How to avoid “if” chains?

问题:

Assuming I have this pseudo-code:

bool conditionA = executeStepA();
if (conditionA){
    bool conditionB = executeStepB();
    if (conditionB){
        bool conditionC = executeStepC();
        if (conditionC){
            ...
        }
    }
}

executeThisFunctionInAnyCase();

Functions executeStepX should be executed if and only if the previous succeed. In any case, the executeThisFunctionInAnyCase function should be called at the end. I'm a newbie in programming, so sorry for the very basic question: is there a way (in C/C++ for example) to avoid that long if chain producing that sort of "pyramid of code", at the expense of the code legibility?

I know that if we could skip the executeThisFunctionInAnyCase function call, the code could be simplified as:

bool conditionA = executeStepA();
if (!conditionA) return;
bool conditionB = executeStepB();
if (!conditionB) return;
bool conditionC = executeStepC();
if (!conditionC) return;

But the constraint is the executeThisFunctionInAnyCase function call. Could the break statement be used in some way?

回答1:

You can use an && (logic AND):

if (executeStepA() && executeStepB() && executeStepC()){
    ...
}
executeThisFunctionInAnyCase();

this will satisfy both of your requirements:

• executeStep<X>() should evaluate only if the previous one succeeded (this is called short circuit evaluation)
• executeThisFunctionInAnyCase() will be executed in any case

回答2:

Just use an additional function to get your second version to work:

void foo()
{
  bool conditionA = executeStepA();
  if (!conditionA) return;

  bool conditionB = executeStepB();
  if (!conditionB) return;

  bool conditionC = executeStepC();
  if (!conditionC) return;
}

void bar()
{
  foo();
  executeThisFunctionInAnyCase();
}

Using either deeply nested ifs (your first variant) or the desire to break out of "part of a function" usually means you do need an extra function.

回答3:

Old school C programmers use goto in this case. It is the one usage of goto that's actually encouraged by the Linux styleguide, it's called the centralized function exit:

int foo() {
    int result = /*some error code*/;
    if(!executeStepA()) goto cleanup;
    if(!executeStepB()) goto cleanup;
    if(!executeStepC()) goto cleanup;

    result = 0;
cleanup:
    executeThisFunctionInAnyCase();
    return result;
}

Some people work around using goto by wrapping the body into a loop and breaking from it, but effectively both approaches do the same thing. The goto approach is better if you need some other cleanup only if executeStepA() was successfull:

int foo() {
    int result = /*some error code*/;
    if(!executeStepA()) goto cleanupPart;
    if(!executeStepB()) goto cleanup;
    if(!executeStepC()) goto cleanup;

    result = 0;
cleanup:
    innerCleanup();
cleanupPart:
    executeThisFunctionInAnyCase();
    return result;
}

With the loop approach you would end up with two levels of loops in that case.

回答4:

This is a common situation and there are many common ways to deal with it. Here's my attempt at a canonical answer. Please comment if I missed anything and I'll keep this post up to date.

This is an Arrow

What you are discussing is known as the arrow anti-pattern. It is called an arrow because the chain of nested ifs form code blocks that expand farther and farther to the right and then back to the left, forming a visual arrow that "points" to the right side of the code editor pane.

Flatten the Arrow with the Guard

Some common ways of avoiding the Arrow are discussed here. The most common method is to use a guard pattern, in which the code handles the exception flows first and then handles the basic flow, e.g. instead of

if (ok)
{
    DoSomething();
}
else
{
    _log.Error("oops");
    return;
}

... you'd use....

if (!ok)
{
    _log.Error("oops");
    return;
} 
DoSomething(); //notice how this is already farther to the left than the example above

When there is a long series of guards this flattens the code considerably as all the guards appear all the way to the left and your ifs are not nested. In addition, you are visually pairing the logic condition with its associated error, which makes it far easier to tell what is going on:

