microformats2 1.0.0

data/.autotest

@@ -0,0 +1,23 @@
+# -*- ruby -*-
+
+require 'autotest/restart'
+
+# Autotest.add_hook :initialize do |at|
+#   at.extra_files << "../some/external/dependency.rb"
+#
+#   at.libs << ":../some/external"
+#
+#   at.add_exception 'vendor'
+#
+#   at.add_mapping(/dependency.rb/) do |f, _|
+#     at.files_matching(/test_.*rb$/)
+#   end
+#
+#   %w(TestA TestB).each do |klass|
+#     at.extra_class_map[klass] = "test/test_misc.rb"
+#   end
+# end
+
+# Autotest.add_hook :run_command do |at|
+#   system "rake build"
+# end

data/History.txt

@@ -0,0 +1,6 @@
+=== 1.0.0 / 2011-06-14
+
+* 1 major enhancement
+
+  * Birthday!
+

data/Manifest.txt

@@ -0,0 +1,8 @@
+.autotest
+History.txt
+Manifest.txt
+README.txt
+Rakefile
+bin/microformats2
+lib/microformats2.rb
+test/test_microformats2.rb

data/README.txt

@@ -0,0 +1,41 @@
+# Microformats
+
+http://github.com/veganstraightedge/microformats2
+
+## DESCRIPTION
+
+Generic Microformats 2 Extractor
+
+## FEATURES/PROBLEMS
+
+* parses and extracts [Microformats 2](http://microformats.org/wiki/microformats-2) syntax
+* needs more test cases
+* needs better docs
+
+## SYNOPSIS
+
+  Microformats2.parse(File.open("http://iamshane.html"))
+
+## REQUIREMENTS
+
+* Hoe
+* Nokogiri
+
+## INSTALL
+
+    sudo gem install microformats2
+
+## DEVELOPERS
+
+After checking out the source, run:
+
+    rake newb
+
+This task will install any missing dependencies,
+run the tests/specs, and generate the RDoc.
+
+## LICENSE
+
+PUBLIC DOMAIN.
+Your heart is as free as the air you breathe.
+The ground you stand on is liberated territory.

data/Rakefile

@@ -0,0 +1,12 @@
+# -*- ruby -*-
+
+require 'rubygems'
+require 'hoe'
+
+Hoe.spec 'microformats2' do
+  developer('Shane Becker', 'veganstraightedge@gmail.com')
+  extra_deps << ['nokogiri', ">= 0"]
+end
+
+
+# vim: syntax=ruby

data/bin/microformats2

@@ -0,0 +1,3 @@
+#!/usr/bin/env ruby
+
+abort "you need to write me"

