Using sed to remove leading and trailing white spaces

Question

I have a file that looks like this:

 2 1 word 78 3 other words 2 some other words here 54 bla bla

I'd like to remove whitespaces and put a comma between values and rest. the output should look like

2,1 word 78,3 other words 2,some other words here 54,bla bla

The command

sed -e 's/\s*([0-9])+\s.+/&/'

does not change anything

tink · Accepted Answer · 2016-03-02 23:50:40Z

5

This should do the trick?

sed -Ee 's/^[[:space:]]+([0-9]+)[[:space:]]+/\1,/' bob 2,1 word 78,3 other words 2,some other words here 54,bla bla

answered Mar 2, 2016 at 22:23

15.6k5 gold badges54 silver badges58 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

illegal option r (under OSX)

Well, clean up you tags, then :) ... you tagged it as Linux Bash Or install a decent sed using homebrew ;}

++; simply use -E instead of -r, and it'll work on OSX (incidentally, GNU sed supports -E too, even though the man page doesn't say so).

@mklement0 : thanks for the -E reference; I'll amend my answer.

Ed Morton · Accepted Answer · 2016-03-02 22:39:46Z

This will work with any POSIX sed:

$ sed 's/^[[:space:]]*//; s/[[:space:]]*$//; s/[[:space:]]/,/' file 2,1 word 78,3 other words 2,some other words here 54,bla bla

Arif Burhan · Accepted Answer · 2016-03-02 22:24:27Z

-2

Try:

sed -e 's/(\d) (\d)(*)/\1,\2\3' infile.txt

answered Mar 2, 2016 at 22:24

5054 silver badges12 bronze badges

sed: 1: "s/(\d) (\d)(*)/\1,\2\3 ": \1 not defined in the RE

@Arif: Did you actually try that ? :)