was obviously zipping and cleaning specified folder...