Shell Programming and Scripting

Ok, so there is a perl script that runs as a server, on my local host. It tells me which port to use. I want to pipe that output into my browser so I can do the whole thing with a single command. The problem is, I think, that the program doesn't actually exit cause it's running a server, so... so I dunno. Anyway the output looks like this:

Code:

Startup complete. You may now point your browser at this address: http://127.0.0.1:8890

... though obviously the port can be different. so, when I run 'foo.pl|grep http,' I just get the line with address, which is exactly what I want. I've tried a few things to get my browser to open this. So, at first I was trying this:

Code:

link=`foo.pl|grep http` firefox "$link"

This doesn't work. It seems to start the server, but I don't get any output out of, so it's useless unless I check my ports manually. The browser doesn't start either. Then I tried this; cutting out the middle-man, as it were:

Code:

/usr/local/diogenes/perl/diogenes-server.pl|grep http|firefox

This starts the sever and the browser, but it doesn't actually point the browser at the server. not really sure what to try next. edit: bb is on the fritz right now or something. Formatting is really messed up

Try this...

Code:

firefox `foo.pl | awk '{print $NF}'`

--ahamed

Not doing it.

whats the output of foo.pl | awk '{print $NF}?

An argument to firefox is working for me.
Try firefox "www.google.com". Is it opening a new browser with the given address?

--ahamed

The output is:

Code:

address:
http://127.0.0.1:8888

The one with grep just outputs the address.

The problem is that the perl script doesn't exit (it maintians the server), so my bash script doesn't make it past that line. I tried it with an '&' but it breaks the syntax

And yes, firefox can be opened from the command line with a url (and I'm actually using uzbl-browser, but I didn't want it to throw anyone off).

so, you mean to say, the foo.pl will never exit but displays the info on the screen?
If that is the case, then why not write the ouput to a file and then read the file & open the browser. Something like this

Code:

foo.pl >/tmp/foo.log & firefox `grep http /tmp/foo.log`

--ahamed

I tried something similar earlier, but I had the grep before the '>,' which must have broken it.

This worked... well, it almost worked. It needs a moment for the output to be there from te browser, so I ended up with this:

Code:

foo.pl > /tmp/foo.log &
sleep .3
firefox `grep http /tmp/foo.log`

... and then a little something to kill the perl script when the browser closes.

Thanks!!