I used the Unix shell to get some insight into my most-used shell commands.
If you are curious to see which shell commands you are using the most, I have a one-liner for you to find out.
I tested it on Linux only, but it should work similar on other Unix systems since I only use standard commands.
Works fine in Bash and Zsh. If you use Fish, you need to handle the different bahavior of the history command.
First, let’s see how many entries are in the history:
$ history | wc -l 2003
Now, let’s list the top 20 commands:
$ history | cut -c8- | cut -d' ' -f1 | sort | uniq -c | sort -nr | head -n20 269 e 242 c 168 l 122 npm 88 p 71 gst 65 git 56 man 55 history 49 .. 35 o 35 e. 34 fg 33 ls 32 cat 30 sudo 30 node 25 find 24 t 22 free
- cut -c8- removes the line numbers of history output.
- cut -d’ ‘ -f1 limits selection to first word in line because we want to ignore command arguments.
- sort | uniq -c counts all unique commands.
- sort -nr orders lines in reverse numerical order.
- head -n20 returns only the first 20 lines.
Interestingly 9 of the first 20 commands are custom shortcuts I created.
To see what ratio the top 20 commands make to the rest of the words, we can combine the above commands with some calculations:
$ TOP20=$(history | cut -c8- | cut -d' ' -f1 | sort | uniq -c | sort -nr | head -n20) $ TOTAL=$(history | wc -l) $ echo "scale=2; ($(echo "$TOP20" | cut -c1-8 | paste -s -d+)) / $TOTAL" | bc .74
- Let’s save the results of the previous commands in variables TOP20 and TOTAL.
- cut -c1-8 gives us only the counts of the previous output.
- paste -s -d+ joins all the lines with a “+” in between.
- scale=2; defines how many decimal places the results has.
- bc does the calculation.
Looks like this top 20 commands make up 74% of all commands I use.
If you think this was helpful, I would be happy to know your most-used commands and most valuable shortcuts.
I am also happy to hear about improvements and variations you came up with!