• Table of Contents
• Previous Chapter
• Next Chapter

Chapter 18: The Standard Template Library

The algorithms can process these abstract data types because they are template based. This chapter does not cover template construction (see chapter 21 for that). Rather, it focuses on the use of the algorithms.

Several elements also used by the standard template library have already been discussed in the C++ Annotations. In chapter 12 abstract containers were discussed, and in section 11.10 function objects were introduced. Also, iterators were mentioned at several places in this document.

The main components of the STL are covered in this and the next chapter. Iterators, adaptors, smart pointers, multi threading and other features of the STL are discussed in coming sections. Generic algorithms are covered in the next chapter (19).

Allocators take care of the memory allocation within the STL. The default allocator class suffices for most applications, and is not further discussed in the C++ Annotations.

All elements of the STL are defined in the standard namespace. Therefore, a using namespace std or a comparable directive is required unless it is preferred to specify the required namespace explicitly. In header files the std namespace should explicitly be used (cf. section 7.11.1).

In this chapter the empty angle bracket notation is frequently used. In code a typename must be supplied between the angle brackets. E.g., plus<> is used in the C++ Annotations, but in code plus<string> may be encountered.

18.1: Predefined function objects

Function objects play important roles in generic algorithms. For example, there exists a generic algorithm sort expecting two iterators defining the range of objects that should be sorted, as well as a function object calling the appropriate comparison operator for two objects. Let's take a quick look at this situation. Assume strings are stored in a vector, and we want to sort the vector in descending order. In that case, sorting the vector stringVec is as simple as:

        sort(stringVec.begin(), stringVec.end(), greater<string>());

The last argument is recognized as a constructor: it is an instantiation of the greater<> class template, applied to strings. This object is called as a function object by the sort generic algorithm. The generic algorithm calls the function object's operator() member to compare two string objects. The function object's operator() will, in turn, call operator> of the string data type. Eventually, when sort returns, the first element of the vector will contain the string having the greatest string value of all.

The function object's operator() itself is not visible at this point. Don't confuse the parentheses in the `greater<string>()' argument with calling operator(). When operator() is actually used inside sort, it receives two arguments: two strings to compare for `greaterness'. Since greater<string>::operator() is defined inline, the call itself is not actually present in the above sort call. Instead sort calls string::operator> through greater<string>::operator().

Now that we know that a constructor is passed as argument to (many) generic algorithms, we can design our own function objects. Assume we want to sort our vector case-insensitively. How do we proceed? First we note that the default string::operator< (for an incremental sort) is not appropriate, as it does case sensitive comparisons. So, we provide our own CaseInsensitive class, which compares two strings case insensitively. Using the POSIX function strcasecmp, the following program performs the trick. It case-insensitively sorts its command-line arguments in ascending alphabetic order:

    #include <iostream>
    #include <string>
    #include <cstring>
    #include <algorithm>
    using namespace std;

    class CaseInsensitive
    {
        public:
            bool operator()(string const &left, string const &right) const
            {
                return strcasecmp(left.c_str(), right.c_str()) < 0;
            }
    };
    int main(int argc, char **argv)
    {
        sort(argv, argv + argc, CaseInsensitive{});
        for (int idx = 0; idx < argc; ++idx)
            cout << argv[idx] << " ";
        cout << '\n';
    }

The comparison function object is often a predefined function object. Predefined function object classes are available for many commonly used operations. In the following sections the available predefined function objects are presented, together with some examples showing their use. Near the end of the section about function objects function adaptors are introduced.

Predefined function objects are used predominantly with generic algorithms. Predefined function objects exists for arithmetic, relational, and logical operations.

18.1.1: Arithmetic function objects

    #include <iostream>
    #include <string>
    #include <functional>
    using namespace std;

    int main(int argc, char **argv)
    {
        plus<size_t> uAdd;              // function object to add size_ts

        cout << "3 + 5 = " << uAdd(3, 5) << '\n';

        plus<string> sAdd;              // function object to add strings

        cout << "argv[0] + argv[1] = " << sAdd(argv[0], argv[1]) << '\n';
    }
    /*
        Output when called as: a.out going

        3 + 5 = 8
        argv[0] + argv[1] = a.outgoing
    */

Suppose we want to perform an operation on a left hand side operand which is always the same variable and a right hand side argument for which, in turn, all elements of an array should be used. E.g., we want to compute the sum of all elements in an array; or we want to concatenate all the strings in a text-array. In situations like these function objects come in handy.

As stated, function objects are heavily used in the context of the generic algorithms, so let's take a quick look ahead at yet another one.

The generic algorithm accumulate visits all elements specified by an iterator-range, and performs a requested binary operation on a common element and each of the elements in the range, returning the accumulated result after visiting all elements specified by the iterator range. It's easy to use this algorithm. The next program accumulates all command line arguments and prints the final string:

    #include <iostream>
    #include <string>
    #include <functional>
    #include <numeric>
    using namespace std;

    int main(int argc, char **argv)
    {
        string result =
                accumulate(argv, argv + argc, string{}, plus<string>());

        cout << "All concatenated arguments: " << result << '\n';
    }

    "All concatenated arguments: "s

in which case the cout statement could simply have been cout << result << '\n'. The string-addition operation is used, called from plus<string>. The final concatenated string is returned.

Now we define a class Time, overloading operator+. Again, we can apply the predefined function object plus, now tailored to our newly defined datatype, to add times:

    #include <iostream>
    #include <string>
    #include <vector>
    #include <functional>
    #include <numeric>
    using namespace std;

    class Time
    {
        friend ostream &operator<<(ostream &str, Time const &time);
        size_t d_days;
        size_t d_hours;
        size_t d_minutes;
        size_t d_seconds;
        public:
            Time(size_t hours, size_t minutes, size_t seconds);
            Time &operator+=(Time const &rhs);
    };
    Time &&operator+(Time const &lhs, Time const &rhs)
    {
        Time ret(lhs);
        return std::move(ret += rhs);
    }
    Time::Time(size_t hours, size_t minutes, size_t seconds)
    :
        d_days(0),
        d_hours(hours),
        d_minutes(minutes),
        d_seconds(seconds)
    {}
    Time &Time::operator+=(Time const &rhs)
    {
        d_seconds   += rhs.d_seconds;
        d_minutes   += rhs.d_minutes   + d_seconds / 60;
        d_hours     += rhs.d_hours     + d_minutes / 60;
        d_days      += rhs.d_days      + d_hours   / 24;
        d_seconds   %= 60;
        d_minutes   %= 60;
        d_hours     %= 24;
        return *this;
    }
    ostream &operator<<(ostream &str, Time const &time)
    {
        return cout << time.d_days << " days, " << time.d_hours <<
                                                    " hours, " <<
                        time.d_minutes << " minutes and " <<
                        time.d_seconds << " seconds.";
    }
    int main(int argc, char **argv)
    {
        vector<Time> tvector;

        tvector.push_back(Time( 1, 10, 20));
        tvector.push_back(Time(10, 30, 40));
        tvector.push_back(Time(20, 50,  0));
        tvector.push_back(Time(30, 20, 30));

        cout <<
            accumulate
            (
                tvector.begin(), tvector.end(), Time(0, 0, 0), plus<Time>()
            ) <<
            '\n';
    }
    //  Displays: 2 days, 14 hours, 51 minutes and 30 seconds.

While this section's first example illustrated using a named function object, the last two examples illustrate how anonymous objects can be passed to the (accumulate) function.

The STL supports the following set of arithmetic function objects. The function call operator (operator()) of these function objects calls the matching arithmetic operator for the objects that are passed to the function call operator, returning that arithmetic operator's return value. The arithmetic operator that is actually called is mentioned below:

• plus<>: calls the binary operator+;
• minus<>: calls the binary operator-;
• multiplies<>: calls the binary operator*;
• divides<>: calls operator/;
• modulus<>: calls operator%;
• negate<>: calls the unary operator-. This arithmetic function object is a unary function object as it expects one argument.

    #include <iostream>
    #include <string>
    #include <functional>
    #include <algorithm>
    using namespace std;

    int main(int argc, char **argv)
    {
        int iArr[] = { 1, -2, 3, -4, 5, -6 };

        transform(iArr, iArr + 6, iArr, negate<int>());

        for (int idx = 0; idx < 6; ++idx)
            cout << iArr[idx] << ", ";
        cout << '\n';
    }
    // Displays:  -1, 2, -3, 4, -5, 6,

18.1.2: Relational function objects

The STL supports the following set of relational function objects. The function call operator (operator()) of these function objects calls the matching relational operator for the objects that are passed to the function call operator, returning that relational operator's return value. The relational operator that is actually called is mentioned below:

• equal_to<>: calls operator==;
• not_equal_to<>: calls operator!=;
• greater<>: calls operator>;
• greater_equal<>: calls operator>=;
• less<>: this object's member operator() calls operator<;
• less_equal<>: calls operator<=.

    #include <iostream>
    #include <string>
    #include <functional>
    #include <algorithm>
    using namespace std;

    int main(int argc, char **argv)
    {
        sort(argv, argv + argc, greater_equal<string>());

        for (int idx = 0; idx < argc; ++idx)
            cout << argv[idx] << " ";
        cout << '\n';

        sort(argv, argv + argc, less<string>());

        for (int idx = 0; idx < argc; ++idx)
            cout << argv[idx] << " ";
        cout << '\n';
    }

Note that argv contains char * values, and that the relational function object expects a string. The promotion from char const * to string is silently performed.

18.1.3: Logical function objects

The STL supports the following set of logical function objects. The function call operator (operator()) of these function objects calls the matching logical operator for the objects that are passed to the function call operator, returning that logical operator's return value. The logical operator that is actually called is mentioned below:

• logical_and<>: calls operator&&;
• logical_or<>: calls operator||;
• logical_not<>: calls operator!.

    #include <iostream>
    #include <string>
    #include <functional>
    #include <algorithm>
    using namespace std;

    int main(int argc, char **argv)
    {
        bool bArr[] = {true, true, true, false, false, false};
        size_t const bArrSize = sizeof(bArr) / sizeof(bool);

        for (size_t idx = 0; idx < bArrSize; ++idx)
            cout << bArr[idx] << " ";
        cout << '\n';

        transform(bArr, bArr + bArrSize, bArr, logical_not<bool>());

        for (size_t idx = 0; idx < bArrSize; ++idx)
            cout << bArr[idx] << " ";
        cout << '\n';
    }
    /*
      Displays:

        1 1 1 0 0 0
        0 0 0 1 1 1
    */

18.1.4: The `std::not_fn' negator

The standard negator is std::not_fn, declared in the <functional> header file.

The function not_fn expects a (movable) object as its argument, returning the negated value of the return value of its argument's function call operator.

As an example consider a main function defining an array of int values:

    int main()
    {
        int arr[] = {1, 2, 3, 4, 5, 6, 7, 8, 9};
    }

To count the number of even values count_if, using a lambda function can be used:

        cout << 
            count_if(arr, arr + size(arr),
                [&](int value)
                {
                    return (value & 1) == 0;
                }
            ) << '\n';

To count the number of odd values, not_fn can be used in the above code like so:

        cout << 
            count_if(arr, arr + size(arr),
                not_fn(
                    [&](int value)
                    {
                        return (value & 1) == 0;
                    }
                )
            ) << '\n';

Of course, in this simple example the lambda function could also easily have been modified. But if instead of a lambda function an existing class implementing a function object had been used it would have been difficult or impossible to change the behavior of that class. If the class offers moving operations then not_fn can be used to negate the values returned by that class's function call operator.

18.2: Iterators

The standard iterator (std::iterator) is now deprecated (http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2016/p0174r1.html#2.1), and the compiler issues a corresponding warning. Consequently, st::iterator should no longer be used when designing your own iterators (section 22.14 describes how to design your own).

Iterators are objects acting like pointers. Iterators have the following general characteristics:

• Two iterators may be compared for (in)equality using the == and != operators. The ordering operators (e.g., >, <) can usually not be used.
• Given an iterator iter, *iter represents the object the iterator points to (alternatively, iter-> can be used to reach the members of the object the iterator points to).
• Given an iterator iter, iter.base() returns the address of *iter. It returns the same type as &*iter. E.g.,

      vector<int> vi{ 1, 2, 3 };
      int *ip = vi.begin().base();
      cout << *ip << '\n';        // shows: 1
  

• ++iter or iter++ advances the iterator to the next element. The notion of advancing an iterator to the next element is consequently applied: several containers support reverse_iterator types, in which the ++iter operation actually reaches a previous element in a sequence.
• Pointer arithmetic may be used with iterators of containers storing their elements consecutively in memory like vector and deque. For such containers iter + 2 points to the second element beyond the one to which iter points. See also section 18.2.1, covering std::distance.
• Merely defining an iterator is comparable to having a 0-pointer. Example:

  #include <vector>
  #include <iostream>
  using namespace std;
  
  int main()
  {
      vector<int>::iterator vi;
  
      cout << &*vi;       // prints 0
  }
  

Whereas reverse iterators can be constructed from ordinary (forward) iterators using reverse_iterator constructors as in:

    string str;
    auto revit = string::reverse_iterator{ str.begin() };

the opposite is not accomplished that way. To retrieve the forward iterator corresponding to a reverse iterator, the reverse_iterator.base() member can be used. E.g., to obtain the forward iterator corresponding to revit use

    auto forward { revit.base() };

Standard practice requires iterator ranges to be left inclusive. The notation [left, right) indicates that left is an iterator pointing to the first element, while right is an iterator pointing just beyond the last element. The iterator range is empty when left == right.

