I have the following code:
#include <iostream>
using namespace std;
class A {
public:
A() { cout << "A::A()" << endl;}
~A() { cout << "A::~A()" << endl; throw "A::exception";}
};
class B {
public:
B() { cout << "B::B()" << endl; throw "B::exception";}
~B() { cout << "B::~B()";}
};
int main() {
try {
cout << "Entering try...catch block" << endl;
A objectA;
B objectB;
cout << "Exiting try...catch block" << endl;
} catch (char const * ex) {
cout << ex << endl;
}
return 0;
}
Now, before stating the question, I would like to point that this code is bad practice (e.g throwing an exception from a constructor will result the object not being fully created, thus the destructor won't be called and it might cause memory leaks or other problems).
Now, The order of the main is this:
Printing "Entering try...catch block".
Calling A's constructor, printing "A::A()"
Calling B's constructor, printing "B::B()", and throws exception.
The exception was thrown, and the line "Exiting try...catch block" will not be printed. The block is exited, so A's destructor is called.
A's destructor prints "A::~A()" and throws another exception.
The second exception (in 5) causes the main to throw an exception, before entering the catch block.
My question is this - is there a way to catch the second exception in the main, without altering the classes A,B?
I have tried to surround both the whole try-catch block and inside the catch block with another try-catch block but that didn't work.
Thanks.
From cppreference.com:
As any other function, a destructor may terminate by throwing an exception [...] however if this destructor happens to be called during stack unwinding, std::terminate is called instead.
So, the attempt to throw an exception from ~A() does not result in a second exception being thrown; it results in the program being terminated. If you need to "catch" this "second exception", you would need to interfere with the termination handler. Or you could find a way to not throw an exception in the destructor. Continuing from cppreference.com:
Although std::uncaught_exception may sometimes be used to detect stack unwinding in progress, it is generally considered bad practice to allow any destructor to terminate by throwing an exception.
You may use set_terminate+longjmp in order to avoid program termination.
#include <iostream>
#include <setjmp.h>
using namespace std;
jmp_buf jmpBuf;
class A {
public:
A() { cout << "A::A()" << endl;}
~A() noexcept(false){ cout << "A::~A()" << endl; throw "A::exception";}
};
class B {
public:
B() { cout << "B::B()" << endl; throw "B::exception";}
~B() { cout << "B::~B()";}
};
int main() {
set_terminate([](){
cout<<"long jump begin" << endl;
longjmp(jmpBuf,1);
});
if(setjmp(jmpBuf)==0) {
try {
cout << "Entering try...catch block" << endl;
A objectA;
B objectB;
cout << "Exiting try...catch block" << endl;
} catch (char const *ex) {
cout << ex << endl;
}
}else{
cout<<"long jump end" << endl;
}
return 0;
}
Related
Does this quote from https://en.cppreference.com/w/cpp/language/function-try-block have a typo?
Reaching the end of a catch clause for a function-try-block on a destructor also automatically rethrows the current exception as if by throw;, but a return statement is allowed.
It seems hard to believe that a destructor automatically re-throws, when every article by every expert I've ever read says that destructors should never, ever throw under any circumstance. In fact, the example code above the quote shows an implicit throw from a constructor, not a destructor.
Therefore, I wonder, is the statement wrong and should have indicated that behavior for a constructor instead?
I had been reviewing this other StackOverflow article when I started thinking about this: C4297 warning in Visual Studio while using function-try-block (function assumed not to throw an exception but does). It already had an answer, but nobody questioned whether the quote was accurate in the first place.
The answer is no, it is not a typo.
The cppreference article did not show an example of a destructor function try block, so I crafted one myself and tested it. Below is the same code. I tested with Microsoft VS2019, using the v142 platform toolset and C++20 dialect.
If you execute this, an abort will be called, which is consistent with the warning that the compiler issues. This suggests that the function catch block does automatically re-throw, even for a destructor. If you uncomment the return statement, it will not throw. Although, I find that to be counter-intuitive, writing a return statement provides a workaround to prevent the implicit throw, just as the referenced StackOverflow article suggests.