Arrow:

ok = DoSomething1();
if (ok)
{
    ok = DoSomething2();
    if (ok)
    {
        ok = DoSomething3();
        if (!ok)
        {
            _log.Error("oops");  //Tip of the Arrow
            return;
        }
    }
    else
    {
       _log.Error("oops");
       return;
    }
}
else
{
    _log.Error("oops");
    return;
}

Guard:

ok = DoSomething1();
if (!ok)
{
    _log.Error("oops");
    return;
} 
ok = DoSomething2();
if (!ok)
{
    _log.Error("oops");
    return;
} 
ok = DoSomething3();
if (!ok)
{
    _log.Error("oops");
    return;
} 
ok = DoSomething4();
if (!ok)
{
    _log.Error("oops");
    return;
} 

This is objectively and quantifiably easier to read because

1. The { and } characters for a given logic block are closer together
2. The amount of mental context needed to understand a particular line is smaller
3. The entirety of logic associated with an if condition is more likely to be on one page
4. The need for the coder to scroll the page/eye track is greatly lessened

How to add common code at the end

The problem with the guard pattern is that it relies on what is called "opportunistic return" or "opportunistic exit." In other words, it breaks the pattern that each and every function should have exactly one point of exit. This is a problem for two reasons:

1. It rubs some people the wrong way, e.g. people who learned to code on Pascal have learned that one function = one exit point.
2. It does not provide a section of code that executes upon exit no matter what, which is the subject at hand.

Below I've provided some options for working around this limitation either by using language features or by avoiding the problem altogether.

Option 1. You can't do this: use finally

Unfortunately, as a c++ developer, you can't do this. But this is the number one answer for languages that contain a finally keyword, since this is exactly what it is for.

try
{
    if (!ok)
    {
        _log.Error("oops");
        return;
    } 
    DoSomething(); //notice how this is already farther to the left than the example above
}
finally
{
    DoSomethingNoMatterWhat();
}

Option 2. Avoid the issue: Restructure your functions

You can avoid the problem by breaking the code into two functions. This solution has the benefit of working for any language, and additionally it can reduce cyclomatic complexity, which is a proven way to reduce your defect rate, and improves the specificity of any automated unit tests.

Here's an example:

void OuterFunction()
{
    DoSomethingIfPossible();
    DoSomethingNoMatterWhat();
}

void DoSomethingIfPossible()
{
    if (!ok)
    {
        _log.Error("Oops");
        return;
    }
    DoSomething();
}

Option 3. Language trick: Use a fake loop

Another common trick I see is using while(true) and break, as shown in the other answers.

while(true)
{
     if (!ok) break;
     DoSomething();
     break;  //important
}
DoSomethingNoMatterWhat();

While this is less "honest" than using goto, it is less prone to being messed up when refactoring, as it clearly marks the boundaries of logic scope. A naive coder who cuts and pastes your labels or your goto statements can cause major problems! (And frankly the pattern is so common now I think it clearly communicates the intent, and is therefore not "dishonest" at all).

There are other variants of this options. For example, one could use switch instead of while. Any language construct with a break keyword would probably work.

Option 4. Leverage the object life cycle

One other approach leverages the object life cycle. Use a context object to carry around your parameters (something which our naive example suspiciously lacks) and dispose of it when you're done.

class MyContext
{
   ~MyContext()
   {
        DoSomethingNoMatterWhat();
   }
}

void MainMethod()
{
    MyContext myContext;
    ok = DoSomething(myContext);
    if (!ok)
    {
        _log.Error("Oops");
        return;
    }
    ok = DoSomethingElse(myContext);
    if (!ok)
    {
        _log.Error("Oops");
        return;
    }
    ok = DoSomethingMore(myContext);
    if (!ok)
    {
        _log.Error("Oops");
    }

    //DoSomethingNoMatterWhat will be called when myContext goes out of scope
}

Note: Be sure you understand the object life cycle of your language of choice. You need some sort of deterministic garbage collection for this to work, i.e. you have to know when the destructor will be called. In some languages you will need to use Dispose instead of a destructor.