The following example shows how all elements of a vector of strings can be inserted into cout using its iterator ranges [begin(), end()), and [rbegin(), rend()). Note that the for-loops for both ranges are identical. Furthermore it nicely illustrates how the auto keyword can be used to define the type of the loop control variable instead of using a much more verbose variable definition like vector<string>::iterator (see also section 3.3.7):

    #include <iostream>
    #include <vector>
    #include <string>
    using namespace std;

    int main(int argc, char **argv)
    {
        vector<string> args(argv, argv + argc);

        for (auto iter = args.begin(); iter != args.end(); ++iter)
            cout << *iter << " ";
        cout << '\n';

        for (auto iter = args.rbegin(); iter != args.rend(); ++iter)
            cout << *iter << " ";
        cout << '\n';
    }

Furthermore, the STL defines const_iterator types that must be used when visiting a series of elements in a constant container. Whereas the elements of the vector in the previous example could have been altered, the elements of the vector in the next example are immutable, and const_iterators are required:

    #include <iostream>
    #include <vector>
    #include <string>
    using namespace std;

    int main(int argc, char **argv)
    {
        vector<string> const args(argv, argv + argc);

        for
        (
            vector<string>::const_iterator iter = args.begin();
                iter != args.end();
                    ++iter
        )
            cout << *iter << " ";
        cout << '\n';

        for
        (
            vector<string>::const_reverse_iterator iter = args.rbegin();
                iter != args.rend();
                    ++iter
        )
            cout << *iter << " ";
        cout << '\n';
    }

The STL defines six types of iterators. These iterator types are expected by generic algorithms, and in order to create a particular type of iterator yourself it is important to know their characteristics. In general, iterators (see also section 22.14) must define:

• operator==, testing two iterators for equality,
• operator!=, testing two iterators for inequality,
• operator++, incrementing the iterator, as prefix operator,
• operator*, to access the element the iterator refers to,

• InputIterators:

  InputIterators are used to read from a container. The dereference operator is guaranteed to work as rvalue in expressions. Instead of an InputIterator it is also possible to use (see below) Forward-, Bidirectional- or RandomAccessIterators. Notations like InputIterator1 and InputIterator2 may be used as well. In these cases, numbers are used to indicate which iterators `belong together'. E.g., the generic algorithm inner_product has the following prototype:

  Type inner_product(InputIterator1 first1, InputIterator1 last1,
                 InputIterator2 first2, Type init);
  

  InputIterator1 first1 and InputIterator1 last1 define a pair of input iterators on one range, while InputIterator2 first2 defines the beginning of another range. Analogous notations may be used with other iterator types.

• OutputIterators:

  OutputIterators can be used to write to a container. The dereference operator is guaranteed to work as an lvalue in expressions, but not necessarily as rvalue. Instead of an OutputIterator it is also possible to use (see below) Forward-, Bidirectional- or RandomAccessIterators.

• ForwardIterators:

  ForwardIterators combine InputIterators and OutputIterators. They can be used to traverse containers in one direction, for reading and/or writing. Instead of a ForwardIterator it is also possible to use (see below) Bidirectional- or RandomAccessIterators.

• BidirectionalIterators:

  BidirectionalIterators can be used to traverse containers in both directions, for reading and writing. Instead of a BidirectionalIterator it is also possible to use (see below) a RandomAccessIterator.

• RandomAccessIterators:

  RandomAccessIterators provide random access to container elements. An algorithm like sort requires a RandomAccessIterator, and can therefore not be used to sort the elements of lists or maps, which only provide BidirectionalIterators.

• ContiguousIterators:

  ContiguousIterators are like random access iterators, but in addition guarantee that the elements these iterators point to are stored contiguously in memory. A container like std::vector offers contiguous iterators.

18.2.1: std::distance and std::size

Using pointer arithmetic to compute the number of elements between two iterators in, e.g., a std::list or std::unordered_map is not possible, as these containers do not store their elements consecutively in memory.

The function std::distance fills in that little gap: std::distance expects two InputIterators and returns the number of elements between them.

Before using distance the <iterator> header file must be included.

If the iterator specified as first argument exceeds the iterator specified as its second argument then the number of elements is non-positive, otherwise it is non-negative. If the number of elements cannot be determined (e.g., the iterators do not refer to elements in the same container), then distance's return value is undefined.

Example:

    #include <iostream>
    #include <unordered_map>
        
    using namespace std;
    
    int main()
    {
        unordered_map<int, int> myMap = {{1, 2}, {3, 5}, {-8, 12}};
    
        cout << distance(++myMap.begin(), myMap.end()) << '\n'; // shows: 2
    }

The iterator header file also defines the function std::size, returning the number of elements in a containers (as returned by the container's size member) or of an array whose dimension is known to the compiler at the point of std::size's call. E.g., if the size of an array data is known to the compiler, then to call a function handler (expecting the address of the first element of an array and the address of the location just beyond that array) the following statement can be used:

    handler(data, data + std::size(data));

As noted, the std::size function is defined in the iterator header. However, it's also guaranteed available when including the header file of a container supporting iterators (including the string header file).

18.2.2: Insert iterators

With the copy algorithm the number of elements to copy is usually available beforehand, since that number can usually be provided by pointer arithmetic. However, situations exist where pointer arithmetic cannot be used. Analogously, the number of resulting elements sometimes differs from the number of elements in the initial range. The generic algorithm unique_copy is a case in point. Here the number of elements that are copied to the destination container is normally not known beforehand.

In situations like these an inserter adaptor function can often be used to create elements in the destination container. There are three types of inserter adaptors:

• back_inserter: calls the container's push_back member to add new elements at the end of the container. E.g., to copy all elements of source in reversed order to the back of destination, using the copy generic algorithm:

  copy(source.rbegin(), source.rend(), back_inserter(destination));
  

• front_inserter calls the container's push_front member, adding new elements at the beginning of the container. E.g., to copy all elements of source to the front of the destination container (thereby also reversing the order of the elements):

  copy(source.begin(), source.end(), front_inserter(destination));
  

• inserter calls the container's insert member adding new elements starting at a specified starting point. E.g., to copy all elements of source to the destination container, starting at the beginning of destination, shifting up existing elements to beyond the newly inserted elements:

  copy(source.begin(), source.end(), inserter(destination,
      destination.begin()));
  

• using value_type = Data, where Data is the data type stored in the class offering push_back, push_front or insert members (Example: using value_type = std::string);
• using const_reference = const &value_type

    using const_reference = DataType const &;

to its interface (where DataType const & is the type of the parameter of the class's member push_back). Example:

    #include <iostream>
    #include <algorithm>
    #include <iterator>
    using namespace std;

    class Insertable
    {
        public:
            using value_type = int;
            using const_reference = int const &;

            void push_back(int const &)
            {}
    };

    int main()
    {
        int arr[] = {1};
        Insertable insertable;

        copy(arr, arr + 1, back_inserter(insertable));
    }

18.2.3: Iterators for `istream' objects

    istream_iterator<Type> identifier(istream &in)

Here, Type is the type of the data elements read from the istream stream. It is used as the `begin' iterator in an iterator range. Type may be any type for which operator>> is defined in combination with istream objects.

The default constructor is used as the end-iterator and corresponds to the end-of-stream. For example,

    istream_iterator<string> endOfStream;

The stream object that was specified when defining the begin-iterator is not mentioned with the default constructor.

Using back_inserter and istream_iterator adaptors, all strings from a stream can easily be stored in a container. Example (using anonymous istream_iterator adaptors):

    #include <iostream>
    #include <iterator>
    #include <string>
    #include <vector>
    #include <algorithm>
    using namespace std;

    int main()
    {
        vector<string> vs;

        copy(istream_iterator<string>(cin), istream_iterator<string>(),
             back_inserter(vs));

        for
        (
            vector<string>::const_iterator begin = vs.begin(), end = vs.end();
                begin != end; ++begin
        )
            cout << *begin << ' ';
        cout << '\n';
    }

18.2.3.1: Iterators for `istreambuf' objects

To read from streambuf objects supporting input operations istreambuf_iterators can be used, supporting the operations that are also available for istream_iterator. Different from the latter iterator type istreambuf_iterators support three constructors:

• istreambuf_iterator<Type>:
  The end iterator of an iterator range is created using the default istreambuf_iterator constructor. It represents the end-of-stream condition when extracting values of type Type from the streambuf.

• istreambuf_iterator<Type>(streambuf *):
  A pointer to a streambuf may be used when defining an istreambuf_iterator. It represents the begin iterator of an iterator range.
• istreambuf_iterator<Type>(istream):
  An istream may be also used when defining an istreambuf_iterator. It accesses the istream's streambuf and it also represents the begin iterator of an iterator range.

18.2.4: Iterators for `ostream' objects

    ostream_iterator<Type> identifier(ostream &outStream);
    ostream_iterator<Type> identifier(ostream &outStream, char const *delim);

Type is the type of the data elements that should be inserted into an ostream. It may be any type for which operator<< is defined in combination with ostream objects. The latter constructor can be used to separate the individual Type data elements by delimiter strings. The former constructor does not use any delimiters.

The example shows how istream_iterators and an ostream_iterator may be used to copy information of a file to another file. A subtlety here is that you probably want to use in.unsetf(ios::skipws). It is used to clear the ios::skipws flag. As a consequence whitespace characters are simply returned by the operator, and the file is copied character by character. Here is the program:

    #include <iostream>
    #include <algorithm>
    #include <iterator>
    using namespace std;

    int main()
    {
        cin.unsetf(ios::skipws);
        copy(istream_iterator<char>(cin), istream_iterator<char>(),
             ostream_iterator<char>(cout));
    }

18.2.4.1: Iterators for `ostreambuf' objects

To write to streambuf objects supporting output operations ostreambuf_iterators can be used, supporting the operations that are also available for ostream_iterator. Ostreambuf_iterators support two constructors:

• ostreambuf_iterator<Type>(streambuf *):
  A pointer to a streambuf may be used when defining an ostreambuf_iterator. It can be used as an OutputIterator.
• ostreambuf_iterator<Type>(ostream):
  An ostream may be also used when defining an ostreambuf_iterator. It accesses the ostream's streambuf and it can also be used as an OutputIterator.

    #include <iostream>
    #include <algorithm>
    #include <iterator>
    using namespace std;

    int main()
    {
        istreambuf_iterator<char> in(cin.rdbuf());
        istreambuf_iterator<char> eof;
        ostreambuf_iterator<char> out(cout.rdbuf());

        copy(in, eof, out);
    }

18.3: The class 'unique_ptr'

When pointers are used to access dynamically allocated memory strict bookkeeping is required to prevent memory leaks. When a pointer variable referring to dynamically allocated memory goes out of scope, the dynamically allocated memory becomes inaccessible and the program suffers from a memory leak. Consequently, the programmer has to make sure that the dynamically allocated memory is returned to the common pool just before the pointer variable goes out of scope.

When a pointer variable points to a dynamically allocated single value or object, bookkeeping requirements are greatly simplified when the pointer variable is defined as a std::unique_ptr object.

Unique_ptrs are objects masquerading as pointers. Since they are objects, their destructors are called when they go out of scope. Their destructors automatically delete the dynamically allocated memory to which they point. Unique_ptrs (and their cousins shared_ptrs (cf. section 18.4) are also called smart pointers).

Unique_ptrs have several special characteristics:

• when assigning a unique_ptr to another move semantics is used. If move semantics is not available compilation fails. On the other hand, if compilation succeeds then the used containers or generic algorithms support the use of unique_ptrs. Here is an example:

  std::unique_ptr<int> up1(new int);
  std::unique_ptr<int> up2(up1);      // compilation error
  

  The second definition fails to compile as unique_ptr's copy constructor is private (the same holds true for the assignment operator). But the unique_ptr class does offer facilities to initialize and assign from rvalue references:

  class unique_ptr                        // interface partially shown
  {
      public:
          unique_ptr(unique_ptr &&tmp);   // rvalues bind here
      private:
          unique_ptr(const unique_ptr &other);
  };
  

  In the next example move semantics is used and so it compiles correctly:

  unique_ptr<int> cp(unique_ptr<int>(new int));
  

• a unique_ptr object should only point to memory that was made available dynamically, as only dynamically allocated memory can be deleted.

• multiple unique_ptr objects should not be allowed to point to the same block of dynamically allocated memory. The unique_ptr's interface was designed to prevent this from happening. Once a unique_ptr object goes out of scope, it deletes the memory it points to, immediately changing any other object also pointing to the allocated memory into a wild pointer.

• When a class Derived is derived from Base, then a newly allocated Derived class object can be assigned to a unique_ptr<Base>, without having to define a virtual destructor for Base. The Base * pointer that is returned by the unique_ptr object can simply be cast statically to Derived, and Derived's destructor is automatically called as well, if the unique_ptr definition is provided with a deleter function address. This is illustrated in the next example:

      class Base
      { ... };
      class Derived: public Base
      {
          ...
          public:
              // assume Derived has a member void process()
     
              static void deleter(Base *bp);
      };
      void Derived::deleter(Base *bp)
      {
          delete static_cast<Derived *>(bp);
      }
      int main()
      {
          unique_ptr<Base, void (*)(Base *)> bp(new Derived, &Derived::deleter);
          static_cast<Derived *>(bp.get())->process(); // OK!
  
