Filtering content in a shell script

I'm currently playing around with ways of changing the output from my SETI@Home client, as the standard outputs (nothing, or a long stream of all of the data) are a bit limited and annoying.

I've written a simple Perl script which takes the STDIN, runs it through a reg. exp looking for something like a percentage sign, then outputs lines with a percentage sign, so that it just outputs the completion percentages and not all the other data. What I'd like to do is use a shell script to do this, and also initialise the program, so it does something to this effect:

* start seti client
* take seti output, put through a reg exp, and output certain text on certain conditions

Can anyone give me a hand here?

if i were you, i would just put it all into the perl script:

to start seti and get the command output...try something like this:

open SETIOUTPUT,"/dir/where/seti/is/setiathome 2>&1|";


regex...


close SETIOUTPUT;

I could But I'n interested in how to do it in a shell script, rather than a Perl script. Thanks for the suggestion though

say the output is:

project seti@home process 10%

and you start the client with:
seticlient|myscript.sh

then myscript.sh should be:

looking at your actual wishes, you should make 2 scripts:
1. "start_seti.sh"

2. "filter_seti.sh":

untested, but should work
the asterisks are used as in good old dos, not regexps.

as you see, regexps are not supported by the shell, but grep and sed can do a great job!

Thanks, that's really cool I'll give it a go a little later. This is a good opportunity for me to finally try and get the hang of shell script syntax lol

Do you know if there's any way of doing it all in one shell script?

hmm. yes, there is. but i can´t remember the exact syntax and my linux machine still suffers from the harddrive breaking down

try this:

seti_client | while read line; do
... process messages here
done;

and the shell itself can do easy pattern matching / string manipulation (no regexps) like this:
$line="testing";

echo ${line%%in}
should output "test"
and:
echo ${line##in}
should output "g"

not sure about the #/%´s and their number (#, ##, %, %% are the four), i always mix them up since i don´t use it reguarly, but you can find out either rtfm´ing /usr/local/doc/packages/bash/bash_manual.html or just trying