app3.scala
author Christian Urban <urbanc@in.tum.de>
Thu, 27 Sep 2012 12:02:45 +0100
changeset 8 5751a3ee41ce
parent 7 73cf4406b773
permissions -rw-r--r--
tuned
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
7
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     1
val my_urls = """urbanc""".r
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     2
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     3
def crawl(url: String, n: Int) : Unit = {
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     4
  if (n == 0) ()
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     5
  else if (my_urls.findFirstIn(url) == None) ()
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     6
  else {
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     7
    println("Visiting: " + n + " " + url)
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     8
    for (u <- get_all_URLs(get_page(url))) crawl(u, n - 1)
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     9
  }
73cf4406b773 updated
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
    10
}