Package collectlinks is useful for only one task:

Given a response from http.Get it will use parse the page and return to you a slice of all the href links found.

Usage:

package main
import (
  "github.com/jackdanger/collectlinks"
  "net/http"
  "fmt"
)

func main() {
  resp, _ := http.Get("http://motherfuckingwebsite.com")
  links := collectlinks.All(resp.Body)
  fmt.Println(links)
}

Running that will output:

[http://twitter.com/thebarrytone http://txti.es]