      } // here ~Derived is called: no polymorphism required.
  

Unique_ptr can also be used with containers and (generic) algorithms. They can properly destruct any type of object, as their constructors accept customizable deleters. In addition, arrays can be handled by unique_ptrs.

18.3.1: Defining `unique_ptr' objects

• The default constructor simply creates a unique_ptr object that does not point to a particular block of memory. Its pointer is initialized to 0 (zero):

  unique_ptr<type> identifier;
  

  This form is discussed in section 18.3.2.

• The move constructor initializes an unique_ptr object. Following the use of the move constructor its unique_ptr argument no longer points to the dynamically allocated memory and its pointer data member is turned into a zero-pointer:

  unique_ptr<type> identifier(another unique_ptr for type);
  

  This form is discussed in section 18.3.3.

• The form that is used most often initializes a unique_ptr object to the block of dynamically allocated memory that is passed to the object's constructor. Optionally deleter can be provided. A (free) function (or function object) receiving the unique_ptr's pointer as its argument can be passed as deleter. It is supposed to return the dynamically allocated memory to the common pool (doing nothing if the pointer equals zero).

  unique_ptr<type> identifier (new-expression [, deleter]);
  

  This form is discussed in section 18.3.4.

18.3.2: Creating a plain `unique_ptr'

    unique_ptr<type> identifier;

The pointer controlled by the unique_ptr object is initialized to 0 (zero). Although the unique_ptr object itself is not the pointer, its value can be compared to 0. Example:

    unique_ptr<int> ip;

    if (!ip)
        cout << "0-pointer with a unique_ptr object\n";

Alternatively, the member get can be used (cf. section 18.3.5).

18.3.3: Moving another `unique_ptr'

    unique_ptr<type> identifier(other unique_ptr object);

The move constructor is used, e.g., in the following example:

    void mover(unique_ptr<string> &&param)
    {
        unique_ptr<string> tmp(move(param));
    }

Analogously, the assignment operator can be used. A unique_ptr object may be assigned to a temporary unique_ptr object of the same type (again move-semantics is used). For example:

    #include <iostream>
    #include <memory>
    #include <string>

    using namespace std;

    int main()
    {
        unique_ptr<string> hello1{ new string{ "Hello world" } };
        unique_ptr<string> hello2{ move(hello1) };
        unique_ptr<string> hello3;

        hello3 = move(hello2);
        cout << // *hello1 << '\n' <<   // would have segfaulted
                // *hello2 << '\n' <<   // same
                *hello3 << '\n';
    }
    // Displays:    Hello world

• hello1 is initialized by a pointer to a dynamically allocated string (see the next section).
• The unique_ptr hello2 grabs the pointer controlled by hello1 using a move constructor. This effectively changes hello1 into a 0-pointer.
• Then hello3 is defined as a default unique_ptr<string>. But then it grabs its value using move-assignment from hello2 (which, as a consequence, is changed into a 0-pointer).

18.3.4: Pointing to a newly allocated object

    unique_ptr<type [, deleter_type]> identifier(new-expression
            [, deleter = deleter_type{}]);

The second (template) argument (deleter(_type)) is optional and may refer to a free function, a function object handling the destruction of the allocated memory, or a lambda function. A deleter is used, e.g., in situations where a double pointer is allocated and the destruction must visit each nested pointer to destroy the allocated memory (see below for an illustration).

Here is an example initializing a unique_ptr pointing to a string object:

    unique_ptr<string> strPtr{ new string{ "Hello world" } };

The argument that is passed to the constructor is the pointer returned by operator new. Note that type does not mention the pointer. The type that is used in the unique_ptr construction is the same as the type that is used in new expressions.

Here is an example showing how an explicitly defined deleter may be used to delete a dynamically allocated array of pointers to strings:

    #include <iostream>
    #include <string>
    #include <memory>
    using namespace std;

    struct Deleter
    {
        size_t d_size;
        Deleter(size_t size = 0)
        :
            d_size(size)
        {}
        void operator()(string **ptr) const
        {
            for (size_t idx = 0; idx < d_size; ++idx)
                delete ptr[idx];
            delete[] ptr;
        }
    };
    int main()
    {
        unique_ptr<string *, Deleter> sp2{ new string *[10], Deleter{ 10 } };

        Deleter &obj = sp2.get_deleter();
    }

A unique_ptr can be used to reach the member functions that are available for objects allocated by the new expression. These members can be reached as if the unique_ptr was a plain pointer to the dynamically allocated object. For example, in the following program the text `C++' is inserted behind the word `hello':

    #include <iostream>
    #include <memory>
    #include <cstring>
    using namespace std;

    int main()
    {
        unique_ptr<string> sp{ new string{ "Hello world" } };

        cout << *sp << '\n';
        sp->insert(strlen("Hello "), "C++ ");
        cout << *sp << '\n';
    }
    /*
        Displays:
            Hello world
            Hello C++ world
    */

18.3.5: Operators and members

• unique_ptr<Type> &operator=(unique_ptr<Type> &&tmp):
  This operator transfers the memory pointed to by the rvalue unique_ptr object to the lvalue unique_ptr object using move semantics. So, the rvalue object loses the memory it pointed at and turns into a 0-pointer. An existing unique_ptr may be assigned to another unique_ptr by converting it to an rvalue reference first using std::move. Example:

  unique_ptr<int> ip1(new int);
  unique_ptr<int> ip2;
  
  ip2 = std::move(ip1);
  

• operator bool() const:
  This operator returns false if the unique_ptr does not point to memory (i.e., its get member, see below, returns 0). Otherwise, true is returned.
• Type &operator*():
  This operator returns a reference to the information accessible via a unique_ptr object . It acts like a normal pointer dereference operator.
• Type *operator->():
  This operator returns a pointer to the information accessible via a unique_ptr object. This operator allows you to select members of an object accessible via a unique_ptr object. Example:

  unique_ptr<string> sp{ new string{ "hello" } };
  cout << sp->c_str();
  

The class unique_ptr supports the following member functions:

• Type *get():
  A pointer to the information controlled by the unique_ptr object is returned. It acts like operator->. The returned pointer can be inspected. If it is zero the unique_ptr object does not point to any memory.
• Deleter &unique_ptr<Type>::get_deleter():
  A reference to the deleter object used by the unique_ptr is returned.
• Type *release():
  A pointer to the information accessible via a unique_ptr object is returned. At the same time the object itself becomes a 0-pointer (i.e., its pointer data member is turned into a 0-pointer). This member can be used to transfer the information accessible via a unique_ptr object to a plain Type pointer. After calling this member the proper destruction of the dynamically allocated memory is the responsibility of the programmer.
• void reset(Type *):
  The dynamically allocated memory controlled by the unique_ptr object is returned to the common pool; the object thereupon controls the memory to which the argument that is passed to the function points. It can also be called without argument, turning the object into a 0-pointer. This member function can be used to assign a new block of dynamically allocated memory to a unique_ptr object.
• void swap(unique_ptr<Type> &):
  Two identically typed unique_ptrs are swapped.

18.3.6: Using `unique_ptr' objects for arrays

With dynamically allocated arrays the following syntax is available:

• the index ([]) notation is used to specify that the smart pointer controls a dynamically allocated array. Example:

  unique_ptr<int[]> intArr(new int[3]);
  

• the index operator can be used to access the array's elements. Example:

  intArr[2] = intArr[0];
  

18.4: The class `shared_ptr'

Before using shared_ptrs the <memory> header file must be included.

The shared pointer automatically destroys its content once its reference count has decayed to zero. As with unique_ptr, when defining a shared_ptr<Base> to store a newly allocated Derived class object, the returned Base * may be cast to a Derived * using a static_cast: polymorphism isn't required, and when resetting the shared_ptr or when the shared_ptr goes out of scope, no slicing occurs, and Derived's destructor (or, if configured: deleter) is called (cf. section 18.3).

Shared_ptrs support copy and move constructors as well as standard and move overloaded assignment operators.

Like unique_ptrs, shared_ptrs may refer to dynamically allocated arrays.

18.4.1: Defining `shared_ptr' objects

• The default constructor simply creates a shared_ptr object that does not point to a particular block of memory. Its pointer is initialized to 0 (zero):

  shared_ptr<type> identifier;
  

  This form is discussed in section 18.4.2.

• The copy constructor initializes a shared_ptr so that both objects share the memory pointed at by the existing object. The copy constructor also increments the shared_ptr's reference count. Example:

  shared_ptr<string> org{ new string{ "hi there" } };
  shared_ptr<string> copy(org);   // reference count now 2
  

• The move constructor initializes a shared_ptr with the pointer and reference count of a temporary shared_ptr. The temporary shared_ptr is changed into a 0-pointer. An existing shared_ptr may have its data moved to a newly defined shared_ptr (turning the existing shared_ptr into a 0-pointer as well). In the next example a temporary, anonymous shared_ptr object is constructed, which is then used to construct grabber. Since grabber's constructor receives an anonymous temporary object, the compiler uses shared_ptr's move constructor:

  shared_ptr<string> grabber{ shared_ptr<string>{ new string{ "hi there" } } };
  

• The form that is used most often initializes a shared_ptr object to the block of dynamically allocated memory that is passed to the object's constructor. Optionally deleter can be provided. A (free) function (or function object) receiving the shared_ptr's pointer as its argument can be passed as deleter. It is supposed to return the dynamically allocated memory to the common pool (doing nothing if the pointer equals zero).

  shared_ptr<type> identifier (new-expression [, deleter]);
  

  This form is discussed in section 18.4.3.

18.4.2: Creating a plain `shared_ptr'

    shared_ptr<type> identifier;

The pointer controlled by the shared_ptr object is initialized to 0 (zero). Although the shared_ptr object itself is not the pointer, its value can be compared to 0. Example:

    shared_ptr<int> ip;

    if (!ip)
        cout << "0-pointer with a shared_ptr object\n";

Alternatively, the member get can be used (cf. section 18.4.4).

18.4.3: Pointing to a newly allocated object

    shared_ptr<type> identifier(new-expression [, deleter]);

The second argument (deleter) is optional and refers to a function object or free function handling the destruction of the allocated memory. A deleter is used, e.g., in situations where a double pointer is allocated and the destruction must visit each nested pointer to destroy the allocated memory (see below for an illustration). It is used in situations comparable to those encountered with unique_ptr (cf. section 18.3.4).

Here is an example initializing a shared_ptr pointing to a string object:

    shared_ptr<string> strPtr{ new string{ "Hello world" } };

The argument that is passed to the constructor is the pointer returned by operator new. Note that type does not mention the pointer. The type that is used in the shared_ptr construction is the same as the type that is used in new expressions.

The next example illustrates that two shared_ptrs indeed share their information. After modifying the information controlled by one of the objects the information controlled by the other object is modified as well:

    #include <iostream>
    #include <memory>
    #include <cstring>
    using namespace std;

    int main()
    {
        shared_ptr<string> sp(new string{ "Hello world" });
        shared_ptr<string> sp2(sp);

        sp->insert(strlen("Hello "), "C++ ");
        cout << *sp << '\n' <<
                *sp2 << '\n';
    }
    /*
        Displays:
            Hello C++ world
            Hello C++ world
    */

18.4.4: Operators and members

• shared_ptr &operator=(shared_ptr<Type> const &other):
  Copy assignment: the reference count of the operator's left hand side operand is reduced. If the reference count decays to zero the dynamically allocated memory controlled by the left hand side operand is deleted. Then it shares the information with the operator's right hand side operand, incrementing the information's reference count.
• shared_ptr &operator=(shared_ptr<Type> &&tmp):
  Move assignment: the reference count of the operator's left hand side operand is reduced. If the reference count decays to zero the dynamically allocated memory controlled by the left hand side operand is deleted. Then it grabs the information controlled by the operator's right hand side operand which is turned into a 0-pointer.
• operator bool() const:
  If the shared_ptr actually points to memory true is returned, otherwise, false is returned.
• Type &operator*():
  A reference to the information stored in the shared_ptr object is returned. It acts like a normal pointer.
• Type *operator->():
  A pointer to the information controlled by the shared_ptr object is returned. Example:

  shared_ptr<string> sp{ new string{ "hello" } };
  cout << sp->c_str() << '\n';
  

The following member function member functions are supported:

• Type *get():
  A pointer to the information controlled by the shared_ptr object is returned. It acts like operator->. The returned pointer can be inspected. If it is zero the shared_ptr object does not point to any memory.
• Deleter &get_deleter():
  A reference to the shared_ptr's deleter (function or function object) is returned.
• void reset(Type *):
  The reference count of the information controlled by the shared_ptr object is reduced and if it decays to zero the memory it points to is deleted. Thereafter the object's information will refer to the argument that is passed to the function, setting its shared count to 1. It can also be called without argument, turning the object into a 0-pointer. This member function can be used to assign a new block of dynamically allocated memory to a shared_ptr object.
• void reset(Type *, DeleterType &&):
  This variant of the previous member accepts a specific Deleter type: if Type is a base-class and derived class objects are used, these derived class objects may require specific actions at destruction time. When the previous member is used, then eventually the newly assigned object's destructor is called without using an explicit deleter function. The current member ensures that by the time the shared counter has decayed to zero the provided deleter is used.
• void shared_ptr<Type>::swap(shared_ptr<Type> &&):
  Two identically typed shared_ptrs are swapped.
• bool unique() const:
  If the current object is the only object referring to the memory controlled by the object true is returned otherwise (including the situation where the object is a 0-pointer) false is returned.
• size_t use_count() const:
  The number of objects sharing the memory controlled by the object is returned.

