author | Christian Urban <christian dot urban at kcl dot ac dot uk> |
Wed, 07 Aug 2013 17:31:42 +0100 | |
changeset 94 | 9ea667baf097 |
parent 93 | 4794759139ea |
child 96 | 9fcd3de53c06 |
permissions | -rw-r--r-- |
val http_pattern = """\"https?://[^\"]*\"""".r val my_urls = """urbanc""".r val email_pattern = """([a-z0-9_\.-]+)@([\da-z\.-]+)\.([a-z\.]{2,6})""".r def crawl(url: String, n: Int) : Unit = { if (n == 0) () else { println("Visiting: " + n + " " + url) val page = get_page(url) println(email_pattern.findAllIn(page).mkString("\n")) for (u <- get_all_URLs(page)) crawl(u, n - 1) } }