data/lib/microformats2.rb

@@ -0,0 +1,165 @@
+require 'nokogiri'
+require 'time'
+require 'date'
+
+module Microformats2
+  VERSION = "1.0.0"
+
+  class LoadError < StandardError; end
+
+  def self.parse(html)
+    raise LoadError, "argument must be a String or File" unless [String, File].include?(html.class)
+
+    doc = Nokogiri::HTML(html)
+    microformats = Hash.new{|hash, key| hash[key] = Array.new}
+    doc.css("*[class^=h-]").each do |microformat|
+      constant_name = classify(microformat.attribute("class").to_s.gsub("-","_"))
+
+      if Object.const_defined?(constant_name)
+        klass = Object.const_get(constant_name)
+      else
+        klass = Class.new
+        Object.const_set constant_name, klass
+      end
+
+      obj = klass.new
+
+      # Add any properties to the object
+      add_properties(microformat, obj)
+      add_urls(microformat, obj)
+      add_dates(microformat, obj)
+      add_times(microformat, obj)
+
+      microformats[constant_name.downcase.to_sym] << obj
+    end
+
+    return microformats
+  end
+
+  def self.add_method(obj, method_name)
+    unless obj.respond_to?(method_name)
+      obj.class.class_eval { attr_accessor method_name }
+    end
+
+    obj
+  end
+
+  def self.add_properties(mf, obj)
+    %w(p n e i).each do |letter|
+      mf.css("*[class|=#{letter}]").each do |property|
+        property.attribute("class").to_s.split.each do |css_class|
+          if css_class =~ /^[pnei]/
+            css_class   = css_class[2..-1].gsub("-","_")
+            method_name = css_class.gsub("-","_")
+            value       = property.text.gsub(/\n+/, " ").gsub(/\s+/, " ").strip
+
+            add_method(obj, method_name)
+
+            if cur = obj.send(method_name)
+              if cur.kind_of? Array
+                cur << value
+              else
+                obj.send("#{method_name}=", [cur, value])
+              end
+            else
+              obj.send("#{method_name}=", value)
+            end
+          end
+        end
+      end
+    end
+  end
+
+  def self.add_urls(mf, obj)
+    mf.css("*[class*=u-]").each do |property|
+      property.attribute("class").to_s.split.each do |css_class|
+        if css_class =~ /^u/
+          css_class   = css_class[2..-1].gsub("-","_")
+          method_name = css_class.gsub("-","_")
+          value       = property.attribute("href").to_s
+
+          add_method(obj, method_name)
+
+          if cur = obj.send(method_name)
+            if cur.kind_of? Array
+              cur << value
+            else
+              obj.send("#{method_name}=", [cur, value])
+            end
+          else
+            obj.send("#{method_name}=", value)
+          end
+        end
+      end
+    end
+  end
+
+  def self.add_dates(mf, obj)
+    mf.css("*[class*=d-]").each do |property|
+      property.attribute("class").to_s.split.each do |css_class|
+        if css_class =~ /^d/
+          css_class   = css_class[2..-1].gsub("-","_")
+          method_name = css_class.gsub("-","_")
+          value       = DateTime.parse((property.attribute("title") || property.text).to_s)
+
+          add_method(obj, method_name)
+
+          if cur = obj.send(method_name)
+            if cur.kind_of? Array
+              cur << value
+            else
+              obj.send("#{method_name}=", [cur, value])
+            end
+          else
+            obj.send("#{method_name}=", value)
+          end
+        end
+      end
+    end
+  end
+
+  def self.add_times(mf, obj)
+    mf.css("*[class*=t-]").each do |property|
+      property.attribute("class").to_s.split.each do |css_class|
+        if css_class =~ /^t/
+          css_class   = css_class[2..-1].gsub("-","_")
+          method_name = css_class.gsub("-","_")
+          value       = Time.parse((property.attribute("title") || property.text).to_s)
+
+          add_method(obj, method_name)
+
+          if cur = obj.send(method_name)
+            if cur.kind_of? Array
+              cur << value
+            else
+              obj.send("#{method_name}=", [cur, value])
+            end
+          else
+            obj.send("#{method_name}=", value)
+          end
+        end
+      end
+    end
+  end
+
+  # Thank you Rails Developers for your unitentional contribution to this project
+  # File activesupport/lib/active_support/inflector/inflections.rb, line 206
+  def self.classify(str)
+    # strip out any leading schema name
+    camelize(singularize(str.to_s.sub(/.*\./, '')))
+  end
+
+  # File activesupport/lib/active_support/inflector/inflections.rb, line 148
+  def self.singularize(word)
+    result = word.to_s.dup
+  end
+
+  # File activesupport/lib/active_support/inflector/methods.rb, line 28
+  def self.camelize(lower_case_and_underscored_word, first_letter_in_uppercase = true)
+    if first_letter_in_uppercase
+      lower_case_and_underscored_word.to_s.gsub(/\/(.?)/) { "::#{$1.upcase}" }.gsub(/(?:^|_)(.)/) { $1.upcase }
+    else
+      lower_case_and_underscored_word.to_s[0].chr.downcase + camelize(lower_case_and_underscored_word)[1..-1]
+    end
+  end
+end

