sed:trying to replace everything before the pattern with an emp string

[fabreeze]

So I'm trying to replace everything before the pattern and everything afterward. and print it out using sed

for instance, i have a text file or html file and i want to only print the URL. so i use $grep pattern filename fist to look up all the line with URLs in it and i was thinking about pipe it to sed to remove the rest of the string before and after it.

<td><....><href="URLs goes here">blah</...>
<td><...><href="URLs goes here">blahblah</...>

Assume that there's no pattern before and after the URLs, as in it could be anything that comes before it....

I'm trying to print just this to standard output:
URLs
URLs
(i'm not allowed to post URLs yet but i hope you get the point)


I was thinking that I should find the way to replace whatever that comes before URLs and after. I have spent hours look up examples and options for sed and i'm so lost. please help!

[histrungalot]

Try:
$ sed -n 's/\(.*<href=\"\)\(.*\)\(\">.*\)/\2/p' yourInputFile.txt
URLs goes here
URLs goes here
$
Check to make sure that its not to greedy

[fabreeze]

Try:
$ sed -n 's/\(.*<href=\"\)\(.*\)\(\">.*\)/\2/p' yourInputFile.txt
URLs goes here
URLs goes here
$
Check to make sure that its not to greedy

I think it should work thanks!
but the thing is that what if whatever that comes before h t t p : / / is not always href.. that's my problem

is there any options in sed that let me print something by finding pattern from the beginning to then end. I can't use ^ or $ because there might not be a space in front of h t t p: / / and at the end of .com . net etc..

$sed some-other-option?/http:\/\/.*.[a-z][a-z][a-z]/..


h t t p:\/\/.*.\.[a-z][a-z][a-z] <-- just printing the pattern?
someone suggest getting rid of everything before and after the pattern but i don't know how to do that..... is that even efficient ?

[histrungalot]

Post example .txt

sed -n 's/\(.*http:\/\/\)\(.*\)\(\">.*\)/\2/p'

Not at my computer to test.