postrank-uri 1.0.0 → 1.0.1

data/.gitignore

@@ -0,0 +1,2 @@
+pkg
+Gemfile.lock

data/lib/postrank-uri/version.rb

@@ -1,5 +1,5 @@
 module PostRank
   module URI
-    VERSION = "1.0.0"
+    VERSION = "1.0.1"
   end
 end

data/lib/postrank-uri.rb

@@ -2,6 +2,7 @@
 
 require 'addressable/uri'
 require 'domainatrix'
+require 'nokogiri'
 require 'yaml'
 
 module PostRank
@@ -64,7 +65,9 @@ module PostRank
     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|
@@ -79,23 +82,41 @@ module PostRank
       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?
@@ -103,7 +124,7 @@ module PostRank
       u
     end
 
-    def self.c18n(uri)
+    def c18n(uri)
       u = parse(uri)
 
       if q = u.query_values(:notation => :flat_array)
@@ -115,7 +136,7 @@ module PostRank
       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}"

data/postrank-uri.gemspec

@@ -8,14 +8,15 @@ Gem::Specification.new do |s|
   s.platform    = Gem::Platform::RUBY
   s.authors     = ["Ilya Grigorik"]
   s.email       = ["ilya@igvita.com"]
-  s.homepage    = "http://rubygems.org/gems/postrank-uri"
+  s.homepage    = "http://github.com/postrank-labs/postrank-uri"
   s.summary     = "URI normalization, c18n, escaping, and extraction"
   s.description = s.summary
 
   s.rubyforge_project = "postrank-uri"
 
-  s.add_dependency "addressable"
+  s.add_dependency "addressable", ">= 2.2.3"
   s.add_dependency "domainatrix"
+  s.add_dependency "nokogiri"
   s.add_development_dependency "rspec"
 
   s.files         = `git ls-files`.split("\n")

data/spec/postrank-uri_spec.rb

@@ -181,6 +181,41 @@ describe PostRank::URI do
         e("http://www.youtube.com/watch?v=w_j4Lda25jA　　とんかつ定食").should == ["http://www.youtube.com/watch?v=w_j4Lda25jA"]
       end
     end
+
+  end
+
+  context "href extract" do
+    it "should extract links from html text" do
+      l = PostRank::URI.extract_href("<a href='google.com'>link to google</a> with text <a href='b.com'>stuff</a>")
+      l.keys.size.should == 2
+
+      l.keys.should include('http://google.com/')
+      l.keys.should include('http://b.com/')
+
+      l['http://google.com/'].should == 'link to google'
+      l['http://b.com/'].should == 'stuff'
+    end
+
+    it "should handle empty hrefs" do
+      lambda do
+        l = PostRank::URI.extract_href("<a>link to google</a> with text <a href=''>stuff</a>")
+        l.should be_empty
+      end.should_not raise_error
+    end
+
+    context "relative paths" do
+      it "should reject relative paths" do
+        l = PostRank::URI.extract_href("<a href='/stuff'>link to stuff</a>")
+        l.should be_empty
+      end
+
+      it "should resolve relative paths if host is provided" do
+        l = PostRank::URI.extract_href("<a href='/stuff'>link to stuff</a>", "igvita.com")
+
+        l.size.should == 1
+        l['http://igvita.com/stuff'].should == 'link to stuff'
+      end
+    end
   end
 
-end
+end

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 1
   - 0
-  - 0
-  version: 1.0.0
+  - 1
+  version: 1.0.1
 platform: ruby
 authors: 
 - Ilya Grigorik
@@ -14,27 +14,27 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-01-20 00:00:00 -05:00
+date: 2011-01-21 00:00:00 -05:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: addressable
   prerelease: false
   requirement: &id001 !ruby/object:Gem::Requirement 
-    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
         segments: 
-        - 0
-        version: "0"
+        - 2
+        - 2
+        - 3
+        version: 2.2.3
   type: :runtime
   version_requirements: *id001
 - !ruby/object:Gem::Dependency 
   name: domainatrix
   prerelease: false
   requirement: &id002 !ruby/object:Gem::Requirement 
-    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
@@ -44,18 +44,29 @@ dependencies:
   type: :runtime
   version_requirements: *id002
 - !ruby/object:Gem::Dependency 
-  name: rspec
+  name: nokogiri
   prerelease: false
   requirement: &id003 !ruby/object:Gem::Requirement 
-    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
         segments: 
         - 0
         version: "0"
-  type: :development
+  type: :runtime
   version_requirements: *id003
+- !ruby/object:Gem::Dependency 
+  name: rspec
+  prerelease: false
+  requirement: &id004 !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        segments: 
+        - 0
+        version: "0"
+  type: :development
+  version_requirements: *id004
 description: URI normalization, c18n, escaping, and extraction
 email: 
 - ilya@igvita.com
@@ -66,6 +77,7 @@ extensions: []
 extra_rdoc_files: []
 
 files: 
+- .gitignore
 - Gemfile
 - README.md
 - Rakefile
@@ -77,7 +89,7 @@ files:
 - spec/helper.rb
 - spec/postrank-uri_spec.rb
 has_rdoc: true
-homepage: http://rubygems.org/gems/postrank-uri
+homepage: http://github.com/postrank-labs/postrank-uri
 licenses: []
 
 post_install_message: 
@@ -86,7 +98,6 @@ rdoc_options: []
 require_paths: 
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement 
-  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
@@ -94,7 +105,6 @@ required_ruby_version: !ruby/object:Gem::Requirement
       - 0
       version: "0"
 required_rubygems_version: !ruby/object:Gem::Requirement 
-  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
@@ -104,7 +114,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: postrank-uri
-rubygems_version: 1.3.7
+rubygems_version: 1.3.6
 signing_key: 
 specification_version: 3
 summary: URI normalization, c18n, escaping, and extraction