#include <iostream>
#include <string>
struct S
{
std::string m;
S(const std::string& str, int idx) try : m(str, idx)
{
std::cout << "S(" << str << ", " << idx << ") constructed, m = " << m << '\n';
}
catch (const std::exception& e)
{
std::cout << "S(" << str << ", " << idx << ") failed: " << e.what() << '\n';
} // implicit "throw;" here
~S() try
{
if (m.length() > 5) {
throw std::exception("shouldn't have been that big!");
}
std::cout << "destroyed!" << std::endl;
}
catch (const std::exception& e)
{
//return;
}
};
int main()
{
S s1{ "ABC", 1 }; // does not throw (index is in bounds)
try
{
S s2{ "ABC", 4 }; // throws (out of bounds)
}
catch (std::exception& e)
{
std::cout << "S s2... raised an exception: " << e.what() << '\n';
}
try
{
S s3("123456", 0);
}
catch (std::exception& e)
{
std::cout << "S s2... raised an exception: " << e.what() << '\n';
}
}
I am analyzing part of code that was part of my lectures.
I have managed to compile it but I cannot understand:
why does my program output "Wyjatek" and 0 instead of "WyjatekNieoznaczony"?
I was pretty sure WyjatekNieoznaczony() should be thrown because a=0 and b=0 as well. Right now i am a little bit confused.
Could you help me, please?
class Wyjatek {};
class WyjatekBledny : public Wyjatek {};
class WyjatekNieoznaczony : public Wyjatek {};
double f(double a, double b) {
if (b == 0) {
if (a == 0)
throw WyjatekNieoznaczony();
else
throw WyjatekBledny();
}
return a / b;
}
double g(double a, double b) throw (int) {
try {
return f(a, b);
}
catch (WyjatekBledny) {
cout << "Wyjatek bledny" << endl;
throw 1;
}
catch (Wyjatek) {
cout << "Wyjatek" << endl;
}
catch (WyjatekNieoznaczony) {
cout << "Wyjatek nieoznaczony" << endl;
throw;
}
return 0;
}
int main()
{
double a = 0, b = 0;
try {
cout << g(a, b) << endl;
}
catch (...)
{
cout << "Inny wyjatek" << endl;
}
system("pause");
return 0;
}
Yes indeed a WyjatekNieoznaczony is thrown, but at the catch site, catch (Wyjatek) { is a match (due to the inheritance) so it's caught there.
A catch site is more like an if else block in behaviour - with each catch possibility being considered in the order they are written - rather than a switch block where you can put the labels in any order you like.
Note also that it's a good idea to catch exceptions by const reference than by value, else you can suffer the pitfalls of object slicing.
If you enabled (and read) compiler warnings, you would've encountered the following diagnostic:
warning: exception of type 'WyjatekNieoznaczony' will be caught [...] by earlier handler for 'Wyjatek'.
This basically means that WyjatekNieoznaczony, by inheriting from Wyjatek, will be first caught by catch(Wyjatek) clause, since it's convertible. The problem is that due to object slicing, it will lose its Nieoznaczonyness.
What I suggest is to reorder the catch clauses so the possibility of slicing disappears (in this case):
catch (WyjatekBledny) {
cout << "Wyjatek bledny" << endl;
throw 1;
}
catch (WyjatekNieoznaczony) {
cout << "Wyjatek nieoznaczony" << endl;
throw;
}
catch (Wyjatek) {
cout << "Wyjatek" << endl;
}
Consider this program
#include <iostream>
using namespace std;
class sample
{
public:
sample()
{
cout << "consructor called" << endl;
throw 5;
}
void test()
{
cout << "Test function" << endl;
}
};
int main()
{
sample *s = nullptr;
try
{
s = new sample[5];
cout << "allocated" << endl;
}
catch(bad_alloc& ba)
{
cout << ba.what() << endl;
}
catch (const int& f)
{
cout << "catcting exception";
}
return 0;
}
I think the flow will be like this.
1. Allocate the memory for 5 object.
2. call the constructor for each object one by one.
But here while calling the constructor, i'm throwing an exception, which is handled. My doubt is why constructor is not getting called for second objects onwards??
The object creation will be in sequence, it can't create all five objects in one go. When the first object gets created your constructor will get called and as it throws exception, it will move control to exception handler block.
You exception Handler will print the appropriate message and have graceful exit.
Try your test removing throw 5;
The new simple[5] will allocate memory for the 5 simples, and then begin constructing them one by one. Since the first one does a throw 5, the other 4 do not get constructed.
So I have an upcoming assignment dealing with exceptions and using them in my current address book program that most of the homework is centered around. I decided to play around with exceptions and the whole try catch thing, and using a class design, which is what I will eventually have to do for my assignment in a couple of weeks. I have working code that check the exception just fine, but what I want to know, is if there is a way to standardize my error message function, (i.e my what() call):
Here s my code:
#include <iostream>
#include <exception>
using namespace std;
class testException: public exception
{
public:
virtual const char* what() const throw() // my call to the std exception class function (doesn't nessasarily have to be virtual).