data/test/test_microformats2.rb

@@ -0,0 +1,163 @@
+require "test/unit"
+require "microformats2"
+
+class TestMicroformats2 < Test::Unit::TestCase
+  def test_throw_exception_on_non_string_params
+    assert_raise Microformats2::LoadError do
+      Microformats2.parse(nil)
+    end
+  end
+
+  def test_returns_hash_of_microformat_objects
+    result = Microformats2.parse("A String")
+    assert_equal Hash, result.class
+  end
+
+  def test_only_parse_microformats
+    result = Microformats2.parse("<html><body><p>Something</p></body></html>")
+    assert_equal 0, result.size
+  end
+
+  def test_extracts_hcard_from_an_html_file
+    hcard = <<-END
+    <html>
+    <head>
+      <title>Simple hCard</title>
+    </head>
+
+    <body>
+      <h1 class="h-card">Chris</h1>
+    </body>
+    </html>
+    END
+    result = Microformats2.parse(File.open(File.join(File.dirname(__FILE__), "hcard.html")))
+    assert_equal HCard, result[:hcard].first.class
+  end
+
+  def test_extracts_hcard_from_html
+    hcard = <<-END
+    <html>
+    <head>
+      <title>Simple hCard</title>
+    </head>
+
+    <body>
+      <h1 class="h-card">Chris</h1>
+    </body>
+    </html>
+    END
+    result = Microformats2.parse(hcard)
+    assert_equal HCard, result[:hcard].first.class
+  end
+
+  def test_constructs_properties_from_hcard
+    hcard = <<-END
+    <html>
+    <head>
+      <title>Simple hCard</title>
+    </head>
+
+    <body>
+      <h1 class="h-card">
+        <a class="p-fn u-url" href="http://factoryjoe.com/">
+          <span class="p-given-name">Chris</span>
+          <abbr class="p-additional-name">R.</abbr>
+          <span class="p-family-name">Messina</span>
+        </a>
+      </h1>
+    </body>
+    </html>
+    END
+    result = Microformats2.parse(hcard)
+    mycard = result[:hcard].first
+
+    assert_equal "Chris", mycard.given_name
+    assert_equal "R.", mycard.additional_name
+    assert_equal "Messina", mycard.family_name
+    assert_equal "Chris R. Messina", mycard.fn
+  end
+
+  def test_constructs_dates
+    hcard = <<-END
+    <html>
+    <head>
+      <title>Simple hCard</title>
+    </head>
+
+    <body>
+      <h1 class="h-card">
+        <span class="d-bday">1979-09-18</span>
+        <span class="d-epoch" title="1970-01-01">EPOCH!</span>
+      </h1>
+    </body>
+    </html>
+    END
+    result = Microformats2.parse(hcard)
+    mycard = result[:hcard].first
+
+    assert_equal DateTime.parse("1979-09-18"), mycard.bday
+    assert_equal DateTime.parse("1970-01-01"), mycard.epoch
+  end
+
+  def test_constructs_times
+    hcard = <<-END
+    <html>
+    <head>
+      <title>Simple hCard</title>
+    </head>
+
+    <body>
+      <h1 class="h-card">
+        <span class="t-start">09:30</span>
+        <span class="t-end" title="6:00">Leaving time</span>
+      </h1>
+    </body>
+    </html>
+    END
+    result = Microformats2.parse(hcard)
+    mycard = result[:hcard].first
+
+    assert_equal Time.parse("09:30"), mycard.start
+    assert_equal Time.parse("06:00"), mycard.end
+  end
+
+  def test_ignores_pattern_matches_not_at_the_beginning_of_class
+    hcard = <<-END
+    <html>
+    <head>
+      <title>Simple hCard</title>
+    </head>
+
+    <body>
+      <h1 class="h-card">
+        <span class="p-n-x">Chris</span>
+      </h1>
+    </body>
+    </html>
+    END
+    result = Microformats2.parse(hcard)
+    mycard = result[:hcard].first
+
+    assert_equal "Chris", mycard.n_x
+    assert mycard.n_x.is_a?(String)
+  end
+
+  def test_constructs_urls_from_hcard
+    hcard = <<-END
+    <html>
+    <head>
+      <title>Simple hCard</title>
+    </head>
+
+    <body>
+      <h1 class="h-card">
+        <a class="p-fn u-url" href="http://factoryjoe.com/">Chris</a>
+      </h1>
+    </body>
+    </html>
+    END
+    result = Microformats2.parse(hcard)
+    mycard = result[:hcard].first
+    assert_equal "http://factoryjoe.com/", mycard.url
+  end
+end

metadata

@@ -0,0 +1,105 @@
+--- !ruby/object:Gem::Specification 
+name: microformats2
+version: !ruby/object:Gem::Version 
+  hash: 23
+  prerelease: 
+  segments: 
+  - 1
+  - 0
+  - 0
+  version: 1.0.0
+platform: ruby
+authors: 
+- Shane Becker
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2011-06-25 00:00:00 Z
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: nokogiri
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  type: :runtime
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
+  name: hoe
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        hash: 17
+        segments: 
+        - 2
+        - 9
+        version: "2.9"
+  type: :development
+  version_requirements: *id002
+description: Generic Microformats 2 Extractor
+email: 
+- veganstraightedge@gmail.com
+executables: 
+- microformats2
+extensions: []
+
+extra_rdoc_files: 
+- History.txt
+- Manifest.txt
+- README.txt
+files: 
+- .autotest
+- History.txt
+- Manifest.txt
+- README.txt
+- Rakefile
+- bin/microformats2
+- lib/microformats2.rb
+- test/test_microformats2.rb
+- .gemtest
+homepage: http://github.com/veganstraightedge/microformats2
+licenses: []
+
+post_install_message: 
+rdoc_options: 
+- --main
+- README.txt
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+requirements: []
+
+rubyforge_project: microformats2
+rubygems_version: 1.8.5
+signing_key: 
+specification_version: 3
+summary: Generic Microformats 2 Extractor
+test_files: 
+- test/test_microformats2.rb