Removing new line after a particular text via bash/awk/sed/perl

Question

I would like to remove all the newline character that occurs after a partiular string and replace it with a tab space. Say for instance my sample.txt is as follows

Answer 1

This might work for you:

echo -e "foo\nbar bar bar bar some text"| sed '/foo$/{N;s/\n/\t/}'
foo     bar bar bar bar some text

Actually:

echo -e "foo\nbar bar bar bar some text"| sed '/foo$/N;s/\n/\t/'
foo     bar bar bar bar some text

The latter solution is less efficient but not by much!

Answer 2

Mmm. not sure this fits your requirements, but I'd do this in vim:

:g/^foo$/join

Or, if foo could be anywhere on a line:

:g/foo$/join

Of course, :%s/foo\n/foo\t/g will do nicely too

Answer 3

The sed version looks like

#!/bin/sed -f
/foo/{
N
s/\n/\t/
}

If this is all you want to do in your sed command, you can simplify the above to a one-liner:

sed -e '/foo/N;y/\n/\t/'  <sample.txt

Explanation: N appends a newline and the next line of input; y/\n/\t replaces all newlines with tabs, or s/\n/\t/ replaces the first newline with a tab.

Note that since sed is line-oriented, the newlines in the input are considered line-separators rather than as part of any line (this is in contrast e.g. to Perl).

Answer 4

Not all tools can do newline replacement. Perl can:

cat file.txt | perl -p -e 's/\n/ /g'

Answer 5

perl -pe 's/(?<=foo)\n/\t/' input

Answer 6

Here is how:

cat input.txt | sed ':a;N;$!ba;s/foo\n/foo\t/g'

More about why simple sed 's/foo\n/foo\t/g' does not work here: http://linuxtopia.org/online_books/linux_tool_guides/the_sed_faq/sedfaq5_009.html