{
return "You can't divide by zero! Error code number 0, restarting the calculator..."; // my error message
}
void noZero();
}myex; //<-this is just a lazy way to create an object
int main()
{
void noZero();
int a, b;
cout << endl;
cout << "Enter a number to be divided " << endl;
cout << endl;
cin >> a;
cout << endl;
cout << "You entered " << a << " , Now give me a number to divide by " << endl;
cin >> b;
try
{
myex.noZero(b); // trys my exception from my class to see if there is an issue
}
catch(testException &te) // if the error is true, then this calls up the eror message and restarts the progrm from the start.
{
cout << te.what() << endl;
return main();
}
cout <<endl;
cout << "The two numbers divided are " << (a / b) << endl; // if no errors are found, then the calculation is performed and the program exits.
return 0;
}
void testException::noZero(int &b) //my function that tests what I want to check
{
if(b == 0 ) throw myex; // only need to see if the problem exists, if it does, I throw my exception object, if it doesn't I just move onto the regular code.
}
What I would like to be able to do is make it so my what() function can return a value dependent on what type of error is being called on. So for instance, if I were calling up an error that looked a the top number,(a), to see if it was a zero, and if it was, it would then set the message to say that "you can't have a numerator of zero", but still be inside the what() function. Here's an example:
virtual const char* what() const throw()
if(myex == 1)
{
return "You can't have a 0 for the numerator! Error code # 1 "
}
else
return "You can't divide by zero! Error code number 0, restarting the calculator..."; // my error message
}
This obviously wouldn't work, but is there a way to make it so I'm not writing a different function for each error message?
Your code contains a lot of misconceptions. The short answer is yes, you can change what() in order to return whatever you want. But let's go step by step.
#include <iostream>
#include <exception>
#include <stdexcept>
#include <sstream>
using namespace std;
class DivideByZeroException: public runtime_error {
public:
DivideByZeroException(int x, int y)
: runtime_error( "division by zero" ), numerator( x ), denominator( y )
{}
virtual const char* what() const throw()
{
cnvt.str( "" );
cnvt << runtime_error::what() << ": " << getNumerator()
<< " / " << getDenominator();
return cnvt.str().c_str();
}
int getNumerator() const
{ return numerator; }
int getDenominator() const
{ return denominator; }
template<typename T>
static T divide(const T& n1, const T& n2)
{
if ( n2 == T( 0 ) ) {
throw DivideByZeroException( n1, n2 );
}
return ( n1 / n2 );
}
private:
int numerator;
int denominator;
static ostringstream cnvt;
};
ostringstream DivideByZeroException::cnvt;
In the first place, runtime_error, derived from exception, is the adviced exception class to derive from. This is declared in the stdexcept header. You only have to initialize its constructor with the message you are going to return in the what() method.
Secondly, you should appropriately name your classes. I understand this is just a test, but a descriptive name will always help to read and understand your code.
As you can see, I've changed the constructor in order to accept the numbers to divide that provoked the exception. You did the test in the exception... well, I've respected this, but as a static function which can be invoked from the outside.
And finally, the what() method. Since we are dividing two numbers, it would be nice to show that two numbers that provoked the exception. The only way to achieve that is the use of ostringstream. Here we make it static so there is no problem of returning a pointer to a stack object (i.e., having cnvt a local variable would introduce undefined behaviour).
The rest of the program is more or less as you listed it in your question:
int main()
{
int a, b, result;
cout << endl;
cout << "Enter a number to be divided " << endl;
cout << endl;
cin >> a;
cout << endl;
cout << "You entered " << a << " , Now give me a number to divide by " << endl;
cin >> b;
try
{
result = DivideByZeroException::divide( a, b );
cout << "\nThe two numbers divided are " << result << endl;
}
catch(const DivideByZeroException &e)
{
cout << e.what() << endl;
}
return 0;
}
As you can see, I've removed your return main() instruction. It does not make sense, since you cannot call main() recursively. Also, the objective of that is a mistake: you'd expect to retry the operation that provoked the exception, but this is not possible, since exceptions are not reentrant. You can, however, change the source code a little bit, to achieve the same effect:
int main()
{
int a, b, result;
bool error;
do {
error = false;
cout << endl;
cout << "Enter a number to be divided " << endl;
cout << endl;
cin >> a;
cout << endl;
cout << "You entered " << a << " , Now give me a number to divide by " << endl;
cin >> b;
try
{
result = DivideByZeroException::divide( a, b ); // trys my exception from my class to see if there is an issue
cout << "\nThe two numbers divided are " << result << endl;
}
catch(const DivideByZeroException &e) // if the error is true, then this calls up the eror message and restarts the progrm from the start.
