This is a fun video: https://www.youtube.com/watch?v=MKb4WD6mioE
Which includes three separate C++ solutions to a simple
(four letters in APL) problem, his preferred C++ solution
being:

  // C++17 Solution
  int maximumWealth(vector<vector<int>>& accounts) {
      return std::transform_reduce(
          accounts.cbegin(),
          accounts.cend(),
          0,
          [](auto a, auto b) { return std::max(a, b); },
          [](auto const& row) { return std::reduce(row.cbegin(), row.cend()); });
  }

Although what he thinks the C++ is trying to be is this Rust:

  pub fn maximum_wealth(accounts: Vec<Vec<i32>>) -> i32 {
      accounts.iter()
              .map(|x| x.iter().sum())
              .max()
              .unwrap()
  }


But D is very competitive here:

  int maximumWealth(const int[][] accounts) {
      import std.algorithm : map, sum, maxElement;
      return accounts.map!sum.maxElement;
  }

Or, aping the APL:

  int maximumWealth(const int[][] accounts) {
      import std.algorithm : fold, map;
      return accounts.map!(fold!"a+b").fold!"a.max(b)";
  }


That's it. I like this RosettaCode stuff and happened across
the video. He does include D in some other similar videos, like
https://www.youtube.com/watch?v=pDbDtGn1PXk&t=278s

On Tuesday, 9 March 2021 at 05:06:03 UTC, mipri wrote:
> But D is very competitive here:
>
>   int maximumWealth(const int[][] accounts) {
>       import std.algorithm : map, sum, maxElement;
>       return accounts.map!sum.maxElement;
>   }
>
> Or, aping the APL:
>
>   int maximumWealth(const int[][] accounts) {
>       import std.algorithm : fold, map;
>       return accounts.map!(fold!"a+b").fold!"a.max(b)";
>   }
>

Golf!

import std;

alias maximumWealth = pipe!(map!sum, maxElement);

unittest { assert([[2, 10], [11]].maximumWealth == 12); }