Option 4.1. Leverage the object life cycle (wrapper pattern)

If you're going to use an object-oriented approach, may as well do it right. This option uses a class to "wrap" the resources that require cleanup, as well as its other operations.

class MyWrapper 
{
   bool DoSomething() {...};
   bool DoSomethingElse() {...}


   void ~MyWapper()
   {
        DoSomethingNoMatterWhat();
   }
}

void MainMethod()
{
    bool ok = myWrapper.DoSomething();
    if (!ok)
        _log.Error("Oops");
        return;
    }
    ok = myWrapper.DoSomethingElse();
    if (!ok)
       _log.Error("Oops");
        return;
    }
}
//DoSomethingNoMatterWhat will be called when myWrapper is destroyed

Again, be sure you understand your object life cycle.

Option 5. Language trick: Use short-circuit evaluation

Another technique is to take advantage of short-circuit evaluation.

if (DoSomething1() && DoSomething2() && DoSomething3())
{
    DoSomething4();
}
DoSomethingNoMatterWhat();

This solution takes advantage of the way the && operator works. When the left hand side of && evaluates to false, the right hand side is never evaluated.

This trick is most useful when compact code is required and when the code is not likely to see much maintenance, e.g you are implementing a well-known algorithm. For more general coding the structure of this code is too brittle; even a minor change to the logic could trigger a total rewrite.

回答5:

Just do

if( executeStepA() && executeStepB() && executeStepC() )
{
    // ...
}
executeThisFunctionInAnyCase();

It's that simple.

---

Due to three edits that each has fundamentally changed the question (four if one counts the revision back to version #1), I include the code example I'm answering to:

bool conditionA = executeStepA();
if (conditionA){
    bool conditionB = executeStepB();
    if (conditionB){
        bool conditionC = executeStepC();
        if (conditionC){
            ...
        }
    }
}

executeThisFunctionInAnyCase();

回答6:

There is actually a way to defer actions in C++: making use of an object's destructor.

Assuming that you have access to C++11:

class Defer {
public:
    Defer(std::function<void()> f): f_(std::move(f)) {}
    ~Defer() { if (f_) { f_(); } }

    void cancel() { f_ = std::function<void()>(); }

private:
    Defer(Defer const&) = delete;
    Defer& operator=(Defer const&) = delete;

    std::function<void()> f_;
}; // class Defer

And then using that utility:

int foo() {
    Defer const defer{&executeThisFunctionInAnyCase}; // or a lambda

    // ...

    if (!executeA()) { return 1; }

    // ...

    if (!executeB()) { return 2; }

    // ...

    if (!executeC()) { return 3; }

    // ...

    return 4;
} // foo

回答7:

There's a nice technique which doesn't need an additional wrapper function with the return statements (the method prescribed by Itjax). It makes use of a do while(0) pseudo-loop. The while (0) ensures that it is actually not a loop but executed only once. However, the loop syntax allows the use of the break statement.

void foo()
{
  // ...
  do {
      if (!executeStepA())
          break;
      if (!executeStepB())
          break;
      if (!executeStepC())
          break;
  }
  while (0);
  // ...
}

回答8:

You could also do this:

bool isOk = true;
std::vector<bool (*)(void)> funcs; //vector of function ptr

funcs.push_back(&executeStepA);
funcs.push_back(&executeStepB);
funcs.push_back(&executeStepC);
//...

//this will stop at the first false return
for (auto it = funcs.begin(); it != funcs.end() && isOk; ++it) 
    isOk = (*it)();
if (isOk)
 //doSomeStuff
executeThisFunctionInAnyCase();

This way you have a minimal linear growth size, +1 line per call, and it's easily maintenable.