18.4.5: Casting shared pointers

    struct Base
    {};
    struct Derived: public Base
    {};

As with unique_ptr, when defining a shared_ptr<Base> to store a newly allocated Derived class object, the returned Base * may be cast to a Derived * using a static_cast: polymorphism isn't required, and when resetting the shared_ptr or when the shared_ptr goes out of scope, no slicing occurs, and Derived's destructor is called (cf. section 18.3).

Of course, a shared_ptr<Derived> can easily be defined. Since a Derived object is also a Base object, a pointer to Derived can be considered a pointer to Base without using casts, but a static_cast could be used to force the interpretation of a Derived * to a Base *:

    Derived d;
    static_cast<Base *>(&d);

However, a plain static_cast cannot be used when initializing a shared pointer to a Base using the get member of a shared pointer to a Derived object. The following code snipped eventually results in an attempt to delete the dynamically allocated Base object twice:

    shared_ptr<Derived> sd{ new Derived };
    shared_ptr<Base> sb{ static_cast<Base *>(sd.get()) };

Since sd and sb point at the same object ~Base will be called for the same object when sb goes out of scope and when sd goes out of scope, resulting in premature termination of the program due to a double free error.

These errors can be prevented using casts that were specifically designed for being used with shared_ptrs. These casts use specialized constructors that create a shared_ptr pointing to memory but shares ownership (i.e., a reference count) with an existing shared_ptr. These special casts are:

• std::static_pointer_cast<Base>(std::shared_ptr<Derived> ptr):
  A shared_ptr to a Base class object is returned. The returned shared_ptr refers to the base class portion of the Derived class to which the shared_ptr<Derived> ptr refers. Example:

  shared_ptr<Derived> dp{ new Derived };
  shared_ptr<Base> bp = static_pointer_cast<Base>(dp);
  

• std::const_pointer_cast<Class>(std::shared_ptr<Class const> ptr):
  A shared_ptr to a Class class object is returned. The returned shared_ptr refers to a non-const Class object whereas the ptr argument refers to a Class const object. Example:

  shared_ptr<Derived const> cp{ new Derived };
  shared_ptr<Derived> ncp = const_pointer_cast<Derived>(cp);
  

• std::dynamic_pointer_cast<Derived>(std::shared_ptr<Base> ptr):
  A shared_ptr to a Derived class object is returned. The Base class must have at least one virtual member function, and the class Derived, inheriting from Base may have overridden Base's virtual member(s). The returned shared_ptr refers to a Derived class object if the dynamic cast from Base * to Derived * succeeded. If the dynamic cast did not succeed the shared_ptr's get member returns 0. Example (assume Derived and Derived2 were derived from Base):

  shared_ptr<Base> bp(new Derived());
  cout << dynamic_pointer_cast<Derived>(bp).get() << ' ' <<
          dynamic_pointer_cast<Derived2>(bp).get() << '\n';
  

  The first get returns a non-0 pointer value, the second get returns 0.

18.4.6: Using `shared_ptr' objects for arrays

But like unique_ptrs, with shared_ptrs referring to arrays the dereferencing operator makes little sense while in these circumstances shared_ptr objects would benefit from index operators.

It is not difficult to create a class shared_array offering such facilities. The class template shared_array, derived from shared_ptr merely should provide an appropriate deleter to make sure that the array and its elements are properly destroyed. In addition it should define the index operator and, when applicable should declare the derefencing operators using delete.

Here is an example showing how shared_array can be defined and used:

    struct X
    {
        ~X()
        {
            cout << "destr\n";  // show the object's destruction
        }
    };
    template <typename Type>
    class shared_array: public shared_ptr<Type>
    {
        struct Deleter          // Deleter receives the pointer
        {                       // and calls delete[]
           void operator()(Type* ptr)
           {
              delete[] ptr;
           }
        };
        public:
            shared_array(Type *p)           // other constructors
            :                               // not shown here
                shared_ptr<Type>(p, Deleter())
            {}
            Type &operator[](size_t idx)    // index operators
            {
                return shared_ptr<Type>::get()[idx];
            }
            Type const &operator[](size_t idx) const
            {
                return shared_ptr<Type>::get()[idx];
            }
            Type &operator*() = delete;     // delete pointless members
            Type const &operator*() const = delete;
            Type *operator->() = delete;
            Type const *operator->() const = delete;
    };
    int main()
    {
        shared_array<X> sp{ new X[3] };
        sp[0] = sp[1];
    }

18.5: Smart `smart pointer' construction: `make_shared' and `make_unique'

    std::shared_ptr<string> sptr{ new std::string{ "hello world" } }

In such statements two memory allocation calls are used: one for the allocation of the std::string and one used interally by std::shared_ptr's constructor itself.

The two allocations can be combined into one single allocation (which is also slightly more efficient than explicitly calling shared_ptr's constructor) using the make_shared template. The function template std::make_shared has the following prototype:

    template<typename Type, typename ...Args>
    std::shared_ptr<Type> std::make_shared(Args ...args);

Before using make_shared the <memory> header file must be included.

This function template allocates an object of type Type, passing args to its constructor (using perfect forwarding, see section 22.5.2), and returns a shared_ptr initialized with the address of the newly allocated Type object.

Here is how the above sptr object can be initialized using std::make_shared. Notice the use of auto which frees us from having to specify sptr's type explicitly:

    auto sptr(std::make_shared<std::string>("hello world"));

After this initialization std::shared_ptr<std::string> sptr has been defined and initialized. It could be used as follows:

    std::cout << *sptr << '\n';

In addition to make_shared the function std::make_unique can be used. It can be used make_shared but returns a std::unique_ptr rather than a shared_ptr.

18.6: Classes having pointer data members

    class Filter
    {
        istream *d_in;
        ostream *d_out;
        public:
            Filter(char const *in, char const *out);
    };

Assume that Filter objects filter information read from *d_in and write the filtered information to *d_out. Using pointers to streams allows us to have them point at any kind of stream like istreams, ifstreams, fstreams or istringstreams. The shown constructor could be implemented like this:

    Filter::Filter(char const *in, char const *out)
    :
        d_in(new ifstream{ in }),
        d_out(new ofstream{ out })
    {
        if (!*d_in || !*d_out)
            throw "Input and/or output stream not available"s;
    }

Of course, the construction could fail. new could throw an exception; the stream constructors could throw exceptions; or the streams could not be opened in which case an exception is thrown from the constructor's body. Using a function try block helps. Note that if d_in's initialization throws, there's nothing to be worried about. The Filter object hasn't been constructed, its destructor is not called and processing continues at the point where the thrown exception is caught. But Filter's destructor is also not called when d_out's initialization or the constructor's if statement throws: no object, and hence no destructor is called. This may result in memory leaks, as delete isn't called for d_in and/or d_out. To prevent this, d_in and d_out must first be initialized to 0 and only then the initialization can be performed:

    Filter::Filter(char const *in, char const *out)
    try
    :
        d_in(0),
        d_out(0)
    {
        d_in = new ifstream{ in };
        d_out = new ofstream{ out };

        if (!*d_in || !*d_out)
            throw "Input and/or output stream not available"s;
    }
    catch (...)
    {
        delete d_out;
        delete d_in;
    }

This quickly gets complicated, though. If Filter harbors yet another data member of a class whose constructor needs two streams then that data cannot be constructed or it must itself be converted into a pointer:

    Filter::Filter(char const *in, char const *out)
    try
    :
        d_in(0),
        d_out(0)
        d_filterImp(*d_in, *d_out)    // won't work
    { ... }

    // instead:

    Filter::Filter(char const *in, char const *out)
    try
    :
        d_in(0),
        d_out(0),
        d_filterImp(0)
    {
        d_in = new ifstream(in);
        d_out = new ofstream(out);
        d_filterImp = new FilterImp(*d_in, *d_out);
        ...
    }
    catch (...)
    {
        delete d_filterImp;
        delete d_out;
        delete d_in;
    }

Although the latter alternative works, it quickly gets hairy. In situations like these smart pointers should be used to prevent the hairiness. By defining the stream pointers as (smart pointer) objects they will, once constructed, properly be destroyed even if the rest of the constructor's code throws exceptions. Using a FilterImp and two unique_ptr data members Filter's setup and its constructor becomes:

    class Filter
    {
        std::unique_ptr<std::ifstream> d_in;
        std::unique_ptr<std::ofstream> d_out;
        FilterImp d_filterImp;
        ...
    };

    Filter::Filter(char const *in, char const *out)
    try
    :
        d_in(new ifstream(in)),
        d_out(new ofstream(out)),
        d_filterImp(*d_in, *d_out)
    {
        if (!*d_in || !*d_out)
            throw "Input and/or output stream not available"s;
    }

We're back at the original implementation but this time without having to worry about wild pointers and memory leaks. If one of the member initializers throws the destructors of previously constructed data members (which are now objects) are always called.

As a rule of thumb: when classes need to define pointer data members they should define those pointer data members as smart pointers if there's any chance that their constructors throw exceptions.

18.7: Comparison classes

Comparison classes are required when implementing the spaceship operator, and to use them (or when declaring and implementing the spachip operator) the <compare> header file must be included.

The weak class types do not support substitutability. Substitutability means that if two objects one and two are equal (so: one == two is true), then fun(one) == fun(two) is also true. Here fun is any function that only uses public const members of its argument that return value types (as compared to pointer types) which also support comparisons (also called comparison-salient state).

The operators of comparison classes expect at least one argument of their own class types. The other argument may either also be of their own class types or it can be the value 0 (or any value that can be considered 0, like nullptr_t).

There are five comparison classes, primarily used when implementing spaceship operators:

• weak_equality, used for classes that only support the == and != operators, but not substitutability;
• strong_equality, used for classes that only support the == and != operators, as well as substitutability;
• partial_ordering used for classes that support all comparison operators, do not support substitutability, and can be used when comparing incomparable arguments (i.e., all comparison operators return false);
• weak_ordering used for classes that support all comparison operators and that do not support substitutability;
• strong_ordering used for classes that support all comparison operators as well as substitutability;

18.7.1: The class `weak_equality'

• weak_equality::equivalent, indicating equality;
• weak_equality::nonequivalent, indicating non-equality;

Note: at the current release of the Gnu C++ compiler (10.0.0) this class is not yet available in the <compare> header file.

18.7.2: The class `strong_equality'

• strong_equality::equal, indicating equality;
• strong_equality::equivalent, indicating equality;
• strong_equality::nonequal, indicating non-equality;
• strong_equality::nonequivalent, indicating non-equality;

Note: at the current release of the Gnu C++ compiler (10.0.0) this class is not yet available in the <compare> header file.

18.7.3: The class `partial_ordering'

The class partial_ordering provides free functions for all comparison operations (==, !=, <, <=, >, and >=) expecting partial_ordering arguments (one argument may be 0). It also defines four static objects which can be returned by the spaceship operator:

• partial_ordering::less, returned when the lhs operand of the spaceship operator should be ordered before the rhs operand;
• partial_ordering::equivalent, indicating equality: there is no ordering preference between the two operands of the spaceship operator;
• partial_ordering::greater, returned when the lhs operand of the spaceship operator should be ordered after the rhs operand;
• partial_ordering::unordered, returned when all comparison operators of a class implementing the spaceship operator should return false (so ==, !=, <, <=, >, and >= all return false).

As an example, consider road taxes. Trucks, cars, and motor cycles have to pay road taxes, but there's no road tax for bicycles. For ordering road taxes a class RoadTax may be used, defining the following spaceship operator (assuming that all types of vehicles are derived from a class Vehicle, and that Vehicle has a (virtual) member double roadTax() returning the amount of road tax that is due for the various types of vehicles; the amount is negative if no road tax is required):

    partial_ordering RoadTax::operator<=>(Vehicle const &lhs, 
                                          Vehicle const &rhs)
    {
        return
          lhs.roadTax() < 0 or rhs.roadTax() < 0 ?
                                          partial_ordering::unordered :
          lhs.roadTax() < rhs.roadTax() ? partial_ordering::less      :
          lhs.roadTax() > rhs.roadTax() ? partial_ordering::greater   :
                                          partial_ordering::equivalent;
    }

18.7.4: The class `weak_ordering'

The class weak_ordering differs from the partial_ordering class in that unordered cannot be used as a comparison result. Like the class partial_ordering it provides free functions for all comparison operations (==, !=, <, <=, >, and >=) expecting partial_ordering arguments (one argument may be 0). It also defines three static objects which can be returned by the spaceship operator:

• weak_ordering::less, returned when the lhs operand of the spaceship operator should be ordered before the rhs operand;
• weak_ordering::equal, indicating equality: there is no ordering preference between the two operands of the spaceship operator;
• weak_ordering::greater, returned when the lhs operand of the spaceship operator should be ordered after the rhs operand;

The example in the previous section can easily adapted to the weak_ordering comparison class: if the roadTax members of vehicles for which no road tax is due return zero then RoadTax's spaceship operator can be implemented this way:

    weak_ordering RoadTax::operator<=>(Vehicle const &lhs, 
                                          Vehicle const &rhs)
    {
        return
          lhs.roadTax() < rhs.roadTax() ? weak_ordering::less      :
          lhs.roadTax() > rhs.roadTax() ? weak_ordering::greater   :
                                          weak_ordering::equal;
    }

18.7.5: The class `strong_ordering'

The class strong_ordering provides free functions for all comparison operations (==, !=, <, <=, >, and >=) expecting partial_ordering arguments (one argument may be 0). It also defines three static objects which can be returned by the spaceship operator:

• strong_ordering::less, returned when the lhs operand of the spaceship operator should be ordered before the rhs operand;
• strong_ordering::equal, indicating equality: there is no ordering preference between the two operands of the spaceship operator;
• strong_ordering::greater, returned when the lhs operand of the spaceship operator should be ordered after the rhs operand;

An example where the class strong_ordering was used has already been provided in section 11.6.2, where the spaceship operator itself was introduced.

