[SOLVED] Using sed to change spaces to a tab

[pnelsonsr]

Have an output that has spaces between columns and these are not a standard amount so I would like to change them to tabs so I can cut the values out. Been trying to do it with sed with something like:

sed "s/\s+/\t/g"

But that doesn't seem to be working.

---------- Post added at 02:36 PM ---------- Previous post was at 02:30 PM ----------

Well I was trying some of the command line arguments for sed and found -r so the following now removes the white space between columns and puts a tab in its place. Here is the working sed statement:

sed -r "s/\s+/\t/g"

[stevea]

sed -e "s/\ /\t/g"

but tr is easier for this one

tr ' ' '\t'

---------- Post added at 06:31 PM ---------- Previous post was at 06:20 PM ----------

Oops - I misread as single space ....
sed -e 's/ [ ]*/\t/g'

[pnelsonsr]

stevea-> sed -e 's/ [ ]*/\t/g'

The [ ]* doesn't work as it takes 0-n which ultimately puts a tab between everything. As noted above I used the + for 1-n.

To specify white space I used /s but you can also use the set [ ] or even [[:space:]]... /s was just shorter but [ ] works as well

Alrighty I'm amazed that there are multiple ways to do it and I just didn't stumble on it at first! Thanks.

[cazo]

Even though you have solved this, unexpand (part of coreutils) will also take care of this.

[pnelsonsr]

Originally Posted by cazo

Even though you have solved this, unexpand (part of coreutils) will also take care of this.

I tried this with unexpand -a and it did only the first space and not all the space between the text in the columns. I also tried it with --first-only and this also did not work correctly. So, I'm not sure how you use unexpand.