lib/postrank-uri.rb in postrank-uri-1.0.0 vs lib/postrank-uri.rb in postrank-uri-1.0.1

@@ -1,9 +1,10 @@
 # -*- encoding: utf-8 -*-
 
 require 'addressable/uri'
 require 'domainatrix'
+require 'nokogiri'
 require 'yaml'
 
 module PostRank
   module URI
 
@@ -62,11 +63,13 @@
 
     URIREGEX[:escape]   = /([^ a-zA-Z0-9_.-]+)/x
     URIREGEX[:unescape] = /((?:%[0-9a-fA-F]{2})+)/x
     URIREGEX.each_pair{|k,v| v.freeze }
 
-    def self.extract(text)
+    module_function
+
+    def extract(text)
       return [] if !text
       urls = []
       text.to_s.scan(URIREGEX[:valid_url]) do |all, before, url, protocol, domain, path, query|
         begin
           url = clean(url).to_s
@@ -77,35 +80,53 @@
       end
 
       urls.compact
     end
 
-    def self.escape(uri)
+    def extract_href(text, host = nil)
+      urls = {}
+      Nokogiri.HTML(text).search('a').each do |a|
+        begin
+          url = normalize(c18n(unescape(a.attr('href'))))
+          if url.host.empty?
+            next if host.nil?
+            url.host = host
+          end
+
+          urls[url.to_s] = a.text
+        rescue
+          next
+        end
+      end
+      urls
+    end
+
+    def escape(uri)
       uri.gsub(URIREGEX[:escape]) do
         '%' + $1.unpack('H2' * $1.size).join('%').upcase
       end.gsub(' ','%20')
     end
 
-    def self.unescape(uri)
+    def unescape(uri)
       uri.tr('+', ' ').gsub(URIREGEX[:unescape]) do
         [$1.delete('%')].pack('H*')
       end
     end
 
-    def self.clean(uri)
+    def clean(uri)
       normalize(c18n(unescape(uri))).to_s
     end
 
-    def self.normalize(uri)
+    def normalize(uri)
       u = parse(uri)
       u.path = u.path.squeeze('/')
       u.query = nil if u.query && u.query.empty?
       u.fragment = nil
       u
     end
 
-    def self.c18n(uri)
+    def c18n(uri)
       u = parse(uri)
 
       if q = u.query_values(:notation => :flat_array)
         q.delete_if { |k,v| C18N[:global].include?(k) }
         q.delete_if { |k,v| C18N[:hosts].find {|r,p| u.host =~ r && p.include?(k) } }
@@ -113,10 +134,10 @@
 
       u.query_values = q
       u
     end
 
-    def self.parse(uri)
+    def parse(uri)
       return uri if uri.is_a? Addressable::URI
 
       uri = uri.index(URIREGEX[:protocol]) == 0 ? uri : "http://#{uri}"
       Addressable::URI.parse(uri).normalize
     end
\ No newline at end of file