Sort a list and separate items with commas instead of newlines

by

I want to list all the filesystems in a single comma-delimited list, in dictionary order.
I’m using this command:

cat /proc/filesystems | sed 's/nodev//' | sed 's/,$//'

My output looks like this:

sysfs
rootfs
ramfs
bdev
proc
cgroup
cpuset
tmpfs
devtmpfs
debugfs
securityfs
sockfs
pipefs
anon_inodefs
devpts
ext3
ext2
ext4
hugetlbfs
vfat
ecryptfs
fuseblk
fuse
fusectl
pstore
mqueue
binfmt_misc
vboxsf

How can I change this to a single line output with commas separating each filesystem?

I figured part of it out by using xargs:

cat /proc/filesystems | sed 's/nodev//' | xargs | sed -e 's/ /,/g'

Now I want to make the output formatted in dictionary order.

Answer

Using awk

awk '{printf "%s%s",(NR>1?",":""),$NF;} END{print""}'  /proc/filesystems

Using sed

sed -r 's/^nodev//; s/^[[:blank:]]*//; H;1h;$!d;x; s/\n/,/g'  /proc/filesystems

How it works

  • s/^nodev//

    This eliminates nodev from the beginning of any line

  • s/^[[:blank:]]*//

    This eliminates the leading whitespace from each line.

  • H

    This appends the current line to the hold space.

  • 1h

    If we are on the first line, overwrite the hold space with this line.

  • $!d

    If we are not on the last line, delete the pattern space and start over on the next line.

  • x; s/\n/,/g

    If we get here, we are one the last line. In that case, move the hold space into the pattern space and replace all newlines with commas.

To understand this, it helps to recognize H;1h;$!d;x as an awk idiom which means read in the whole file.

Extra: sorting

To sort the names:

awk '{print $NF;}'  /proc/filesystems | sort | awk '{printf "%s%s",(NR>1?",":""),$NF;} END{print""}'

Attribution
Source : Link , Question Author : TommyLan , Answer Author : John1024

Leave a Comment