Re: Allow tail to display data after stream stops sending input after N seconds

[Kaz Kylheku]

On 2022-04-24 03:45, konsolebox wrote:
> A command like this doesn't display data:
> 
> tail -f /var/log/messages -n +1  | grep -e something --line-buffered | tail
> 
> probably because the last tail waits for the pipe to terminate.
> 
> But this could work if there exists an option like -S which allows
> tail to wait for N seconds before considering input as "stopped" and
> then displaying currently absorbed input.

Hi konsolebox,

I was inspired by this thread to hack up a little utility called pw (Pipe Watch)
over a couple hours late last night.

Here is a link to the man page:

https://www.kylheku.com/cgit/pw/tree/pw.1

You pipe your input to pw, similarly in concept to a pager. It doesn't behave
like a pager though; it maintains a small FIFO of the last bunch of lines
of the data source, which are updated to an interactive display. The FIFO
is continuously pumping data; the display just takes snapshots.

I managed to cram a few features into it, like scrolling left-right over long
lines, a suspend mode which freezes the display, and some Vi-like
colon commands to save/append what is displayed to a file, or send it to a pipe.

If you have time, give it a try.

Enjoy!

P. S.

I have an idea for a new feature also. Since this is basically a kind of data
capture device for live input (like an oscilloscope), it would be useful if
it had a trigger mode. Instead of time-based refresh, this would look for
some pattern in the data to update the display.  

In logs, there are sometimes repeating patterns (like a periodic signal).
If you can lock on to a repeating pattern, you can get it to appear to freeze,
so you get a better visual, making it easy to see the parts in the pattern
that are changing.