Truncate output after X lines and print message if and only if output was truncated

Question

Say I only need the first 5 lines of an output for logging purposes. I also need to know if and when the log has been truncated.

I am trying to use head to do the job, the seq command below outputs 20 lines that get truncated by head, and I echo a truncating information:

 > seq -f 'log line %.0f' 20 | head -n 5 && echo '...Output truncated. Only showing first 5 lines...'
log line 1
log line 2
log line 3
log line 4
log line 5
...Output truncated. Only showing first 5 lines...

But if the seq command outputs less than 5 lines, using the same above construction, I get a wrong "truncated" status:

seq -f ' log line %.0f' 3 | head -n 5 && echo '...Output truncated. Only showing first 5 lines...'
log line 1
log line 2
log line 3
...Output truncated. Only showing first 5 lines...

Is there a way for the head command (or another tool) to tell me if it truncated anything so that I only display the "...truncated..." message when needed?

Answer 1

A note of warning:

When you do:

cmd | head

and if the output is truncated, that could cause cmd to be killed by a SIGPIPE, if it writes more lines after head has exited. If it's not what you want, if you want cmd to keep running afterwards, even if its output is discarded, you'd need to read but discard the remaining lines instead of exiting after 10 lines have been output (for instance, with sed '1,10!d' or awk 'NR<=10' instead of head).

So, for the two different approaches:

output truncated, cmd may be killed

cmd | awk 'NR>5 {print "TRUNCATED"; exit}; {print}'
cmd | sed  '6{s/.*/TRUNCATED/;q;}'

Note that the mawk implementation of awk accumulates a buffer-full of input before starting processing it, so cmd may not be killed until it has written a buffer-full (8KiB on my system AFAICT) of data. That can be worked-around by using the -Winteractive option.

Some sed implementations also read one line in advance (to be able to know which is the last line when using the $ address), so with those, cmd may only be killed after it has output its 7^th line.

output truncated, the rest discarded so cmd is not killed

cmd | awk 'NR<=5; NR==6{print "TRUNCATED"}'
cmd | sed '1,6!d;6s/.*/TRUNCATED/'

Answer 2

You can use AWK:

seq -f 'line %.0f' 20 | awk 'NR <= 5; NR > 5 { print "...Output truncated. Only showing first 5 lines..."; exit }'

This prints the first five lines (if any) as-is; if it sees a line beyond that, it outputs the truncation message and exits.

You can specify the exit code to use, if you want to implement conditional processing:

seq -f 'line %.0f' 20 | awk 'NR <= 5; NR > 5 { exit 1 }' || echo ...Output truncated. Only showing first 5 lines...