Created
July 31, 2012 04:32
-
-
Save rhizome/3213628 to your computer and use it in GitHub Desktop.
Retrieve a list of all URLs within a site's pages
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env ruby | |
require 'anemone' | |
require 'pp' | |
require 'optparse' | |
options = OpenStruct.new | |
OptionParser.new do |o| | |
o.banner = "geturls.rb [options] URL" | |
o.on("-s", "--sortby [url|count]", [:url,:count], "Sort by URL or count") do |s| | |
options.sortby = s | |
end | |
o.on("-l", "--[no-]links", "Include links within pages") do |l| | |
options.links = l | |
end | |
o.on_tail("-h", "--help", "Show this message") do | |
puts o | |
exit | |
end | |
end.parse! | |
options.url = ARGV | |
lnks = Hash.new | |
externals = Hash.new | |
Anemone.crawl(options.url) do |anem| | |
anem.on_every_page do |page| | |
lnks["#{page.url}"] ||= 0 | |
lnks["#{page.url}"] += 1 | |
if options.links | |
if page.doc.present? | |
if page.doc.css('a') | |
page.doc.css('a').map { |link| | |
externals["#{link['href']}"] ||= 0 | |
externals["#{link['href']}"] += 1 | |
} | |
end | |
end | |
end | |
end | |
end | |
lnks.sort_by {|url,count| options.sortby.nil? ? count : options.sortby }.each { |k,v| | |
puts "#{k}: #{v}" | |
} | |
externals.sort_by {|url,count| options.sortby.nil? ? count : options.sortby }.each { |k,v| | |
puts "#{k}: #{v}" | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment