sed removing extra character from end

Shell Programming and Scripting
1

Hi,

Searching through forum I found "sed 's/[ \t]*$//'" can be used to remove trailing whitespaces and tabs from file. The command works fine but I see minor issue as below. Can you please suggest if I am doing something wrong here.

$ cat a.txt
upg_prod_test
upg_prod_new

$ cat a.txt |sed 's/[ \t]$//'
upg_prod_tes
upg_prod_new

In above example the sed is removing trailing character if word is terminating with "t".

Thanks for your time.

2

Looks like your sed doesn't like to interpret the escape sequence "\t"; other versions do.
If your shell allows for it, try

sed "s/[ "$'\t'"]$//" a.txt

or simply

sed 's/[  ]$//' a.txt
         ^--- use CTRL-V <TAB> for this char

.

3

Still the same issue.

$ sed "s/[ "$'\t'"]$//" a.txt
upg_prod_tes
upg_prod_new

$ sed "s/[ "$'\t'"]*$//" a.txt
upg_prod_tes
upg_prod_new

Not sure if I would be able to make use the second option suggested as the "sed" is part of a SH which is copied from another server using scp command line.

Thanks.

4

Another approach:

sed 's/[ \t]*$//' a.txt

Or us CTRL-V <TAB> for \t

5

Strange. Works in bash on linux and (Free)BSD.

cat file
upg_prod_test
upg_prod_new    

sed "s/[ "$'\t'"]$//" file
upg_prod_test
upg_prod_new

sed 's/[        ]$//' file
upg_prod_test
upg_prod_new

What's your shell?
Pls try (with CTRL-V <TAB>)

sed 'p;s/[      ]$//' file

and post result

1 Like
6

I am using KSH on AIX 6.1

7 
spc=`printf '[ \t]'`
<a.txt sed 's/'"$spc"'$//'
1 Like
8

Try

TAB=$(printf "\011")
sed "p;s/[ $TAB]$//" file
9 
sed 's/'"$spc"'*$//' a.txt

Worked for me .. Thanks !!

---------- Post updated at 08:41 AM ---------- Previous update was at 08:40 AM ----------

Thanks RudiC/MadeInGermany for your time :slight_smile:

10

A general solution that should work with any (POSIX) sed (and independent of the capabilities of the shell it runs in) would be:

sed 's/[[:blank:]]*$//' a.txt
11

Will it be able to remove tabs as well?

12

Yes.

1 Like