18.8: Regular Expressions

Before using the specific C++ implementations of regular expressions the header file <regex> must be included.

Regular expressions are extensively documented elsewhere (e.g., regex(7), Friedl, J.E.F Mastering Regular Expressions, O'Reilly). The reader is referred to these sources for a refresher on the topic of regular expressions. In essence, regular expressions define a small meta-language recognizing textual units (like `numbers', `identifiers', etc.). They are extensively used in the context of lexical scanners (cf. section 25.6.1) when defining the sequence of input characters associated with tokens. But they are also intensively used in other situations. Programs like sed(1) and grep(1) use regular expressions to find pieces of text in files having certain characteristics, and a program like perl(1) adds some `sugar' to the regular expression language, simplifying the construction of regular expressions. However, though extremely useful, it is also well known that regular expressions tend to be very hard to read. Some even call the regular expression language a write-only language: while specifying a regular expression it's often clear why it's written in a particular way. But the opposite, understanding what a regular expression is supposed to represent if you lack the proper context, can be extremely difficult. That's why, from the onset and as a rule of thumb, it is stressed that an appropriate comment should be provided, with each regular expression, as to what it is supposed to match.

In the upcoming sections first a short overview of the regular expression language is provided, which is then followed by the facilities C++ is currently offering for using regular expressions. These facilities mainly consist of classes helping you to specify regular expression, matching them to text, and determining which parts of the text (if any) match (parts of) the text being analyzed.

18.8.1: The regular expression mini language

C++'s default definition of regular expressions distinguishes the following atoms:

• x: the character `x';

• .: any character except for the newline character;

• [xyz]: a character class; in this case, either an `x', a `y', or a `z' matches the regular expression. See also the paragraph about character classes below;

• [abj-oZ]: a character class containing a range of characters; this regular expression matches an `a', a `b', any letter from `j' through `o', or a `Z'. See also the paragraph about character classes below;

• [^A-Z]: a negated character class: this regular expression matches any character but those in the class beyond ^. In this case, any character except for an uppercase letter. See also the paragraph about character classes below;

• [:predef:]: a predefined set of characters. See below for an overview. When used, it is interpreted as an element in a character class. It is therefore always embedded in a set of square brackets defining the character class (e.g., [[:alnum:]]);

• \X: if X is `a', `b', `f', `n', `r', `t', or `v', then the ANSI-C interpretation of `\x'. Otherwise, a literal `X' (used to escape operators such as *);

• (r): the regular expression r. It is used to override precedence (see below), but also to define r as a marked sub-expression whose matching characters may directly be retrieved from, e.g., a std::smatch object (cf. section 18.8.3);

• (?:r): the regular expression r. It is used to override precedence (see below), but it is not regarded as a marked sub-expression;

In addition to these basic atoms, the following special atoms are available (which can also be used in character classes):

• \s: a whitespace character;

• \S: any character but a whitespace character;

• \d: a decimal digit character;

• \D: any character but a decimal digit character;

• \w: an alphanumeric character or an underscore (_) character;

• \W: any character but an alphanumeric character or an underscore (_) character.

Atoms may be concatenated. If r and s are atoms then the regular expression rs matches a target text if the target text matches r and s, in that order (without any intermediate characters inside the target text). E.g., the regular expression [ab][cd] matches the target text ac, but not the target text a:c.

Atoms may be combined using operators. Operators bind to the preceding atom. If an operator should operate on multiple atoms the atoms must be surrounded by parentheses (see the last element in the previous itemization). To use an operator character as an atom it can be escaped. Eg., * represents an operator, \* the atom character star. Note that character classes do not recognize escape sequences: [\*] represents a character class consisting of two characters: a backslash and a star.

The following operators are supported (r and s represent regular expression atoms):

• r*: zero or more rs;

• r+: one or more rs;

• r?: zero or one rs (that is, an optional r);

• r{m, n}: where 1 <= m <= n: matches `r' at least m, but at most n times;

• r{m,}: where 1 <= m: matches `r' at least m times;

• r{m}: where 1 <= m: matches `r' exactly m times;

• r|s: matches either an `r' or an `s'. This operator has a lower priority than any of the multiplication operators;

• ^r : ^ is a pseudo operator. This expression matches `r', if appearing at the beginning of the target text. If the ^-character is not the first character of a regular expression it is interpreted as a literal ^-character;

• r$: $ is a pseudo operator. This expression matches `r', if appearing at the end of the target text. If the $-character is not the last character of a regular expression it is interpreted as a literal $-character;

When a regular expression contains marked sub-expressions and multipliers, and the marked sub-expressions are multiply matched, then the target's final sub-string matching the marked sub-expression is reported as the text matching the marked sub-expression. E.g, when using regex_search (cf. section 18.8.4.3), marked sub-expression (((a|b)+\s?)), and target text a a b, then a a b is the fully matched text, while b is reported as the sub-string matching the first and second marked sub-expressions.

18.8.1.1: Character classes

To add a closing bracket to a character class use [] immediately following the initial open-bracket, or start with [^] for a negated character class not containing the closing bracket. Minus characters are used to define character ranges (e.g., [a-d], defining [abcd]) (be advised that the actual range may depend on the locale being used). To add a literal minus character to a character class put it at the very beginning ([-, or [^-) or at the very end (-]) of a character class.

Once a character class has started, all subsequent characters are added to the class's set of characters, until the final closing bracket (]) has been reached.

In addition to characters and ranges of characters, character classes may also contain predefined sets of character. They are:

         [:alnum:] [:alpha:] [:blank:]
         [:cntrl:] [:digit:] [:graph:]
         [:lower:] [:print:] [:punct:]
         [:space:] [:upper:] [:xdigit:]

These predefined sets designate sets of characters equivalent to the corresponding standard C isXXX function. For example, [:alnum:] defines all characters for which isalnum(3) returns true.

18.8.2: Defining regular expressions: std::regex

The types std::regex and std::wregex define regular expression patterns. They define, respectively the types basic_regex<char> and basic_regex<wchar_t> types. Below, the class regex is used, but in the examples wregex could also have been used.

Regular expression facilities were, to a large extent, implemented through templates, using, e.g., the basic_string<char> type (which is equal to std::string). Likewise, generic types like OutputIter (output iterator) and BidirConstIter (bidirectional const iterator) are used with several functions. Such functions are function templates. Function templates determine the actual types from the arguments that are provided at call-time.

These are the steps that are commonly taken when using regular expressions:

• First, a regular expression is defined. This involves defining or modifying a regex object.
• Then the regular expression is provided with a target text, which may result in sections of the target text matching the regular expression.
• The sections of the target text matching (or not matching) the regular expression are retrieved to be processed elsewhere, or:
• The sections of the target text matching (or not matching) the regular expression are directly modified by existing regular expression facilities, after which the modified target text may be processed elsewhere.

The way regex objects handle regular expressions can be configured using a bit_or combined set of std::regex_constants values, defining a regex::flag_type value. These regex_constants are:

• std::regex_constants::awk:
  awk(1)'s (POSIX) regular expression grammar is used to specify regular exressions (e.g., regular expressions are delimited by /-characters, like /\w+/; for further details and for details of other regular expression grammars the reader should consult the man-pages of the respective programs);

• std::regex_constants::basic:
  the basic POSIX regular expression grammar is used to specify regular expressions;

• std::regex_constants::collate:
  the character range operator (-) used in character classes defines a locale sensitive range (e.g., [a-k]);

• std::regex_constants::ECMAScript:
  this flag_type is used by default by regex constructors. The regular expression uses the Modified ECMAScript regular expression grammar;

• std::regex_constants::egrep:
  egrep(1)'s (POSIX) regular expression grammar is used to specify regular expressions. This is the same grammar as used by regex_constants::extended, with the addition of the newline character ('\n') as an alternative for the '|'-operator;

• std::regex_constants::extended:
  the extended POSIX regular expression grammar is used to specify regular expressions;

• std::regex_constants::grep:
  grep(1)'s (POSIX) regular expression grammar is used to specify regular expressions. This is the same grammar as used by regex_constants::basic, with the addition of the newline character ('\n') as an alternative for the '|'-operator;

• std::regex_constants::icase:
  letter casing in the target string is ignored. E.g., the regular expression A matches a and A;

• std::regex_constants::nosubs:
  When performing matches, all sub-expressions ((expr)) are treated as non-marked (?:expr);

• std::regex_constants::optimize:
  optimizes the speed of matching regular expressions, at the cost of slowing down the construction of the regular expression somewhat. If the same regular expression object is frequently used then this flag may substantially improve the speed of matching target texts;

Constructors

The default, move and copy constructors are available. Actually, the default constructor defines one parameter of type regex::flag_type, for which the value regex_constants::ECMAScript is used by default.

• regex():
  the default constructor defines a regex object not containing a regular expression;

• explicit regex(char const *pattern):
  defines a regex object containing the regular expression found at pattern;

• regex(char const *pattern, std::size_t count):
  defines a regex object containing the regular expression found at the first count characters of pattern;

• explicit regex(std::string const &pattern):
  defines a regex object containing the regular expression found at pattern. This constructor is defined as a member template, accepting a basic_string-type argument which may also use non-standard character traits and allocators;

• regex(ForwardIterator first, ForwardIterator last):
  defines a regex object containing the regular expression found at the (forward) iterator range [first, last). This constructor is defined as a member template, accepting any forward iterator type (e.g., plain char pointers) which can be used to define the regular expression's pattern;

• regex(std::initializer_list<Char> init):
  defines a regex object containing the regular expression from the characters in the initializer list init.

  Here are some examples:

      std::regex re("\\w+");      // matches a sequence of alpha-numeric
                                  // and/or underscore characters 
  
      std::regex re{'\\', 'w', '+'} ;     // idem
  
      std::regex re(R"(\w+xxx")", 3);     // idem
  

Member functions

• regex &operator=(RHS):
  The copy and move assignment operators are available. Otherwise, RHS may be:
  • an NTBS (of type char const *);
  • a std::string const & (or any compatible std::basic_string);
  • a std::initializer_list<char>;

• regex &assign(RHS):
  This member accepts the same arguments as regex's constructors, including the (optional) regex_constants values;

• regex::flag_type flag() const:
  Returns the regex_constants flags that are active for the current regex object. E.g.,

      int main()
      {
          regex re;
  
          regex::flag_type flags = re.flags();
      
          cout <<                                 // displays: 16 0 0
              (re.flags() & regex_constants::ECMAScript) << ' '  <<
              (re.flags() & regex_constants::icase) << ' '  <<
              (re.flags() & regex_constants::awk) << ' '  << '\n';
      }
  

  Note that when a combination of flag_type values is specified at construction-time that only those flags that were specified are set. E.g., when re(regex_constants::icase) would have been specified the cout statement would have shown 0 1 0. It's also possible to specify conflicting combinations of flag-values like regex_constants::awk | regex_constants::grep. The construction of such regex objects succeeds, but should be avoided.

• locale_type get_loc() const:
  Returns the locale that is associated with the current regex object;

• locale_type imbue(locale_type locale):
  Replaces the regex object's current locale setting with locale, returning the replaced locale;

• unsigned mark_count() const:
  The number of marked sub-expressions in the regex object is returned. E.g.,

      int main()
      {
          regex re("(\\w+)([[:alpha:]]+)"); 
          cout << re.mark_count() << '\n';        // displays: 2
      }
  

• void swap(regex &other) noexcept:
  Swaps the current regex object with other. Also available as a free function: void swap(regex &lhs, regex &rhs), swapping lhs and rhs.

18.8.3: Retrieving matches: std::match_results

• regex_match merely matches a target text against a regular expression, informing the caller whether a match was found or not;
• regex_search also matches a target text against a regular expression, but allows retrieval of matches of marked sub-expressions (i.e., parenthesized regular expressions);
• regex_replace matches a target text against a regular expression, and replaces pieces of matched sections of the target text by another text.

These functions must be provided with a target text and a regex object (which is not modified by these functions). Usually another argument, a std::match_results object is also passed to these functions, to contain the results of the regular expression matching procedure.

Before using the match_results class the <regex> header file must be included.

Examples of using match_results objects are provided in section 18.8.4. This and the next section are primarily for referential purposes.

Various specializations of the class match_results exist. The specialization that is used should match the specializations of the used regex class. E.g., if the regular expression was specified as a char const * the match_results specialization should also operate on char const * values. The various specializations of match_results have been given names that can easily be remembered, so selecting the appropriate specialization is simple.

The class match_results has the following specializations:

• cmatch:
  defines match_results<char const *>, using a char const * type of iterator. It should be used with a regex(char const *) regular expression specification;
• wcmatch:
  defines match_results<wchar_ const *>, using a wchar_t const * type of iterator. It should be used with a regex(wchar_t const *) regular expression specification;
• smatch:
  defines match_results<std::string::const_iterator>, using a std::string::const_iterator type of iterator. It should be used with a regex(std::string const &) regular expression specification;
• wsmatch:
  defines match_results<std::wstring::const_iterator>, using a std::wstring::const_iterator type of iterator. It should be used with a regex(wstring const &) regular expression specification.

