awk and sed challenge: adjust the text width

Question

So let's see how can we do this: trim the text width within a certain value, say, 10. For lines longer than 10, break it into multiple lines.

Example: A text file:

01234567
01234567890123456789abcd
0123

should be changed to:

01234567
0123456789
0123456789
abcd
0123

So how can we do it using sed or awk as short as possible?

Answer 1

$ sed -e 's/\(..........\)/\1\\n/g' foo.txt

or, if that doesn't work (eg, don't have a sufficiently new gnu sed), just insert a newline and make sure it's quoted:

$ sed -e 's/\(..........\)/\1\\
/g' foo.txt

You can pretty much transliterate that into awk, too:

$ awk '{ gsub(/........../, "&\n" ) ; print}' foo.txt

Answer 2

In awk with a variable width:

awk -v WIDTH=5 '{ gsub(".{"WIDTH"}", "&\n"); printf $0 }; !/\n$/ { print "" }'

The final statement prevents the printing of extra newlines when the line is an exact multiple of the maximum line width.

Answer 3

Use the proper tool for the job...

fold -w 10