---

EDIT: (Thanks @Unda) Not a big fan because you loose visibility IMO :

bool isOk = true;
auto funcs { //using c++11 initializer_list
    &executeStepA,
    &executeStepB,
    &executeStepC
};

for (auto it = funcs.begin(); it != funcs.end() && isOk; ++it) 
    isOk = (*it)();
if (isOk)
 //doSomeStuff
executeThisFunctionInAnyCase();

回答9:

Would this work? I think this is equivalent with your code.

bool condition = true; // using only one boolean variable
if (condition) condition = executeStepA();
if (condition) condition = executeStepB();
if (condition) condition = executeStepC();
...
executeThisFunctionInAnyCase();

回答10:

Assuming the desired code is as I currently see it:

bool conditionA = executeStepA();
if (conditionA){
    bool conditionB = executeStepB();
    if (conditionB){
        bool conditionC = executeStepC();
        if (conditionC){
            ...
        }
    }
}    
executeThisFunctionInAnyCase();

I would say that the correct approach, in that it's the simplest to read and easiest to maintain, would have fewer levels of indentation, which is (currently) the stated purpose of the question.

// Pre-declare the variables for the conditions
bool conditionA = false;
bool conditionB = false;
bool conditionC = false;

// Execute each step only if the pre-conditions are met
conditionA = executeStepA();
if (conditionA)
    conditionB = executeStepB();
if (conditionB)
    conditionC = executeStepC();
if (conditionC) {
    ...
}

// Unconditionally execute the 'cleanup' part.
executeThisFunctionInAnyCase();

This avoids any need for gotos, exceptions, dummy while loops, or other difficult constructs and simply gets on with the simple job at hand.

回答11:

Could break statement be used in some way?

Maybe not the best solution but you can put your statements in a do .. while (0) loop and use break statements instead of return.

回答12:

You could put all the if conditions, formatted as you want it in a function of their own, the on return execute the executeThisFunctionInAnyCase() function.

From the base example in the OP, the condition testing and execution can be split off as such;

void InitialSteps()
{
  bool conditionA = executeStepA();
  if (!conditionA)
    return;
  bool conditionB = executeStepB();
  if (!conditionB)
    return;
  bool conditionC = executeStepC();
  if (!conditionC)
    return;
}

And then called as such;

InitialSteps();
executeThisFunctionInAnyCase();

If C++11 lambdas are available (there was no C++11 tag in the OP, but they may still be an option), then we can forgo the seperate function and wrap this up into a lambda.

// Capture by reference (variable access may be required)
auto initialSteps = [&]() {
  // any additional code
  bool conditionA = executeStepA();
  if (!conditionA)
    return;
  // any additional code
  bool conditionB = executeStepB();
  if (!conditionB)
    return;
  // any additional code
  bool conditionC = executeStepC();
  if (!conditionC)
    return;
};

initialSteps();
executeThisFunctionInAnyCase();

回答13:

If you dislike goto and dislike do { } while (0); loops and like to use C++ you can also use a temporary lambda to have the same effect.

[&]() { // create a capture all lambda
  if (!executeStepA()) { return; }
  if (!executeStepB()) { return; }
  if (!executeStepC()) { return; }
}(); // and immediately call it

executeThisFunctionInAnyCase();

回答14:

The chains of IF/ELSE in your code is not the language issue, but the design of your program. If you're able to re-factor or re-write your program I'd like to suggest that you look in Design Patterns (http://sourcemaking.com/design_patterns) to find a better solution.

Usually, when you see a lot of IF's & else's in your code , it is an opportunity to implement the Strategy Design Pattern (http://sourcemaking.com/design_patterns/strategy/c-sharp-dot-net) or maybe a combination of other patterns.

I'm sure there're alternatives to write a long list of if/else , but I doubt they will change anything except that the chain will look pretty to you (However, the beauty is in the eye of the beholder still applies to code too:-) ) . You should be concerned about things like (in 6 months when I have a new condition and I don't remember anything about this code , will I be able to add it easily? Or what if the chain changes, how quickly and error-free will I be implement it)

回答15:

You just do this..

coverConditions();
executeThisFunctionInAnyCase();

function coverConditions()
 {
 bool conditionA = executeStepA();
 if (!conditionA) return;
 bool conditionB = executeStepB();
 if (!conditionB) return;
 bool conditionC = executeStepC();
 if (!conditionC) return;
 }

99 times of 100, this is the only way to do it.