Constructors

The default, copy, and move constructors are available. The default constructor defines an Allocator const & parameter, which by default is initialized to the default allocator. Normally, objects of the class match_results receive their match-related information by passing them to the above-mentioned functions, like regex_match. When returning from these functions members of the class match_results can be used to retrieve specific results of the matching process.

Member functions

• match_results &operator=:
  The copy and move assignment operators are available;

• std::string const &operator[](size_t idx) const:
  Returns a (const) reference to sub-match idx. With idx value 0 a reference to the full match is returned. If idx >= size() (see below) a reference to an empty sub-range of the target string is returned. The behavior of this member is undefined if the member ready() (see below) returns false;

• Iterator begin() const:
  Returns an iterator to the first sub-match. Iterator is a const-iterator for const match_results objects;

• Iterator cbegin() const:
  Returns an iterator to the first sub-match. Iterator is a const-iterator;

• Iterator cend() const:
  Returns an iterator pointing beyond the last sub-match. Iterator is a const-iterator;

• Iterator end() const:
  Returns an iterator pointing beyond the last sub-match. Iterator is a const-iterator for const match_results objects;

• ReturnType format(Parameters) const:
  As this member requires a fairly extensive description, it would break the flow of the current overview. This member is used in combination with the regex_replace function, and it is therefore covered in detail in that function's section (18.8.4.5);

• allocator_type get_allocator() const:
  Returns the object's allocator;

• bool empty() const:
  Returns true if the match_results object contains no matches (which is also returned after merely using the default constructor). Otherwise it returns false;

• int length(size_t idx = 0) const:
  Returns the length of sub-match idx. By default the length of the full match is returned. If idx >= size() (see below) 0 is returned;

• size_type max_size() const:
  Returns the maximum number of sub-matches that can be contained in a match_results object. This is an implementation dependent constant value;

• int position(size_t idx = 0) const:
  Returns the offset in the target text of the first character of sub-match idx. By default the position of the first character of the full match is returned. If idx >= size() (see below) -1 is returned;

• std::string const &prefix() const:
  Returns a (const) reference to a sub-string of the target text that ends at the first character of the full match;

• bool ready() const:
  No match results are available from a default constructed match_results object. It receives its match results from one of the mentioned matching functions. Returns true once match results are available, and false otherwise.

• size_type size() const:
  Returns the number of sub-matches. E.g., with a regular expression (abc)|(def) and target defcon three submatches are reported: the total match (def); the empty text for (abc); and def for the (def) marked sub-expression.

  Note: when multipliers are used only the last match is counted and reported. E.g., for the pattern (a|b)+ and target aaab two sub-matches are reported: the total match aaab, and the last match (b);

• std::string str(size_t idx = 0) const:
  Returns the characters defining sub-match idx. By default this is the full match. If idx >= size() (see below) an empty string returned;

• std::string const &suffix() const:
  Returns a (const) reference to a sub-string of the target text that starts beyond the last character of the full match;

• void swap(match_results &other) noexcept:
  Swaps the current match_results object with other. Also available as a free function: void swap(match_results &lhs, match_results &rhs), swapping lhs and rhs.

18.8.4: Regular expression matching functions

There are three major families of functions that can be used to match a target text against a regular expression. Each of these functions, as well as the match_results::format member, has a final std::regex_constants::match_flag_type parameter (see the next section), which is given the default value regex_constants::match_default which can be used to fine-tune the way the regular expression and the matching process is being used. This final parameter is not explicitly mentioned with the regular expression matching functions or with the format member. The three families of functions are:

• bool std::regex_match(Parameters):
  This family of functions is used to match a regular expression against a target text. Only if the regular expression matches the full target text true is returned; otherwise false is returned. Refer to section 18.8.4.2 for an overview of the available overloaded regex_match functions;

• bool std::regex_search(Parameters):
  This family of functions is also used to match a regular expression against a target text. This function returns true once the regular expression matches a sub-string of the target text; otherwise false is returned. See below for an overview of the available overloaded regex_search functions;

• ReturnType std::regex_replace(Parameters):
  This family of functions is used to produce modified texts, using the characters of a target string, a regex object and a format string. This member closely resembles the functionality of the match_results::format member discussed in section 18.8.4.4.

18.8.4.1: The std::regex_constants::match_flag_type flags

The match_flag_type enumeration defines the following values (below, `[first, last)' refers to the character sequence being matched).

• format_default (not a bit-mask value, but a default value which is equal to 0). With just this specification ECMAScript rules are used to construct strings in std::regex_replace;

• format_first_only: std::regex_replace only replaces the first match;

• format_no_copy: non-matching strings are not passed to the output by std::regex_replace;

• format_sed: POSIX sed(1) rules are used to construct strings in std::regex_replace;

• match_any: if multiple matches are possible, then any match is an acceptable result;

• match_continuous: sub-sequences are only matching if they start at first;

• match_not_bol: the first character in [first, last) is treated as an ordinary character: ^ does not match [first, first);

• match_not_bow: \b does not match [first, first);

• match_default (not a bit-mask value, but equal to 0): the default value of the final argument that's passed to the regular expression matching functions and match_results::format member. ECMAScript rules are used to construct strings in std::regex_replace;

• match_not_eol: the last character in [first, last) is treated as an ordinary character: $ does not match [last,last);

• match_not_eow: \b does not match [last, last);

• match_not_null: empty sequences are not considered matches;

• match_prev_avail: --first refers to a valid character position. When specified match_not_bol and match_not_bow are ignored;

18.8.4.2: Matching full texts: std::regex_match

The following overloaded variants of this function are available:

• bool regex_match(BidirConstIter first, BidirConstIter last, std::match_results &results, std::regex const &re):
  BidirConstIter is a bidirectional const iterator. The range [first, last) defines the target text. The match results are returned in results. The types of the iterators must match the type of the match_results that's used. E.g., a cmatch should be used if the iterators are of char const * types, and a smatch should be used if the iterators are of string::const_iterator types. Similar correspondence requirements hold true for the other overloaded versions of this function;

• bool regex_match(BidirConstIter first, BidirConstIter last, std::regex const &re):
  this function behaves like the previous function, but does not return the results of the matching process in a match_results object;

• bool regex_match(char const *target, std::match_results &results, std::regex const &re):
  this function behaves like the first overloaded variant, using the characters in target as its target text;

• bool regex_match(char const *str, std::regex const &re):
  this function behaves like the previous function but does not return the match results;

• bool regex_match(std::string const &target, std::match_results &results, std::regex const &re):
  this function behaves like the first overloaded variant, using the characters in target as its target text;

• bool regex_match(std::string const &str, std::regex const &re):
  this function behaves like the previous function but does not return the match results;

• bool regex_match(std::string const &&, std::match_results &, std::regex &) = delete (the regex_match function does not accept temporary string objects as target strings, as this would result in invalid string iterators in the match_result argument.)

    #include <iostream>
    #include <regex>
    
    using namespace std;
    
    int main(int argc, char const **argv)
    {
        regex re("(\\d{5})[[:alpha:]]+"); 

        cmatch results;

        if (not regex_match(argv[1], results, re))
            cout << "No match\n";
        else
            cout << "size: " << results.size() << ": " << 
                    results.str(1) << " -- " << results.str() << '\n';
    }

18.8.4.3: Partially matching text: std::regex_search

The following overloaded variants of this function are available:

• bool regex_search(BidirConstIter first, BidirConstIter last, std::match_results &results, std::regex const &re):
  BidirConstIter is a bidirectional const iterator. The range [first, last) defines the target text. The match results are returned in results. The types of the iterators must match the type of the match_results that's used. E.g., a cmatch should be used if the iterators are of char const * types, and a smatch should be used if the iterators are of string::const_iterator types. Similar correspondence requirements hold true for the other overloaded versions of this function;

• bool regex_search(BidirConstIter first, BidirConstIter last, std::regex const &re):
  this function behaves like the previous function, but does not return the results of the matching process in a match_results object;

• bool regex_search(char const *target, std::match_results &results, std::regex const &re):
  this function behaves like the first overloaded variant, using the characters in target as its target text;

• bool regex_search(char const *str, std::regex const &re):
  this function behaves like the previous function but does not return the match results;

• bool regex_search(std::string const &target, std::match_results &results, std::regex const &re):
  this function behaves like the first overloaded variant, using the characters in target as its target text;

• bool regex_search(std::string const &str, std::regex const &re):
  this function behaves like the previous function but does not return the match results;

• bool regex_search(std::string const &&, std::match_results &, std::regex &) = delete:
  the regex_search function does not accept temporary string objects as target strings, as this would result in invalid string iterators in the match_result argument.

     1: #include <iostream>
     2: #include <string>
     3: #include <regex>
     4: 
     5: using namespace std;
     6: 
     7: int main()
     8: {
     9:     while (true)
    10:     {
    11:         cout << "Enter a pattern or plain Enter to stop: ";
    12: 
    13:         string pattern;
    14:         if (not getline(cin, pattern) or pattern.empty())
    15:             break;
    16: 
    17:         regex re(pattern);
    18:         while (true)
    19:         {
    20:             cout << "Enter a target text for `" << pattern << "'\n"
    21:                     "(plain Enter for the next pattern): ";
    22: 
    23:             string text;
    24:             if (not getline(cin, text) or text.empty())
    25:                 break;
    26: 
    27:             smatch results;
    28:             if (not regex_search(text, results, re))
    29:                 cout << "No match\n";
    30:             else
    31:             {
    32:                 cout << "Prefix: "  << results.prefix() << "\n"
    33:                         "Match:  "  << results.str()    << "\n"
    34:                         "Suffix: "  << results.suffix() << "\n";
    35:                 for (size_t idx = 1; idx != results.size(); ++idx)
    36:                     cout << "Match " << idx << " at offset " <<
    37:                                 results.position(idx) << ": " <<
    38:                                 results.str(idx) << '\n';
    39:             }
    40:         }
    41:     }
    42: }

18.8.4.4: The member std::match_results::format

The format member operates on (sub-)matches contained in a match_results object, using a format string, and producing text in which format specifiers (like $&) are replaced by matching sections of the originally provided target text. In addition, the format member recognizes all standard C escape sequences (like \n). The format member is used to create text that is modified with respect to the original target text.

As a preliminary illustration: if results is a match_results object and match[0] (the fully matched text) equals `hello world', then calling format with the format string this is [$&] produces the text this is [hello world]. Note the specification $& in this format string: this is an example of a format specifier. Here is an overview of all supported format specifiers:

• $`: corresponds to the text returned by the prefix member: all characters in the original target text up to the first character of the fully matched text;
• $&: corresponds to the fully matched text (i.e., the text returned by the match_results::str member);
• $n: (where n is an integral natural number): corresponds to the text returned bu operator[](n);
• $': corresponds to the text returned by the suffix member: all characters in the original target string beyond the last character of the fully matched text;
• $$: corresponds to the single $ character.

Four overloaded versions of the format members are available. All overloaded versions define a final regex_constants::match_flag_type parameter, which is by default initialized to match_default. This final parameter is not explicitly mentioned in the following coverage of the format members.

To further illustrate the way the format members can be used it is assumed that the following code has been executed:

     1:     regex re("([[:alpha:]]+)\\s+(\\d+)");  // letters blanks digits
     2: 
     3:     smatch results;
     4:     string target("this value 1024 is interesting");
     5: 
     6:     if (not regex_search(target, results, re))
     7:         return 1;

The first two overloaded format functions expect an output-iterator to where the formatted text is written. These overloaded members return the final output iterator, pointing just beyond the character that was last written.

• OutputIter format(OutputIter out, char const *first, char const *last) const:
  the characters in the range [first, last) are applied to the sub-expressions stored in the match_results object, and the resulting string is inserted at out. An illustration is provided with the next overloaded version;

• OutputIter format(OutputIter out, std::string const &fmt) const:
  the content of fmt is applied to the sub-expressions stored in the match_results object, and the resulting string is inserted at out. The next line of code inserts the value 1024 into cout (note that fmt must be a std::string, hence the explicit use of the string constructor):

  results.format(ostream_iterator<char>(cout, ""), "$2"s);
  

The remaining two overloaded format members expect a std::string or an NTBS defining the format string. Both members return a std::string containing the formatted text:

• std::string format(std::string const &fmt) const
• std::string format(char const *fmt) const

    string reverse(results.format("$2 and $1"));

18.8.4.5: Modifying target strings: std::regex_replace

• OutputIt regex_replace(OutputIter out, BidirConstIter first, BidirConstIter last, std::regex const &re, std::string const &fmt):
  OutputIter is an output iterator; BidirConstIter a bidirectional const iterator.

  The function returns the possibly modified text in an iterator range [out, retvalue), where out is the output iterator passed as the first argument to regex_replace, and retvalue is the output iterator returned by regex_replace.

