Sentinels and Concepts with Ranges Algorithms

The ranges library in C++20 supports sentinels. Sentinels stand for the end of a range and can be regarded as generalized end iterators.

A range provided by a begin iterator and an end sentinel specifies a group of items you can iterate over. The containers of the STL are ranges because their end iterator marks the end of the range.

Sentinel

The following example uses sentinels for a C-string and a std::vector<int>.

// sentinel.cpp

#include <iostream>
#include <algorithm>
#include <compare>
#include <vector>

struct Space {                        // (1)
bool operator== (auto pos) const {
        return *pos == ' '; 
    }
};

struct NegativeNumber {               // (2)
    bool operator== (auto num) const {
        return *num < 0;   
    }
};

struct Sum {                          // (7)
    void operator()(auto n) { sum += n; }
    int sum{0};
};

int main() {

    std::cout << '\n';

    const char* rainerGrimm = "Rainer Grimm";
   
    std::ranges::for_each(rainerGrimm, Space{}, [] (char c) { std::cout << c; });  // (3)
    std::cout << '\n';
    for (auto c: std::ranges::subrange{rainerGrimm, Space{}}) std::cout << c;      // (4)
    std::cout << '\n';

    std::ranges::subrange rainer{rainerGrimm, Space{}};                            // (5)
    std::ranges::for_each(rainer, [] (char c) { std::cout << c << ' '; });         // (6)
    std::cout << '\n';
    for (auto c: rainer) std::cout << c << ' ';
    std::cout << '\n';
  

    std::cout << "\n";


    std::vector<int> myVec{5, 10, 33, -5, 10};

    for (auto v: myVec) std::cout << v << " ";
    std::cout << '\n';

    auto [tmp1, sum] = std::ranges::for_each(myVec, Sum{});
    std::cout << "Sum: " << sum.sum << '\n';                                   // (8)

    auto [tmp2, sum2] = std::ranges::for_each(std::begin(myVec), NegativeNumber{}, 
                                              Sum{} );            
    std::cout << "Sum: " << sum2.sum << '\n';                                   // (9)

    std::ranges::transform(std::begin(myVec), NegativeNumber{},                 // (10)
                           std::begin(myVec), [](auto num) { return num * num; });
    std::ranges::for_each(std::begin(myVec), NegativeNumber{},                  // (11)
                          [](int num) { std::cout << num << " "; });
    std::cout << '\n';
    for (auto v: std::ranges::subrange{ std::begin(myVec), NegativeNumber{}}) { // (12)
        std::cout << v << " ";
    }

    std::cout << "\n\n";
    
}
The program defines two sentinels: Space (line 1) and NegativeNumber (line 2). Both define the equal operator. Thanks to the <compare> header, the compiler auto-generates the non-equal operator. The non-equal operator is required when using algorithms such as std::ranges_for_each or std::ranges::transform with a sentinel. Let me start with the sentinel Space.
Line (3) applies the sentinel Space{} directly onto the string “rainerGrimm“. Creating a std::ranges::subrange (line 4) allows it to use the sentinel in a range-based for-loop. You can also define a std::ranges::subrange and use it directly in the algorithm std::ranges::for_each (line 5) or in a range-based for-loop (line 6).
My second example uses a std::vector<int>, filled with the values {5, 10, 33, -5, 10}. The sentinel NegativeNumber checks if a number is negative. First, I sum up all values using the function object Sum (line 7). std::ranges::for_each returns a pair (it, func)it is the successor of the sentinel and func the function object applied to the range. Thanks to the structured binding, I can directly define the variables sum and sum2 and display their values (lines 8 and 9). std::ranges::for_each uses the sentinel NegativeNumber. Consequently, sum2 has the sum up to the sentinel. The call std::ranges::transform (line 10) transforms each element to its square: [](auto num){ return num * num}. The transformation stops with the sentinel NegativeNumber. Line 11 and line 12 display the transformed values.
Finally, here is the output of the program.
sentinel

You may ask yourself, should I use a classical algorithm of the STL or the ranges pendant on a container? Let me answer this question by comparing both.

std Algorithms versus std::ranges Algorithms

Before I dive into the details in my comparison, I want to provide the big picture:

AlgorithmsComparison

Range does not support numeric

The ranges does support the functions of the functional, and the algorithm header, but the function of the numeric header.  The numeric header includes mathematical functions such as  std::gcd, std::midpoint, std::iota, or std::accumulate.

Let me write about more interesting differences.

 

Rainer D 6 P2 500x500Modernes C++ Mentoring

  • "Fundamentals for C++ Professionals" (open)
  • "Design Patterns and Architectural Patterns with C++" (open)
  • "C++20: Get the Details" (open)
  • "Concurrency with Modern C++" (open)
  • "Generic Programming (Templates) with C++": October 2024
  • "Embedded Programming with Modern C++": October 2024
  • "Clean Code: Best Practices for Modern C++": March 2025
  • Do you want to stay informed: Subscribe.

