arachnid 0.1.3 → 0.2.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/lib/arachnid.rb +4 -1
- metadata +1 -1
data/lib/arachnid.rb
CHANGED
|
@@ -21,7 +21,10 @@ class Arachnid
|
|
|
21
21
|
|
|
22
22
|
def crawl(options = {})
|
|
23
23
|
|
|
24
|
+
#defaults to 1 thread so people don't do a stupid amount of crawling on unsuspecting domains
|
|
24
25
|
threads = options[:threads] ? options[:threads] : 1
|
|
26
|
+
#defaults to -1 so it will always keep running until it runs out of urls
|
|
27
|
+
max_urls = options[:max_urls] ? options[:max_urls] : -1
|
|
25
28
|
|
|
26
29
|
@hydra = Typhoeus::Hydra.new(:max_concurrency => threads)
|
|
27
30
|
@global_visited = BloomFilter::Native.new(:size => 1000000, :hashes => 5, :seed => 1, :bucket => 8, :raise => false)
|
|
@@ -29,7 +32,7 @@ class Arachnid
|
|
|
29
32
|
|
|
30
33
|
@global_queue << @start_url
|
|
31
34
|
|
|
32
|
-
while(@global_queue.size > 0)
|
|
35
|
+
while(@global_queue.size > 0 && @global_visited.size != max_urls)
|
|
33
36
|
temp_queue = @global_queue
|
|
34
37
|
|
|
35
38
|
temp_queue.each do |q|
|