  The function matches the text at the range [first, last) against the regular expression stored in re. If the regular expression does not match the target text in the range [first, last) then the target text is literally copied to out. If the regular expression does match the target text then

  • first, the match result's prefix is copied to out. The prefix equals the initial characters of the target text up to the very first character of the fully matched text.
  • next, the matched text is replaced by the content of the fmt format string, in which the format specifiers can be used that were described in the previous section (section 18.8.4.4), and the replaced text is copied to out;
  • finally, the match result's suffix is copied to out. The suffix equals all characters of the target text beyond the last character of the matched text.
  The workings of regex_replace is illustrated in the next example:

   1:     regex re("([[:alpha:]]+)\\s+(\\d+)");  // letters blanks digits
   2: 
   3:     string target("this value 1024 is interesting");
   4: 
   5:     regex_replace(ostream_iterator<char>(cout, ""), target.begin(),
   6:                     target.end(), re, "$2"s);
  

  In line 5 regex_replace is called. Its format string merely contains $2, matching 1024 in the target text. The prefix ends at the word value, the suffix starts beyond 1024, so the statement in line 5 inserts the text

  this 1024 is interesting
  

  into the standard output stream.

• OutputIt regex_replace( OutputIter out, BidirConstIter first, BidirConstIter last, std::regex const &re, char const *fmt):
  This variant behaves like the first variant. When using, in the above example, "$2" instead of "$2"s, then this variant would have been used;

• std::string regex_replace(std::string const &str, std::regex const &re, std::string const &fmt):
  This variant returns a std::string containing the modified text, and expects a std::string containing the target text. Other than that, it behaves like the first variant. To use this overloaded variant in the above example the statement in line 5 could have been replaced by the following statement, initializing the string result:

  string result(regex_replace(target, re, "$2"s));
  

• std::string regex_replace(std::string const &str, std::regex const &re, char const *fmt):
  After changing, in the above statement, "$2"s into "$2", this variant is used, behaving exactly like the previous variant;

• std::string regex_replace(char const *str, std::regex const &re, std::string const &fmt):
  This variant uses a char const * to point to the target text, and behaves exactly like the previous but one variant;

• std::string regex_replace(char const *str, std::regex const &re, char const *fmt):
  This variant also uses a char const * to point to the target text, and also behaves exactly like the previous but one variant;

18.9: Randomization and Statistical Distributions

The STL offers several standard mathematical (statistical) distributions. These distributions allow programmers to obtain randomly selected values from a selected distribution.

These statistical distributions need to be provided with a random number generating object. Several of such random number generating objects are provided, extending the traditional rand function that is part of the C standard library.

These random number generating objects produce pseudo-random numbers, which are then processed by the statistical distribution to obtain values that are randomly selected from the specified distribution.

Although the STL offers various statistical distributions their functionality is fairly limited. The distributions allow us to obtain a random number from these distributions, but probability density functions or cumulative distribution functions are currently not provided by the STL. These functions (distributions as well as the density and the cumulative distribution functions) are, however, available in other libraries, like the boost math library (specifically:
http://www.boost.org/doc/libs/1_44_0/libs/math/doc/sf_and_dist/html/index.html).

It is beyond the scope of the C++ Annotations to discuss the mathematical characteristics of the various statistical distributions. The interested reader is referred to the pertinent mathematical textbooks (like Stuart and Ord's (2009) Kendall's Advanced Theory of Statistics, Wiley) or to web-locations like http://en.wikipedia.org/wiki/Bernoulli_distribution.

18.9.1: Random Number Generators

The linear_congruential_engine random number generator computes

    linear_congruential_engine<int, 10, 3, 13> lincon;

The linear_congruential generator may be seeded by providing its constructor with a seeding-argument. E.g., lincon(time(0)).

The subtract_with_carry_engine random number generator computes

    subtract_with_carry_engine<int, 13, 3, 13> subcar;

The subtract_with_carry_engine generator may be seeded by providing its constructor with a seeding-argument. E.g., subcar(time(0)).

The predefined mersenne_twister_engine mt19937 (predefined as type in the <random> header file) is used in the examples below. It can be constructed using `mt19937 mt' or it can be seeded by providing its constructor with an argument (e.g., mt19937 mt(time(0))). Its function call operator returns a random unsigned integral value.

Other ways to initialize the mersenne_twister_engine are beyond the scope of the C++ Annotations (but see Lewis et al. ( Lewis, P.A.W., Goodman, A.S., and Miller, J.M. (1969), A pseudorandom number generator for the System/360, IBM Systems Journal, 8, 136-146.) (1969)).

The random number generators may also be seeded by calling their members seed accepting unsigned long values or generator functions (as in lc.seed(time(0)), lc.seed(mt)).

The random number generators offer members min and max returning, respectively, their minimum and maximum values (inclusive). If a reduced range is required the generators can be nested in a function or class adapting the range.

Here's a small example showing how the mersenne_twister_engine mt19937 can be used to generate random numbers:

    #include <iostream>
    #include <ctime>
    #include <random>

    using namespace std;

        // arguments: 1st: number of random numbers to generate
        //            2nd: lowest positve random number,
        //            3rd: highest positive random number

    int main(int argc, char **argv)
    {
        mt19937 mt( time(0) );      // seed with the current time in secs.

        for (
            size_t nGenerate = stoul(argv[1]), lowest = stoul(argv[2]),
                                               mod = stoul(argv[3]) + 1 - lowest;
                nGenerate--;
        )
            cout << (lowest + mt() % mod) << ' ';

        cout << '\n';
    }

18.9.2: Statistical distributions

All distributions offer the following members (result_type refers to the type name of the values returned by the distribution):

• result_type max() const
  returns the distribution's least upper bound;
• result_type min() const
  returns the distribution's greatest lower bound;
• param_type param() const
  returns the object's param_type struct;
• void param(const param_type &param) redefines the parameters of the distribution;
• void reset(): clears all of its cached values;

All distributions support the following operators (distribution-name should be replaced by the name of the intended distribution, e.g., normal_distribution):

• template<typename URNG> result_type operator()(URNG &urng)
  returns the next random value from the statistical distribution, with the function object urng returning the next random number selected from a uniform random distribution;
• template<typename URNG> result_type operator()
(URNG &urng, param_type &param)
  returns the next random value from the statistical distribution initialized with the parameters provided by the param struct. The function object urng returns the next random number selected from a uniform random distribution;
• std::istream &operator>>(std::istream &in, distribution-name &object): The parameters of the distribution are extracted from an std::istream;
• std::ostream &operator<<(std::ostream &out, distribution-name const &bd): The parameters of the distribution are inserted into an std::ostream

The following example shows how the distributions can be used. Replacing the name of the distribution (normal_distribution) by another distribution's name is all that is required to switch distributions. All distributions have parameters, like the mean and standard deviation of the normal distribution, and all parameters have default values. The names of the parameters vary over distributions and are mentioned below at the individual distributions. Distributions offer members returning or setting their parameters.

Most distributions are defined as class templates, requiring the specification of a data type that is used for the function's return type. If so, an empty template parameter type specification (<>) will get you the default type. The default types are either double (for real valued return types) or int (for integral valued return types). The template parameter type specification must be omitted with distributions that are not defined as template classes.

Here is an example showing the use of the statistical distributions, applied to the normal distribution:

#include <iostream>
#include <ctime>
#include <random>
using namespace std;

int main()
{
    std::mt19937 engine(time(0));
    std::normal_distribution<> dist;

    for (size_t idx = 0; idx < 10; ++idx)
        std::cout << "a random value: " << dist(engine) << "\n";

    cout << '\n' <<
        dist.min() << " " << dist.max() << '\n';
}

18.9.2.1: Bernoulli distribution

The bernoulli distribution is not defined as a class template.

Defined types:

    using result_type = bool;
    struct param_type
    {
      explicit param_type(double prob = 0.5);
      double p() const;                     // returns prob
    };

Constructor and members:

• bernoulli_distribution(double prob = 0.5)
  constructs a bernoulli distribution with probability prob of returning true;
• double p() const
  returns prob;
• result_type min() const
  returns false;
• result_type max() const
  returns true;

18.9.2.2: Binomial distribution

The template type parameter IntType defines the type of the generated random value, which must be an integral type.

Defined types:

    using result_type =  IntType;
    struct param_type
    {
      explicit param_type(IntType trials, double prob = 0.5);
      IntType t() const;                    // returns trials
      double p() const;                     // returns prob
    };

Constructors and members and example:

• binomial_distribution<>(IntType trials = 1, double prob = 0.5) constructs a binomial distribution for trials experiments, each having probability prob of success.
• binomial_distribution<>(param_type const &param) constructs a binomial distribution according to the values stored in the param struct.
• IntType t() const
  returns trials;
• double p() const
  returns prob;
• result_type min() const
  returns 0;
• result_type max() const
  returns trials;

18.9.2.3: Cauchy distribution

The mean and standard deviation of the Cauchy distribution are undefined.

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType a = RealType(0),
                            RealType b = RealType(1));

        double a() const;
        double b() const;
    };

Constructors and members:

• cauchy_distribution<>(RealType a = RealType(0), RealType b = RealType(1)) constructs a cauchy distribution with specified a and b parameters.
• cauchy_distribution<>(param_type const &param) constructs a cauchy distribution according to the values stored in the param struct.
• RealType a() const
  returns the distribution's a parameter;
• RealType b() const
  returns the distribution's b parameter;
• result_type min() const
  returns the smallest positive result_type value;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.4: Chi-squared distribution

Note that even though the distribution's parameter n usually is an integral value, it doesn't have to be integral, as the chi_squared distribution is defined in terms of functions (exp and Gamma) that take real arguments (see, e.g., the formula shown in the <bits/random.h> header file, provided with the GNU g++ compiler distribution).

The chi-squared distribution is used, e.g., when testing the goodness of fit of an observed distribution to a theoretical one.

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType n = RealType(1));

        RealType n() const;
    };

Constructors and members:

• chi_squared_distribution<>(RealType n = 1) constructs a chi_squared distribution with specified number of degrees of freedom.
• chi_squared_distribution<>(param_type const &param) constructs a chi_squared distribution according to the value stored in the param struct;
• IntType n() const
  returns the distribution's degrees of freedom;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.5: Extreme value distribution

It has two parameters: a location parameter a and scale parameter b. See also
http://www.itl.nist.gov/div898/handbook/apr/section1/apr163.htm

Defined types:

using result_type = RealType;

struct param_type
{
    explicit param_type(RealType a = RealType(0),
                        RealType b = RealType(1));

    RealType a() const;     // the location parameter
    RealType b() const;     // the scale parameter
};

Constructors and members:

• extreme_value_distribution<>(RealType a = 0, RealType b = 1) constructs an extreme value distribution with specified a and b parameters;
• extreme_value_distribution<>(param_type const &param) constructs an extreme value distribution according to the values stored in the param struct.
• RealType a() const
  returns the distribution's location parameter;
• RealType stddev() const
  returns the distribution's scale parameter;
• result_type min() const
  returns the smallest positive value of result_type;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.6: Exponential distribution

Its parameter prob defines the distribution's lambda parameter, called its rate parameter. Its expected value and standard deviation are both 1 / lambda.

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType lambda = RealType(1));

        RealType lambda() const;
    };

Constructors and members:

• exponential_distribution<>(RealType lambda = 1) constructs an exponential distribution with specified lambda parameter.
• exponential_distribution<>(param_type const &param) constructs an exponential distribution according to the value stored in the param struct.
• RealType lambda() const
  returns the distribution's lambda parameter;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.7: Fisher F distribution

It is characterized by two parameters, being the degrees of freedom of the two chi-squared distributions.

Note that even though the distribution's parameter n usually is an integral value, it doesn't have to be integral, as the Fisher F distribution is constructed from Chi-squared distributions that accept a non-integral parameter value (see also section 18.9.2.4).

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType m = RealType(1),
                            RealType n = RealType(1));

        RealType m() const; // The degrees of freedom of the nominator
        RealType n() const; // The degrees of freedom of the denominator
    };

Constructors and members:

• fisher_f_distribution<>(RealType m = RealType(1), RealType n = RealType(1)) constructs a fisher_f distribution with specified degrees of freedom.
• fisher_f_distribution<>(param_type const &param) constructs a fisher_f distribution according to the values stored in the param struct.
• RealType m() const
  returns the degrees of freedom of the nominator;
• RealType n() const
  returns the degrees of freedom of the denominator;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.8: Gamma distribution

It has two parameters, alpha and beta. Its expected value is alpha * beta and its standard deviation is alpha * beta².

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType alpha = RealType(1),
                            RealType beta = RealType(1));

        RealType alpha() const;
        RealType beta() const;
    };

Constructors and members:

• gamma_distribution<>(RealType alpha = 1, RealType beta = 1) constructs a gamma distribution with specified alpha and beta parameters.
• gamma_distribution<>(param_type const &param) constructs a gamma distribution according to the values stored in the param struct.
• RealType alpha() const
  returns the distribution's alpha parameter;
• RealType beta() const
  returns the distribution's beta parameter;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.9: Geometric distribution

It has one parameter, prob, representing the probability of success in an individual bernoulli trial.