{
cout << e.what() << endl;
error = true;
}
} while( error );
return 0;
}
As you can see, in case of an error the execution follows until a "proper" division is entered.
Hope this helps.
You can create your own exception class for length error like this
class MyException : public std::length_error{
public:
MyException(const int &n):std::length_error(to_string(n)){}
};
class zeroNumerator: public std::exception
{
const char* what() const throw() { return "Numerator can't be 0.\n"; }
};
//...
try
{
myex.noZero(b); // trys my exception from my class to see if there is an issue
if(myex==1)
{
throw zeroNumerator(); // This would be a class that you create saying that you can't have 0 on the numerator
}
}
catch(testException &te)
{
cout << te.what() << endl;
return main();
}
You should always use std::exception&e. so do
catch(std::exception & e)
{
cout<<e.what();
}
You should consider a hierarchy of classes.
The reason for it might not be obvious when trying to use exceptions just for transferring a string, but actual intent of using exceptions should be a mechanism for advanced handling of exceptional situations. A lot of things are being done under the hood of C++ runtime environment while call stack is unwound when traveling from 'throw' to corresponded 'catch'.
An example of the classes could be:
class CalculationError : public std::runtime_error {
public:
CalculationError(const char * message)
:runtime_error(message)
{
}
};
class ZeroDeviderError : public CalculationError {
public:
ZeroDeviderError(int numerator, const char * message)
: CalculationError(message)
, numerator (numerator)
{
}
int GetNumerator() const { return numerator; }
private:
const int numerator;
};
Providing different classes for the errors, you give developers a chance to handle different errors in particular ways (not just display an error message)
Providing a base class for the types of error, allows developers to be more flexible - be as specific as they need.
In some cases, they might want to be specific
} catch (const ZeroDividerError & ex) {
// ...
}
in others, not
} catch (const CalculationError & ex) {
// ...
}
Some additional details:
You should not create objects of your exceptions before throwing in the manner you did. Regardless your intention, it is just useless - anyway, you are working with a copy of the object in the catch section (don't be confused by access via reference - another instance of the exception object is created when throwing)
Using a const reference would be a good style catch (const testException &te) unless you really need a non-constant object.
Also, please note that the type (classes) used for exceptions are not permitted to throw exceptions out of their copy constructors since, if the initial exception is attempted to be caught by value, a call of copy constructor is possible (in case is not elided by the compiler) and this additional exception will interrupt the initial exception handling before the initial exception is caught, which causes calling std::terminate.
Since C++11 compilers are permitted to eliminate the copying in some cases when catching, but both the elision is not always sensible and, if sensible, it is only permission but not obligation (see https://en.cppreference.com/w/cpp/language/copy_elision for details; before C++11 the standards of the language didn’t regulate the matter).
Also, you should avoid exceptions (will call them the additional) to be thrown out of constructors and move constructors of your types (classes) used for exceptions (will call them initial) since the constructors and move constructors could be called when throwing objects of the types as initial exceptions, then throwing out an additional exception would prevent creation of an initial exception object, and the initial would just be lost. As well as an additional exception from a copy constructor, when throwing an initial one, would cause the same.
I wrote a very simple solution however someone laughed and found a flaw as shown here http://ideone.com/IcWMEf
#include <iostream>
#include <ostream>
#include <functional>
#include <exception>
using namespace std;
// Wrong scope(failure)
class FailBlockT
{
typedef function<void()> T;
public:
T t;
FailBlockT(T t)
{
this->t=t;
}
~FailBlockT()
{
if (std::uncaught_exception())
{
t();
}
}
};
struct Test
{
~Test()
{
try
{
FailBlockT f([]()
{
cout << "failure" << endl;
});
// there is no any exception here, but "failure" is printed.
// See output below
}
catch(...)
{
cout << "some exception" << endl;
}
}
};
int main()
{
try
{
Test t;
throw 1;
}
catch(int){}
return 0;
}
In short the problem is my code looks at std::uncaught_exception(). When an exception is thrown and a normal destructor is executed. If i use scope failure there it will look at std::uncaught_exception() and think the object scope is lost due to exception rather then simply walking out of scope.
