I’ve been exploring the /usr directory as a part of a mission to better understand the Linux Filesystem Hierarchy Standard — the topography of the modern server. One question that arose during my travels through the filesystem was how many files are in /bin vs. /user/bin? That’s when I discovered this helpful pipeline.

ls | wc -l

Pipelines are composed of two or more commands connected with pipes on the command line. The pipe “|” redirects the output of the preceding command into the standard input of the next command, which acts on the data received. In this case, the list of filenames produced byls is passed to wc -l which counts the number of newlines in the list (i.e. every filename.) The wc command prints the count of newline, words, and bytes of a given file to standard output in that order. Here the -l option limits the count returned to just newlines. Below is the command in action.

\[user@host /\]$ ls /usr/bin | wc -l  
654

There are 654 files in /usr/bin of my Linux distro before any additional software has been installed. How does this compare to /bin? I could run this same pipeline for /bin and compare the result, but that still wouldn’t tell me if these directories have the same contents. Enter diff.

\[user@host /\]$ diff /usr/bin /bin

diff compares files line by line and outputs the difference. If the command is silent, then the two files are identical (though this behavior can be changed by using one of diff's many options). I ran diff to compare /usr/bin and /bin and got nothing. But what if they share subdirectories with the same name but different contents?

\[user@host /\]$ diff -r /usr/bin /bin

Silence. The -r option compared my directories recursively and turned up no differences. Now I feel confident the contents are the same, but my original mission to understand /usr is not complete. Onwards!

(Tips, corrections, or comments are always welcome.)