Never, ever, ever try to do something "tricky" in computer code.

---

By the way, I'm pretty sure the following is the actual solution you had in mind...

The continue statement is critical in algorithmic programming. (Much as, the goto statement is critical in algorithmic programming.)

In many programming languages you can do this:

-(void)_testKode
    {
    NSLog(@"code a");
    NSLog(@"code b");
    NSLog(@"code c\n");

    int x = 69;

    {

    if ( x == 13 )
        {
        NSLog(@"code d---\n");
        continue;
        }

    if ( x == 69 )
        {
        NSLog(@"code e---\n");
        continue;
        }

    if ( x == 13 )
        {
        NSLog(@"code f---\n");
        continue;
        }

    }

    NSLog(@"code g");
    }

(Note first of all: naked blocks like that example are a critical and important part of writing beautiful code, particularly if you are dealing with "algorithmic" programming.)

Again, that's exactly what you had in your head, right? And that's the beautiful way to write it, so you have good instincts.

However, tragically, in the current version of objective-c (Aside - I don't know about Swift, sorry) there is a risible feature where it checks if the enclosing block is a loop.

Here's how you get around that...

-(void)_testKode
    {
    NSLog(@"code a");
    NSLog(@"code b");
    NSLog(@"code c\n");

    int x = 69;

    do{

    if ( x == 13 )
        {
        NSLog(@"code d---\n");
        continue;
        }

    if ( x == 69 )
        {
        NSLog(@"code e---\n");
        continue;
        }

    if ( x == 13 )
        {
        NSLog(@"code f---\n");
        continue;
        }

    }while(false);

    NSLog(@"code g");
    }

So don't forget that ..

do { } while(false);

just means "do this block once".

ie, there is utterly no difference between writing do{}while(false); and simply writing {} .

This now works perfectly as you wanted...here's the output...

So, it's possible that's how you see the algorithm in your head. You should always try to write what's in your head. ( Particularly if you are not sober, because that's when the pretty comes out! :) )

In "algorithmic" projects where this happens a lot, in objective-c, we always have a macro like...

#define RUNONCE while(false)

... so then you can do this ...

-(void)_testKode
    {
    NSLog(@"code a");
    int x = 69;

    do{
    if ( x == 13 )
        {
        NSLog(@"code d---\n");
        continue;
        }
    if ( x == 69 )
        {
        NSLog(@"code e---\n");
        continue;
        }
    if ( x == 13 )
        {
        NSLog(@"code f---\n");
        continue;
        }
    }RUNONCE

    NSLog(@"code g");
    }

There are two points:

a, even though it's stupid that objective-c checks the type of block a continue statement is in, it's troubling to "fight that". So it's a tough decision.

b, there's the question should you indent, in the example, that block? I lose sleep over questions like that, so I can't advise.

Hope it helps.

回答16:

Have your execute functions throw an exception if they fail instead of returning false. Then your calling code could look like this:

try {
    executeStepA();
    executeStepB();
    executeStepC();
}
catch (...)

Of course I'm assuming that in your original example the execution step would only return false in the case of an error occuring inside the step?

回答17:

A lot of good answers already, but most of them seem to tradeoff on some (admittedly very little) of the flexibility. A common approach which doesn't require this tradeoff is adding a status/keep-going variable. The price is, of course, one extra value to keep track of:

bool ok = true;
bool conditionA = executeStepA();
// ... possibly edit conditionA, or just ok &= executeStepA();
ok &= conditionA;

if (ok) {
    bool conditionB = executeStepB();
    // ... possibly do more stuff
    ok &= conditionB;
}
if (ok) {
    bool conditionC = executeStepC();
    ok &= conditionC;
}
if (ok && additionalCondition) {
    // ...
}

executeThisFunctionInAnyCase();
// can now also:
return ok;

