import io.Source+ −
import scala.util.matching.Regex+ −
+ −
// gets the first ~10K of a page+ −
def get_page(url: String) : String = { + −
try {+ −
Source.fromURL(url).take(10000).mkString + −
}+ −
catch {+ −
case e => {+ −
println(" Problem with: " + url)+ −
""+ −
}+ −
}+ −
}+ −
+ −
// staring URL for the crawler+ −
val startURL = """http://www.inf.kcl.ac.uk/staff/urbanc/"""+ −
+ −
// regex for URLs+ −
val http_pattern = """\"https?://[^\"]*\"""".r+ −
val my_urls = """urbanc""".r+ −
+ −
def unquote(s: String) = s.drop(1).dropRight(1)+ −
+ −
def get_all_URLs(page: String) : Set[String] = {+ −
(http_pattern.findAllIn(page)).map { unquote(_) }.toSet+ −
}+ −
+ −
// naive version - seraches until a given depth+ −
// visits pages potentially more than once+ −
def crawl(url: String, n: Int) : Unit = {+ −
if (n == 0) ()+ −
else if (my_urls.findFirstIn(url) == None) ()+ −
else {+ −
println("Visiting: " + n + " " + url)+ −
for (u <- get_all_URLs(get_page(url))) crawl(u, n - 1)+ −
}+ −
}+ −
+ −
// can now deal with depth 3+ −
// start on command line+ −
crawl(startURL, 4)+ −
+ −