I can't think of any good solutions to differentiate leaving scope normally VS having an exception thrown IN it. Yes i know throwing is a bad idea in dtors BUT thats why I fail to notice this problem, because I never throw in exceptions.
How do I differentiate/solve this?
No exception was thrown but it thinks it has.
An exception was thrown, just not from right there.
There is no mechanism in C++11 to ask, "Was an exception thrown from code just below me, but not from code elsewhere in the call-stack?" std::uncaught_exception is doing exactly what it is supposed to do: say whether there is an exception currently in the process of being resolved at the time the function is called. And there is, so it returns true.
C++17 adds std::uncaught_exceptions (note the plural), which can be used to detect the difference. With such a tool, you can make your FailBlock object work:
template<typename Func>
class FailBlockT
{
private:
int e_count_;
T t_;
public:
FailBlockT(T t) : e_count_(std::uncaught_exceptions()), t_(t) {}
FailBlock(const FailBlock &) = delete; //The type should not be mobile.
~FailBlockT()
{
if (std::uncaught_exceptions() != e_count_)
{
t_();
}
}
};
std::uncaught_exceptions() returns the number of exceptions that are provoking stack unwinding at the time the call was made. If the number is the same during the constructor and destructor of an object (assuming it's a stack object), then the destructor is not being called due to an exception being thrown through where this type was used.
But without this tool, it, there's not much you can do to differentiate between an exception provoking the exiting of the scope rather than exiting a scope when exception unwinding just happens to be going on. So you're going to have to bite the bullet and catch the exception like everyone else.
Or just don't put this FailBlock thing in destructors. It seems to me that those should go directly into regular functions that can actually throw (and destructors should never throw). It seems to me that you're worried about a corner case that doesn't make any real sense.
I can't think of any good solutions to differentiate leaving scope normally VS having an exception thrown IN it.
Check stack_unwinding library - I have implemented scope(failure) and scope(success) features in C++.
It is based on platform specific function uncaught_exception_count. It is similar to std::uncaught_exception from standard library, but instead of boolean result it returns unsigned int showing current count of uncaught exceptions.
Currently it is tested on {Clang 3.2, GCC 3.4.6, GCC 4.1.2, GCC 4.4.6, GCC 4.4.7, MSVC2005SP1, MSVC2008SP1, MSVC2010SP1, MSVC2012} x {x32, x64}.
In C++11 folowing syntax is available:
try
{
int some_var=1;
cout << "Case #1: stack unwinding" << endl;
scope(exit)
{
cout << "exit " << some_var << endl;
++some_var;
};
scope(failure)
{
cout << "failure " << some_var << endl;
++some_var;
};
scope(success)
{
cout << "success " << some_var << endl;
++some_var;
};
throw 1;
} catch(int){}
{
int some_var=1;
cout << "Case #2: normal exit" << endl;
scope(exit)
{
cout << "exit " << some_var << endl;
++some_var;
};
scope(failure)
{
cout << "failure " << some_var << endl;
++some_var;
};
scope(success)
{
cout << "success " << some_var << endl;
++some_var;
};
}
In C++98 it is a bit more noisier:
try
{
cout << "Case #1: stack unwinding" << endl;
BOOST_SCOPE_EXIT(void) { cout << "exit" << endl; } BOOST_SCOPE_EXIT_END
SCOPE_FAILURE(void) { cout << "failure" << endl; } SCOPE_FAILURE_END
SCOPE_SUCCESS(void) { cout << "success" << endl; } SCOPE_SUCCESS_END
throw 1;
} catch(int){}
{
cout << "Case #2: normal exit" << endl;
BOOST_SCOPE_EXIT(void) { cout << "exit" << endl; } BOOST_SCOPE_EXIT_END
SCOPE_FAILURE(void) { cout << "failure" << endl; } SCOPE_FAILURE_END
SCOPE_SUCCESS(void) { cout << "success" << endl; } SCOPE_SUCCESS_END
}
Also, library has UNWINDING_AWARE_DESTRUCTOR feature. Example:
struct DestructorInClass
{
UNWINDING_AWARE_DESTRUCTOR(DestructorInClass,unwinding)
{
cout << "DestructorInClass, unwinding: "
<< ( unwinding ? "true" : "false" ) << endl;
}
};
However, there are some cases where UNWINDING_AWARE_DESTRUCTOR may give wrong results (though scope(success) and scope(failure) features are not affected by such issues).