回答18:

In C++ (the question is tagged both C and C++), if you can't change the functions to use exceptions, you still can use the exception mechanism if you write a little helper function like

struct function_failed {};
void attempt(bool retval)
{
  if (!retval)
    throw function_failed(); // or a more specific exception class
}

Then your code could read as follows:

try
{
  attempt(executeStepA());
  attempt(executeStepB());
  attempt(executeStepC());
}
catch (function_failed)
{
  // -- this block intentionally left empty --
}

executeThisFunctionInAnyCase();

If you're into fancy syntax, you could instead make it work via explicit cast:

struct function_failed {};
struct attempt
{
  attempt(bool retval)
  {
    if (!retval)
      throw function_failed();
  }
};

Then you can write your code as

try
{
  (attempt) executeStepA();
  (attempt) executeStepB();
  (attempt) executeStepC();
}
catch (function_failed)
{
  // -- this block intentionally left empty --
}

executeThisFunctionInAnyCase();

回答19:

If your code is as simple as your example and your language supports short-circuit evaluations, you could try this:

StepA() && StepB() && StepC() && StepD();
DoAlways();

If you are passing arguments to your functions and getting back other results so that your code cannot be written in the previous fashion, many of the other answers would be better suited to the problem.

回答20:

For C++11 and beyond, a nice approach might be to implement a scope exit system similar to D's scope(exit) mechanism.

One possible way to implement it is using C++11 lambdas and some helper macros:

template<typename F> struct ScopeExit 
{
    ScopeExit(F f) : fn(f) { }
    ~ScopeExit() 
    { 
         fn();
    }

    F fn;
};

template<typename F> ScopeExit<F> MakeScopeExit(F f) { return ScopeExit<F>(f); };

#define STR_APPEND2_HELPER(x, y) x##y
#define STR_APPEND2(x, y) STR_APPEND2_HELPER(x, y)

#define SCOPE_EXIT(code)\
    auto STR_APPEND2(scope_exit_, __LINE__) = MakeScopeExit([&](){ code })

This will allow you to return early from the function and ensure whatever cleanup code you define is always executed upon scope exit:

SCOPE_EXIT(
    delete pointerA;
    delete pointerB;
    close(fileC); );

if (!executeStepA())
    return;

if (!executeStepB())
    return;

if (!executeStepC())
    return;

The macros are really just decoration. MakeScopeExit() can be used directly.

回答21:

Why is nobody giving the simplest solution ? :D

If all your functions have the same signature then you can do it this way (for C language):

bool (*step[])() = {
    &executeStepA,
    &executeStepB,
    &executeStepC,
    ... 
};

for (int i = 0; i < numberOfSteps; i++) {
    bool condition = step[i]();

    if (!condition) {
        break;
    }
}

executeThisFunctionInAnyCase();

For a clean C++ solution, you should create an interface class that contains an execute method and wraps your steps in objects.
Then, the solution above will look like this:

Step *steps[] = {
    stepA,
    stepB,
    stepC,
    ... 
};

for (int i = 0; i < numberOfSteps; i++) {
    Step *step = steps[i];

    if (!step->execute()) {
        break;
    }
}

executeThisFunctionInAnyCase();

回答22:

Assuming you don't need individual condition variables, inverting the tests and using the else-falthrough as the "ok" path would allow you do get a more vertical set of if/else statements:

bool failed = false;

// keep going if we don't fail
if (failed = !executeStepA())      {}
else if (failed = !executeStepB()) {}
else if (failed = !executeStepC()) {}
else if (failed = !executeStepD()) {}

runThisFunctionInAnyCase();

Omitting the failed variable makes the code a bit too obscure IMO.

Declaring the variables inside is fine, no worry about = vs ==.

