Some time ago, I saw an inspiring talk from CppCon 2013: "C++ Seasoning" by Sean Parent. One of the main points of this presentation was not to use raw loops. Instead, prefer to use existing algorithms or write functions that 'wraps' such loops. I was curious about this idea and searched for nice code examples. Here is my short list of usage of algorithms from the C++ std library that might help in writing better code.
Of course. I could not skip two prominent examples from the original "C++ Seasoning" talk: slide and gather.
The Plan
Insertion Sort
In just two lines of code!
for (auto i = start; i != end; ++i)
std::rotate(std::upper_bound(start, i, *i), i, std::next(i));
How It Works?
Rotate(first, middle, last)- takes a range [first, last) and rotates it so that the middle element becomes the first in that range.
upper_bound - Returns an iterator pointing to the first element in the range [first,last) which compares greater than val. The range should be already sorted (or at least partitioned).
How do those two elements combine into Insertion sort?
std::upper_bound(start, i, *i) returns position of the first element greater than *i. Then, the range is shifted, so that i-th element becomes first.
Let's look at one example:
Pretty nice!
Quick Sort
Found on Stack Overflow:
template<class FwdIt, class Compare = std::less<>>
void quickSort(FwdIt first, FwdIt last, Compare cmp = Compare{})
{
auto const N = std::distance(first, last);
if (N <= 1) return;
auto const pivot = std::next(first, N / 2);
std::nth_element(first, pivot, last, cmp);
quickSort(first, pivot, cmp);
quickSort(pivot, last, cmp);
}
How It Works?
I will not describe quick sort algorithm... you should know how it works already! In this implementation, std::nth_element is used to do most of the job. This function partially sorts the range so that given n-th elements is placed in proper position. All of the elements before n-th element are less than or equal to the elements after the n-th element.
Slide
Example from the Sean Parent's talk:
template <typename It>
auto slide(It first, It last, It pos) -> std::pair<It, It>
{
if (pos < first) return { pos, std::rotate(pos, frist, last) };
if (last < pos) return { std::rotate(first, last, pos), pos };
return { first, llast };
}
How It Works?
As an example, you can imagine a list of items on a UI dialog. User selects a continuous range and then algorithm takes this range and moves it into some other place of the list.
- This function uses
std::rotate: to move elements forward or backward. - It returns two iterators - the start and the end of the new sequence. In C++11
std::rotate got new version and now can return iterator to the new position of p element. - If you are not interested in returning of this iterator pair, you can simplify this code much more.
Implementation note:
- In GCC 4.9 (and previous versions)
std::rotate does not return an iterator, but only void. So currently, this code will not work there.
Gather
Another example from Sean Parent's talk:
template <typename BiIt, typename UnPred>
auto gather(BiIt f, BiIt l, BiIt p, UnPred s) -> std::pair <BiIt, BiIt>
{
return { stable_partition(f, p, not1(s)),
stable_partition(p, l, s) };
}
How It Works?
Its use case can be similar to slide: select elements - using a predicate s (so this time continuous range is not needed), then gather those elements into a range and move this range to position around p. It returns the start and the end of the selected range.
UnPred is a predicate that returns if a given element is selected or not.
std::stable_partition: from cppreference
Reorders the elements in a given range in such a way that all elements for which the predicate returns true precede the elements for which predicate returns false. Relative order of the elements is preserved.
std::stable_partition is used twice:
Implementation note:
std::not1 does not work with the code correctly, so there is a proposal to use simple lambda. Read more here in Sean's comment.
String trim
Found on Stack Overflow
std::string trim(const std::string &s) {
return trimLeft(trimRight(s));
}
std::string trimLeft(const std::string &s) {
auto temp = s;
temp.erase(std::begin(temp),
std::find_if(std::begin(temp), std::end(temp),
[](char c){return !std::isspace(c, std::locale());
}));
return temp;
}
std::string trimRight(const std::string &s) {
auto temp = s;
temp.erase(std::find_if(std::rbegin(temp), std::rend(temp),
[](char c){return !std::isspace(c, std::locale()); }).base(),
std::end(temp));
return temp;
}
How It Works?
Another beautiful usage of Standard Library:
- In order to trim the
string, we trim from right and then from the left (what a discovery!) - trim left:
std::find_if returns iterator to the first non space character in the string. Then we erase those characters. - trim right: also uses
std::find_if but this time we use reverse iterators
Note: You can also use boost string algorithm to make life even easier.
Bonus :)
What does this code do?
while (std::next_permutation(start, end));
Simple, one line of code... should be nice! But...
Answer: It's another and 'wonderful' method of sorting containers - permutation sort! But please do not use it at home. :)
Complexity: O((n+1)!)
This algorithm is a variation of Bogosort and other similar 'sorting' algorithms. Read more on wiki. As victor_zverovich noticed, in Bogosort the next permutation is chosen at random, but std::next_permutation gives the next lexicographically greater permutation.
Sumup
I've showed several, I think nice, code examples where algorithms from C++ Standard Library are heavily used. Maybe next time, when I'll be writing some ugly piece of code I'll stop, think for a minute, and maybe some existing algorithm/function could be called instead.
Do you know some more interesting examples? My list, definitely, does not show all of them!
Resources
CodeProject
