mjhjmtu

I have a file as below

"IN001~24Apr16~Hi,
what a way?
oh no!~
not here~"
"IN003~29Apr16~
what a way?
~oh no!
say again.
not again~"

and I want the output in below format

"IN001~24Apr16~Hi,what a way?oh no!~not here~"
"IN003~29Apr16~what a way?~oh no!say again.not again~"

Any solution using awk or sed is much appreciated.

awk ' printf "%s",$0 ; /"$/ print ""; ;' inputfile

There shall be a newline in the output only if the input line ends with double quotes. Thus this awk code outputs all input lines without a newline. After that it checks whether the input line ends in a " and if so then a newline is output.

$ sed -n -e '/^"/h;n;' -e H -e '/"$/g;s/n//g;p;' file
"IN001~24Apr16~Hi,what a way?oh no!~not here~"
"IN003~29Apr16~what a way?~oh no!say again.not again~"

The sed command will...

• /^"/h;n;: copy all lines that start with " to the hold space (h), overwriting whatever was there, and immediately move on to the next line of input (n).




• H: Other lines are appended to the hold space.




• /"$/g;s/n//g;p;: If a line ends with ", the hold space will be fetched (g) and newlines removed from it (these are added when appending lines to the hold space), and the resulting string is printed.

The command line can be compressed a bit:

$ sed -n '/^"/h;n;;H;/"$/g;s/n//g;p;' file