// keep going if we don't fail
if (bool failA = !executeStepA())      {}
else if (bool failB = !executeStepB()) {}
else if (bool failC = !executeStepC()) {}
else if (bool failD = !executeStepD()) {}
else {
     // success !
}

runThisFunctionInAnyCase();

This is obscure, but compact:

// keep going if we don't fail
if (!executeStepA())      {}
else if (!executeStepB()) {}
else if (!executeStepC()) {}
else if (!executeStepD()) {}
else { /* success */ }

runThisFunctionInAnyCase();

回答23:

This looks like a state machine, which is handy because you can easily implement it with a state-pattern.

In Java it would look something like this:

interface StepState{
public StepState performStep();
}

An implementation would work as follows:

class StepA implements StepState{ 
    public StepState performStep()
     {
         performAction();
         if(condition) return new StepB()
         else return null;
     }
}

And so on. Then you can substitute the big if condition with:

Step toDo = new StepA();
while(toDo != null)
      toDo = toDo.performStep();
executeThisFunctionInAnyCase();

回答24:

As Rommik mentioned, you could apply a design pattern for this, but I would use the Decorator pattern rather than Strategy since you are wanting to chain calls. If the code is simple, then I would go with one of the nicely structured answers to prevent nesting. However, if it is complex or requires dynamic chaining, then the Decorator pattern is a good choice. Here is a yUML class diagram:

Here is a sample LinqPad C# program:

void Main()
{
    IOperation step = new StepC();
    step = new StepB(step);
    step = new StepA(step);
    step.Next();
}

public interface IOperation 
{
    bool Next();
}

public class StepA : IOperation
{
    private IOperation _chain;
    public StepA(IOperation chain=null)
    {
        _chain = chain;
    }

    public bool Next() 
    {
        bool localResult = false;
        //do work
        //...
        // set localResult to success of this work
        // just for this example, hard coding to true
        localResult = true;
        Console.WriteLine("Step A success={0}", localResult);

        //then call next in chain and return
        return (localResult && _chain != null) 
            ? _chain.Next() 
            : true;
    }
}

public class StepB : IOperation
{
    private IOperation _chain;
    public StepB(IOperation chain=null)
    {
        _chain = chain;
    }

    public bool Next() 
    {   
        bool localResult = false;

        //do work
        //...
        // set localResult to success of this work
        // just for this example, hard coding to false, 
            // to show breaking out of the chain
        localResult = false;
        Console.WriteLine("Step B success={0}", localResult);

        //then call next in chain and return
        return (localResult && _chain != null) 
            ? _chain.Next() 
            : true;
    }
}

public class StepC : IOperation
{
    private IOperation _chain;
    public StepC(IOperation chain=null)
    {
        _chain = chain;
    }

    public bool Next() 
    {
        bool localResult = false;
        //do work
        //...
        // set localResult to success of this work
        // just for this example, hard coding to true
        localResult = true;
        Console.WriteLine("Step C success={0}", localResult);
        //then call next in chain and return
        return (localResult && _chain != null) 
            ? _chain.Next() 
            : true;
    }
}

The best book to read on design patterns, IMHO, is Head First Design Patterns.

回答25:

Several answers hinted at a pattern that I saw and used many times, especially in network programming. In network stacks there is often a long sequence of requests, any of which can fail and will stop the process.

The common pattern was to use do { } while (false);

I used a macro for the while(false) to make it do { } once; The common pattern was:

do
{
    bool conditionA = executeStepA();
    if (! conditionA) break;
    bool conditionB = executeStepB();
    if (! conditionB) break;
    // etc.
} while (false);

This pattern was relatively easy to read, and allowed objects to be used that would properly destruct and also avoided multiple returns making stepping and debugging a bit easier.