     

    Concept support

    The std::ranges algorithms are the poster child for concepts.

    Let’s start with a comparison of the classical std::sort and the new std::ranges::sort. std::sort and std::ranges::sort require a random-access iterator that can access each range element in constant time. Here are the two relevant overloads for std::sort and std::ranges::sort.
    • std::sort

    template< class RandomIt >
    constexpr void sort( RandomIt first, RandomIt last );
    
    • std:ranges::sort

    template <std::random_access_iterator I, std::sentinel_for<I> S,
             class Comp = ranges::less, class Proj = std::identity>
    requires std::sortable<I, Comp, Proj>
    constexpr I sort(I first, S last, Comp comp = {}, Proj proj = {});
    
    What happens when you invoke std::sort or std::ranges::sort with a container such as std::list only supporting a bidirectional iterator?

    std::sort

    // sortVector.cpp
    
    #include <algorithm>
    #include <list>
     
    int main() {
       
       std::list<int> myList{1, -5, 10, 20, 0};
       std::sort(myList.begin(), myList.end());
        
    }
    
    Compiling the program sortVector.cpp with the GCC causes an epic error message of 1090 lines.
     sortVector

    std::ranges::sort

    // sortRangesVector.cpp
    
    #include <algorithm>
    #include <list>
     
    int main() {
       
       std::list<int> myList{1, -5, 10, 20, 0};
       std::ranges::sort(myList.begin(), myList.end());
        
    }
    

    Using std::ranges::sort instead of std::sort reduces the error message drastically. Now, I get 57 error lines.

    sortRangesVectorError

    Honestly, the error message of GCC should be easier to read, but I don’t blame them. We are still in the early process of supporting concepts. Here are the first 10 lines of the 57 lines. I marked the critical message in red.
    sortRangesVectorErrorLines

    Which mentoring program should I implement next?

    I’m happy that the current mentoring program, “Fundamentals for C++ Professionals”, is a big success and has more than 35 participants. Now, I will implement an additional mentoring program. All of them are based on my C++ books, posts, and classes.

    Make your choice here: https://www.modernescpp.com/index.php/my-next-mentoring-program

    What’s next?

    I’m not done with my comparison of  std and std::ranges algorithms. In my next post, I will write about the unified lookup rules that std::ranges algorithms provide and additional safety guarantees.

    Thanks a lot to my Patreon Supporters: Matt Braun, Roman Postanciuc, Tobias Zindl, G Prvulovic, Reinhold Dröge, Abernitzke, Frank Grimm, Sakib, Broeserl, António Pina, Sergey Agafyin, Андрей Бурмистров, Jake, GS, Lawton Shoemake, Jozo Leko, John Breland, Venkat Nandam, Jose Francisco, Douglas Tinkham, Kuchlong Kuchlong, Robert Blanch, Truels Wissneth, Mario Luoni, Friedrich Huber, lennonli, Pramod Tikare Muralidhara, Peter Ware, Daniel Hufschläger, Alessandro Pezzato, Bob Perry, Satish Vangipuram, Andi Ireland, Richard Ohnemus, Michael Dunsky, Leo Goodstadt, John Wiederhirn, Yacob Cohen-Arazi, Florian Tischler, Robin Furness, Michael Young, Holger Detering, Bernd Mühlhaus, Stephen Kelley, Kyle Dean, Tusar Palauri, Juan Dent, George Liao, Daniel Ceperley, Jon T Hess, Stephen Totten, Wolfgang Fütterer, Matthias Grün, Phillip Diekmann, Ben Atakora, Ann Shatoff, Rob North, Bhavith C Achar, Marco Parri Empoli, Philipp Lenk, Charles-Jianye Chen, Keith Jeffery, Matt Godbolt, and Honey Sukesan.

    Thanks, in particular, to Jon Hess, Lakshman, Christian Wittenhorst, Sherhy Pyton, Dendi Suhubdy, Sudhakar Belagurusamy, Richard Sargeant, Rusty Fleming, John Nebel, Mipko, Alicja Kaminska, Slavko Radman, and David Poole.

    My special thanks to Embarcadero
    My special thanks to PVS-Studio
    My special thanks to Tipi.build 
    My special thanks to Take Up Code
    My special thanks to SHAVEDYAKS

    Modernes C++ GmbH

    Modernes C++ Mentoring (English)

    Do you want to stay informed about my mentoring programs? Subscribe Here

    Rainer Grimm
    Yalovastraße 20
    72108 Rottenburg

    Mobil: +49 176 5506 5086
    Mail: schulung@ModernesCpp.de
    Mentoring: www.ModernesCpp.org

    Modernes C++ Mentoring,

     

     

    0 replies

    Leave a Reply

    Want to join the discussion?
    Feel free to contribute!

    Leave a Reply

    Your email address will not be published. Required fields are marked *