author | Christian Urban <christian dot urban at kcl dot ac dot uk> |
Wed, 27 Nov 2013 21:36:30 +0000 | |
changeset 208 | bd5a8a6b3871 |
parent 97 | 60a3ba90dd53 |
child 254 | dcd4688690ce |
permissions | -rw-r--r-- |
7 | 1 |
val my_urls = """urbanc""".r |
2 |
||
3 |
def crawl(url: String, n: Int) : Unit = { |
|
4 |
if (n == 0) () |
|
5 |
else if (my_urls.findFirstIn(url) == None) () |
|
6 |
else { |
|
96
9fcd3de53c06
updated
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
93
diff
changeset
|
7 |
println(s"Visiting: $n $url") |
97
60a3ba90dd53
added
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
96
diff
changeset
|
8 |
for (u <- get_all_URLs(get_page(url))) |
60a3ba90dd53
added
Christian Urban <christian dot urban at kcl dot ac dot uk>
parents:
96
diff
changeset
|
9 |
crawl(u, n - 1) |
7 | 10 |
} |
11 |
} |