回答26:

To improve on Mathieu's C++11 answer and avoid the runtime cost incurred through the use of std::function, I would suggest to use the following

template<typename functor>
class deferred final
{
public:
    template<typename functor2>
    explicit deferred(functor2&& f) : f(std::forward<functor2>(f)) {}
    ~deferred() { this->f(); }

private:
    functor f;
};

template<typename functor>
auto defer(functor&& f) -> deferred<typename std::decay<functor>::type>
{
    return deferred<typename std::decay<functor>::type>(std::forward<functor>(f));
}

This simple template class will accept any functor that can be called without any parameters, and does so without any dynamic memory allocations and therefore better conforms to C++'s goal of abstraction without unnecessary overhead. The additional function template is there to simplify use by template parameter deduction (which is not available for class template parameters)

Usage example:

auto guard = defer(executeThisFunctionInAnyCase);
bool conditionA = executeStepA();
if (!conditionA) return;
bool conditionB = executeStepB();
if (!conditionB) return;
bool conditionC = executeStepC();
if (!conditionC) return;

Just as Mathieu's answer this solution is fully exception safe, and executeThisFunctionInAnyCase will be called in all cases. Should executeThisFunctionInAnyCase itself throw, destructors are implicitly marked noexceptand therefore a call to std::terminate would be issued instead of causing an exception to be thrown during stack unwinding.

回答27:

It's seems like you want to do all your call from a single block. As other have proposed it, you should used either a while loop and leave using break or a new function that you can leave with return (may be cleaner).

I personally banish goto, even for function exit. They are harder to spot when debugging.

An elegant alternative that should work for your workflow is to build a function array and iterate on this one.

const int STEP_ARRAY_COUNT = 3;
bool (*stepsArray[])() = {
   executeStepA, executeStepB, executeStepC
};

for (int i=0; i<STEP_ARRAY_COUNT; ++i) {
    if (!stepsArray[i]()) {
        break;
    }
}

executeThisFunctionInAnyCase();

回答28:

an interesting way is to work with exceptions.

try
{
    executeStepA();//function throws an exception on error
    ......
}
catch(...)
{
    //some error handling
}
finally
{
    executeThisFunctionInAnyCase();
}

If you write such code you are going somehow in the wrong direction. I wont see it as "the problem" to have such code, but to have such a messy "architecture".

Tip: discuss those cases with a seasoned developer which you trust ;-)

回答29:

Another approach - do - while loop, even though it was mentioned before there was no example of it which would show how it looks like:

do
{
    if (!executeStepA()) break;
    if (!executeStepB()) break;
    if (!executeStepC()) break;
    ...

    break; // skip the do-while condition :)
}
while (0);

executeThisFunctionInAnyCase();

_{(Well there's already an answer with while loop but do - while loop does not redundantly check for true (at the start) but instead at the end xD (this can be skipped, though)).}

回答30:

Because you also have [...block of code...] between executions, I guess you have memory allocation or object initializations. In this way you have to care about cleaning all you already initialized at exit, and also clean it if you will meet problem and any of functions will return false.

In this case, best what I had in my experience (when I worked with CryptoAPI) was creating small classes, in constructor you initialize your data, in destructor you uninitialize it. Each next function class have to be child of previous function class. If something went wrong - throw exception.

class CondA
{
public:
    CondA() { 
        if (!executeStepA()) 
            throw int(1);
        [Initialize data]
    }
    ~CondA() {        
        [Clean data]
    }
    A* _a;
};

class CondB : public CondA
{
public:
    CondB() { 
        if (!executeStepB()) 
            throw int(2);
        [Initialize data]
    }
    ~CondB() {        
        [Clean data]
    }
    B* _b;
};

class CondC : public CondB
{
public:
    CondC() { 
        if (!executeStepC()) 
            throw int(3);
        [Initialize data]
    }
    ~CondC() {        
        [Clean data]
    }
    C* _c;
};

And then in your code you just need to call:

shared_ptr<CondC> C(nullptr);
try{
    C = make_shared<CondC>();
}
catch(int& e)
{
    //do something
}
if (C != nullptr)
{
   C->a;//work with
   C->b;//work with
   C->c;//work with
}
executeThisFunctionInAnyCase();

I guess it is best solution if every call of ConditionX initialize something, allocs memory and etc. Best to be sure everything will be cleaned.