Shell Programming and Scripting

this didn't work. I got one of those "Broken Pipe" error messages

Check out reborgs' solution.

this seems to work a little bit. but there are still lots of unwanted characters attached to the strings that i need to get rid off. the thing is, i guess you guys would have to actually see the output i'm trying to work on for u to see how u can help me.

i could probably do this in shell but they dont want shell programming, they want this done in perl.

i would greatly appreciate it if anyone would point me to a perl website that deals specifically and only with grabbing and filtering through strings. i'm not interested in anything else.

It certainly wouldn't hurt. It'd also help to tell us how things are not working, a minor change may be all that's needed. Who is they? Perl people tend to yell RTFM after having scattered their documentation in 9,000 different places, with ridiculous keywords/names only perl experts would ever guess; the web is not as useful a resource for perl as it could be, unfortunately. Try 'man perlretut'. If you don't have this manpage on your system, I've uploaded it here.

Try this

Code:

grep -r akamai *  | perl -ane '$F[2] =~ s/^.*"(.*)"/$1/;print "$F[0] $F[2]"'

If this doesn't work show your result of grep -r akamai *

below is a better example of the outputs i have to filter through.


./NLT/sports/index.html: var myFlashObject = new FlashObject("http://a740.g.akamai.net/f/740/606/1d/image.pathfinder.com/TFK/media/shockflash/DSWMedia/feature_v1d.swf?bg_color=6600FF&head_text_1=Remembering+Steve+Irwin&head_size_1=26&head_lcnt_1=3&hea d_text_2=&head_size_2=0&head_lcnt_2=0&head_text_3=&head_size_3=0&head_lcnt_3=0&deck_text=The+adventu rous+�crocodile+hunter�+and+conservation+hero+dies+after+a+stingray+attack&deck_size=14&xref_link=/TFK/news/story/0,6260,1531665,00.html", "TFK News cover", "270", "145",5, "#6600FF");


and the above is only from one file. so imagine when i do a grep -r * on a directory full with files that contain strings that are similar. the list is very long and pretty much impossible (at least i think) to script.

so basically, what i need from that abysmal output is the file name which the output was found in and that would be == /NLT/sports/index.html and I also need the http string that contains "akamai.net". everything else needs to be filtered out.

try this

Code:

grep -r akamai * | perl -ne '$_=~ s/^([^:]*).*(http.*akamai[^"]*).*$/\1: \2/; print $_'