find / -type f -print|xargs ls -l|sort -rnk5|head -10
[download]

{grin}

dir /b /o:-s | u:head -10
[download]

---

Examine what is said, not who speaks.

"Efficiency is intelligent laziness." -David Dunham
"Think for yourself!" - Abigail
Timing (and a little luck) are everything!

Good job, but in that environment, I would just ask Clippy what my largest files were! :)

Yes, you could. I realized that, but didn't present it, because it's neither efficient memory wise (as you are processing all the file names at the same time), nor is it run-time efficient, as you are sorting all the file names, and that's Omega (N log N). The solution where you keep a sorted buffer takes linear time.

If you want to factor in the amount of files (let's say k) to be reported, the solution you present takes O (N log N + k), while my solution takes O (k N). If you replace the array with a heap, you can reduce that to O (k + N log k). The used memory in your solution is O (N), and O (k) in my solution.

All mentioned upperbounds are tight.

Abigail