Defined types:

    using result_type = IntType;

    struct param_type
    {
        explicit param_type(double prob = 0.5);
        double p() const;
    };

Constructors, members and example:

• geometric_distribution<>(double prob = 0.5) constructs a geometric distribution for bernoulli trials each having probability prob of success.
• geometric_distribution<>(param_type const &param) constructs a geometric distribution according to the values stored in the param struct.
• double p() const
  returns the distribution's prob parameter;
• param_type param() const
  returns the object's param_type structure;
• void param(const param_type &param) redefines the parameters of the distribution;
• result_type min() const
  returns the distribution's lower bound (= 0);
• result_type max() const
  returns the distribution's upper bound;
• template<typename URNG> result_type operator()(URNG &urng)
  returns the next random value from the geometric distribution
• template<typename URNG> result_type operator()
(URNG &urng, param_type &param)
  returns the next random value from a geometric distribution initialized by the provided param struct.
• The random number generator that is passed to the generating functions must return integral values. Here is an example showing how the geometric distribution can be used:

  #include <iostream>
  #include <ctime>
  #include <random>
  
  int main()
  {
      std::linear_congruential_engine<unsigned, 7, 3, 61> engine(0);
  
      std::geometric_distribution<> dist;
  
      for (size_t idx = 0; idx < 10; ++idx)
          std::cout << "a random value: " << dist(engine) << "\n";
  
      std::cout << '\n' <<
          dist.min() << " " << dist.max() << '\n';
  
  }
  

18.9.2.10: Log-normal distribution

It has two parameters, m and s representing, respectively, the mean and standard deviation of ln(X).

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType m = RealType(0),
                            RealType s = RealType(1));

        RealType m() const;
        RealType s() const;
    };

Constructor and members:

• lognormal_distribution<>(RealType m = 0, RealType s = 1) constructs a log-normal distribution for a random variable whose mean and standard deviation is, respectively, m and s.
• lognormal_distribution<>(param_type const &param) constructs a log-normal distribution according to the values stored in the param struct.
• RealType m() const
  returns the distribution's m parameter;
• RealType stddev() const
  returns the distribution's s parameter;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.11: Normal distribution

It has two parameters, mean and standard deviation.

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType mean = RealType(0),
                            RealType stddev = RealType(1));

        RealType mean() const;
        RealType stddev() const;
    };

Constructors and members:

• normal_distribution<>(RealType mean = 0, RealType stddev = 1) constructs a normal distribution with specified mean and stddev parameters. The default parameter values define the standard normal distribution;
• normal_distribution<>(param_type const &param) constructs a normal distribution according to the values stored in the param struct.
• RealType mean() const
  returns the distribution's mean parameter;
• RealType stddev() const
  returns the distribution's stddev parameter;
• result_type min() const
  returns the lowest positive value of result_type;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.12: Negative binomial distribution

It has two parameters: (IntType) k (> 0), being the number of failures until the experiment is stopped and (double) p the probability of success in each individual experiment.

Defined types:

    using result_type = IntType;

    struct param_type
    {
        explicit param_type(IntType k = IntType(1), double p = 0.5);

        IntType k() const;
        double p() const;
    };

Constructors and members:

• negative_binomial_distribution<>(IntType k = IntType(1), double p = 0.5) constructs a negative_binomial distribution with specified k and p parameters;
• negative_binomial_distribution<>(param_type const &param) constructs a negative_binomial distribution according to the values stored in the param struct.
• IntType k() const
  returns the distribution's k parameter;
• double p() const
  returns the distribution's p parameter;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.13: Poisson distribution

It has one parameter, mean, specifying the expected number of events in the interval under consideration. E.g., if on average 2 events are observed in a one-minute interval and the duration of the interval under study is 10 minutes then mean = 20.

Defined types:

    using result_type = IntType;

    struct param_type
    {
        explicit param_type(double mean = 1.0);

        double mean() const;
    };

Constructors and members:

• poisson_distribution<>(double mean = 1) constructs a poisson distribution with specified mean parameter.
• poisson_distribution<>(param_type const &param) constructs a poisson distribution according to the values stored in the param struct.
• double mean() const
  returns the distribution's mean parameter;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.14: Student t distribution

It is characterized by one parameter: the degrees of freedom, which is equal to the sample size - 1.

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType n = RealType(1));

        RealType n() const;    // The degrees of freedom
    };

Constructors and members:

• student_t_distribution<>(RealType n = RealType(1)) constructs a student_t distribution with indicated degrees of freedom.
• student_t_distribution<>(param_type const &param) constructs a student_t distribution according to the values stored in the param struct.
• RealType n() const
  returns the degrees of freedom;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.9.2.15: Uniform int distribution

It has two parameters, a and b, specifying, respectively, the lowest value that can be returned and the highest value that can be returned.

Defined types:

    using result_type = IntType;

    struct param_type
    {
        explicit param_type(IntType a = 0, IntType b = max(IntType));

        IntType a() const;
        IntType b() const;
    };

Constructors and members:

• uniform_int_distribution<>(IntType a = 0, IntType b = max(IntType)) constructs a uniform_int distribution for the specified range of values.
• uniform_int_distribution<>(param_type const &param) constructs a uniform_int distribution according to the values stored in the param struct.
• IntType a() const
  returns the distribution's a parameter;
• IntType b() const
  returns the distribution's b parameter;
• result_type min() const
  returns the distribution's a parameter;
• result_type max() const
  returns the distribution's b parameter;

18.9.2.16: Uniform real distribution

It has two parameters, a and b, specifying, respectively, the half-open range of values ([a, b)) that can be returned by the distribution.

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType a = 0, RealType b = max(RealType));

        RealType a() const;
        RealType b() const;
    };

Constructors and members:

• uniform_real_distribution<>(RealType a = 0, RealType b = max(RealType)) constructs a uniform_real distribution for the specified range of values.
• uniform_real_distribution<>(param_type const &param) constructs a uniform_real distribution according to the values stored in the param struct.
• RealType a() const
  returns the distribution's a parameter;
• RealType b() const
  returns the distribution's b parameter;
• result_type min() const
  returns the distribution's a parameter;
• result_type max() const
  returns the distribution's b parameter;

18.9.2.17: Weibull distribution

It has two or three parameters and the two-parameter variant is offered by the STL. The three parameter variant has a shape (or slope) parameter, a scale parameter and a location parameter. The two parameter variant implicitly uses the location parameter value 0. In the two parameter variant the shape parameter (a) and the scale parameter (b) are provided. See
http://www.weibull.com/hotwire/issue14/relbasics14.htm for an interesting coverage of the meaning of the Weibull distribution's parameters.

Defined types:

    using result_type = RealType;

    struct param_type
    {
        explicit param_type(RealType a = RealType{ 1 },
                            RealType b = RealType{ 1 });

        RealType a() const;     // the shape (slope) parameter
        RealType b() const;     // the scale parameter
    };

Constructors and members:

• weibull_distribution<>(RealType a = 1, RealType b = 1) constructs a weibull distribution with specified a and b parameters;
• weibull_distribution<>(param_type const &param) constructs a weibull distribution according to the values stored in the param struct.
• RealType a() const
  returns the distribution's shape (or slope) parameter;
• RealType stddev() const
  returns the distribution's scale parameter;
• result_type min() const
  returns 0;
• result_type max() const
  returns the maximum value of result_type;

18.10: tie

    pair<int, int> factory()
    {
        return { 1, 2 };
    }
    
    void fun()
    {
        auto [one, two] = factory();
    
        cout << one << ' ' << two << '\n';
    }

Being able to use structured bindings is very useful in cases like these.

But what if we want to assign the fields of a struct to variables that have already been defined or that were passed to a function via its parameters? In those situations structured bindings offer no help. E.g., in the following code snippet a function retrieve is defined having an int & parameter and an int local variable and we want to assign the values returned by factory to those variables:

    void retrieve(int &one)
    {
        int two;
        // ...  = factory() ??
    }

Structured bindings cannot be used here: the elements of structured bindings cannot be references. Although it is possible to define a std::pair<int &, int &> such an object cannot be initialized with the references of one and two which are directly referring to the fields returned by factory. These statements won't compile:

    pair<int &, int &> p{one, two} = factory();
    pair<int &, int &>{one, two} = factory();

While it is possible to first define a pair<int &, int &> object and then assign factory's return value to it, that approach clearly is less elegant than what's offered by structured bindings:

    pair<int &, int &> p{one, two};
    p = factory();

Fortunately, there is a better alternative. After including the <tuple> header file (see also section 22.6) std::tie is available allowing us to `tie' references to fields of structured data types. Using std::tie it is very easy to associate the variables one and two of the function retrieve with the fields of the pair returned by factory:

    void retrieve(int &one)
    {
        int two;
    
        tie(one, two) = factory();
    
        cout << one << ' ' << two << '\n';
    }

When Executing these statements:

    int one = 0;
    int two = 0;

    cout << one << ' ' << two << '\n';

    retrieve(one);

    cout << one << ' ' << two << '\n';

the following output is obtained:

    0 0
    1 2
    1 0

In addition to the above the std::tie function also supports ordering and (in)equality comparisons. The struct Data in the next example defines three fields: an int, a std::string and a double. Each of these fields support ordering and (in)equality comparisons. In those cases, all comparison operators can easily be implemented through the spaceship operator (cf. section 11.6.2) using std::tie:

    struct Data
    {
        int d_int;
        string d_string;
        double d_double;
    };
    
    
    bool operator==(Data const &lhs, Data const &rhs)
    {
        return tie(lhs.d_int, lhs.d_string, lhs.d_double) ==
               tie(rhs.d_int, rhs.d_string, rhs.d_double);
    }
    
    partial_ordering operator<=>(Data const &lhs, Data const &rhs)
    {
        return tie(lhs.d_int, lhs.d_string, lhs.d_double) <=>
               tie(rhs.d_int, rhs.d_string, rhs.d_double);
    }

Note that struct Data's spaceship operator returns partial_ordering values (cf. section 18.7.3). Although int and std::string's spaceship operators return strong_ordering values, double's spaceship operator doesn't. Instead it returns partial_ordering values. Consequently, struct Data's spaceship operator also returns partial_ordering values.

18.11: Optional return values

Consider a function returning subsequent lines from a stream. That function could be a member function reading from a stream which was opened by its object. A first implementation of such a member function could be

    std::string Class::nextLine()
    {
        string line;
        getline(d_stream, line);
        return line;
    }

Common ways to handle failures in these situations are

• the functions return pointers to strings, which are 0 when getline fails, and point to strings containing the lines if getline succeeds;
• the functions return bools and define string pointers or reference parameters: the functions' return values indicate whether or not getline succeeded;
• the functions return std::pair or std::tuple objects of which one field is a bool and the other field is a std::string.

The standard template library offers yet another way to handle situations like these: the template class

    template <typename DataType>
    class optional;

The interpretation of std::optional objects is simple: either it contains an object of its DataType or it doesn't. If it does contain a DataType object then that object is available as object instead of a pointer to an object (which might have been dynamically allocated) of the specified DataType. At the same type the optional object can be interpreted as a bool. If the optional object contains a DataType object the optional's bool value is true. If it doesn't contain a DataType value, then its bool value is false.

The class std::optional offers the following facilities:

• Constructors:
  The default constructor (e.g., std::optional<std::string> opt;) does not contain a value;
  Copy- and move constructors are available;
  Objects may be constructed from values that are convertible to the optional's DataType (e.g., an optional<string> can be initialized from a NTBS). If the initializing value is an rvalue reference then the DataType object is move-constructed from the initializing value;

• operator=
  The assignment operator can be used to reassign the DataType values of optional objects or to reassign the optional objects from another optional object using the same DataType. Copy- and move-assignment operators are available;

• Accessors:
  explicit operator bool() and the has_value() members return true if the optional object contains a DataType value;
  value(), operator*(), and operator->() return references to the optional's DataType value. The reference is a const reference if called from an optional<DataType> const object; it is an rvalue-reference if called from an rvalue reference to an optional<DataType> object. Notes:
  • the operator* and operator-> members act like value, but do not imply that the optional's DataType member itself is stored as a pointer.
  • value() checks whether the optional object actually contains a DataType object, and throws a std::bad_optional_access exception if not.

  value_or(Type &&defaultValue) returns a copy of the optional object's DataType if the object contains a value or returns DataType{ defaultValue } if not. Note that DataType must be constructible from Type;

• Modifiers:
  swap(optional<DataType> &other): swaps the current and other optional objects' content.
  reset(): erases the optional's DataType member. Following reset() has_value() returns false.
  emplace(Args &&...args), emplace(initialize_list, Args &&...args): the first emplace forwards args to DataType's constructor; the second emplace forwards args to the initializer_list, and forwards that list to DataType's constructor;

• Comparison operators:
  All comparison operators (including operator<=>) are available (if defined for the optional's DataType) to compare the DataType objects of two optional objects;

• std::optional<DataType> make_optional(...)
  returns an optional object constructed from a DataType lvalue or rvalue reference, or constructed from the same arguments that are accepted by emplace.

Here is the implementation of a function nextLine, using std::optional<std::string> and a simple main function illustrating its workings:

    #include <iostream>
    #include <sstream>
    #include <string>
    #include <optional>

    using namespace std;

    optional<string> nextLine(istream &in)
    {
        std::optional<std::string> opt;

        string line;
        if (getline(in, line))
            opt = move(line);

        cout << "internal: has value: " << opt.has_value() <<
                                    ", value = " << *opt << '\n';
        return opt;
    }

    int main()
    {
        istringstream in{ "hello world\n" };

        auto opt = nextLine(in);

        cout << "main:     has value: " << opt.has_value() <<
                                    ", value = " << *opt << '\n';
        opt = nextLine(in);

        cout << "main:     has value: " << opt.has_value() <<
                                    ", value = " << *opt << '\n';
    }

The ouput of this program is:

    internal: has value: 1, value = hello world
    main:     has value: 1, value = hello world
    internal: has value: 0, value = 
    main:     has value: 0, value = hello world

• Table of Contents
• Previous Chapter
• Next Chapter