Today I had a need to process large numbers of PostgreSQL database tables and report how many rows were in each. Once I had that information, I needed to do some additional math on this data and then print it out in a nice formatted manner. And for reasons I won’t go into, I had to use shell.
But it was all very trivial…until I got to that last part. It involves commification.
Is that even a word? My browser’s spell checker thinks not and it doesn’t know commify either. While you might think commification is a 1950s word for the spread of Marxism, it means “adding commas to numbers”. In other words, turning 100000000 into 100,000,000.
Of course, in some countries they’d look at 100,000,000 and ask why you have two decimal places. And in India they’d expect this to be formatted as 1,00,00,00,00 which is as confused to me as 100,000,000 would be to an Indian. But regardless of your locale, the issue is the same.
There are Linux tools to transform, chop, parse, split, join, fold, bend, substitute, replace, and apply just about any contortion to a string of letters, numbers, and symbols. Indeed, text processing is one of Unix’s greatest strengths. Indeed, early on that was what Unix was mainly used for.
However, at first glance, I was unable to find any simple commifier for Linux. I have little snippets on how to do it in Python and Perl but there should be an easy way to do this on the command line. Right?
Wrong.
Some people have written extensive scripts to do this. Removing comments and blank lines from that script, it’s 20 lines. 20 lines of interpreted shell code just to format a number!
Fortunately, we don’t need to do this. We can do it in a single line:
$ echo 100000000 | sed ':a;s/\(^\|[^0-9.]\)\([0-9]\+\)\([0-9]\{3\}\)/\1\2,\3/g;ta'
$ 100,000,000
There, that’s simple enough, right? Let’s walk through that sed command letter by letter…wait, don’t go! I was just kidding. BTW, that only works for GNU sed – there are other options discussed in the sed FAQ.
But please don’t waste your time. Let me show you the simple way that’s small enough to easily commit to memory.
Check this out:
$ echo 100000000 | printf "%'d\n" 100000000
$ 100,000,000
Nice! This works in sufficiently modern versions of bash and also ksh (93+ – if you’re still using ksh88 in 2022 you need counseling).
It’s the apostrophe after the % sign that tells printf to commify.
Any suggestion that I wrote this article so I could find it the next time I google how to do this is probably accurate.
Related Posts:
I also enjoy using LEBs! I have an empire of little guys for VPNs, larger guys for apps, and big guys for databases and storage. When I'm not in front a screen I'm into German Shepherd dogs, quality knives, target shooting, theology, tabletop roleplaying games, and forest hiking.
I enjoy writing technical articles here on LowEndBox to help people get more out of their systems. You can find me on LowEndTalk @raindog308
Figured I’d point this out – we’d write 100000000 in Indian system as 10,00,00,000 and not 1,00,00,00,00, and read it as 10 crore.
This…
echo 100000000 | printf “%’d\n” 100000000
…doesn’t actually work because printf is getting the number from the end, not the one being piped in. If you want to pipe it in you have to do it via xargs, like so.
grep MemTotal /proc/meminfo | tr -s ‘ ‘ | cut -d ‘ ‘ -f 2 | xargs printf “%’d\n”
Any suggestion that I wrote this comment so I could find it the next time I google how to do this is probably accurate.