seotracker 0.1.0 → 0.2.0

data/Gemfile

@@ -3,3 +3,4 @@ source :rubygems
 gem 'minitest'
 gem 'mechanize'
 gem 'jeweler'
+#gem 'pry'

data/README.rdoc

@@ -6,3 +6,5 @@ seot = Seotracker::Google.new
 seot.get_position('serialmaniak.ru', 'серийный маньяк')
 
 -> 1 
+
+Больше информации на странице: http://kiote.github.com/SeoTracker

data/VERSION

@@ -1 +1 @@
-0.1.0
+0.2.0

data/lib/seotracker/yandex/direct.rb

@@ -0,0 +1,30 @@
+# encoding: utf-8
+
+class Seotracker::Yandex::Direct < Seotracker::Yandex
+  def special(word)
+    get_cookie
+    url = SEARCH_URL + "text=#{word}&p=0"
+    page = @agent.get(url, [], nil, {'cookie' => @cookie})
+    begin
+      texts = page.root.xpath('/html/body/div[3]/div/div/div/div/div[2]/div/h2/a')
+      hrefs = page.root.xpath('/html/body/div[3]/div/div/div/div/div/div/div/div/span')
+
+      # pre_result - массив вида ['text','text','text','url','url','url']
+      pre_result = []
+      texts.map { |t| pre_result << t.content.strip.squeeze("\n") }
+      hrefs.map { |h| pre_result << h.content.scan(/\w+\.\w+/).first unless h.content.scan(/\w+\.\w+/)== [] }
+      raise Seotracker::Yandex::Direct::Exception if pre_result.count != 6
+
+      # приводим pre_result к виду [{ad: text, url: url}}]
+      result = []
+      (0...3).to_a.each { |i| result << { ad: pre_result[i], url: pre_result[i + 3] } }
+
+      result
+    rescue Exception => e
+      debug "can't parse yandex direct:" + e.message
+      'error'
+    end
+  end
+end
+
+class Seotracker::Yandex::Direct::Exception < Exception; end

data/lib/seotracker/yandex.rb

@@ -1,24 +1,41 @@
 class Seotracker::Yandex < Seotracker
-  SEARCH_URL = "http://yandex.ru/yandsearch?"
+  SEARCH_URL = 'http://yandex.ru/yandsearch?'
+  WORDSTAT_URL = 'http://wordstat.yandex.ru/?cmd=words&page=1&&geo=&text_geo=&text='
+
+  # получаем стастистику wordstat
+  # word - слово, по которому получаем статистику
+  def get_wordstat(word)
+    url = WORDSTAT_URL + word
+    get_cookie if @cookie.nil?
+    page = @agent.get(url, [], nil, {'cookie' => @cookie})
+    res = page.root.xpath('/html/body/form/table[2]/tbody/tr/td[4]/table/tbody/tr[3]/td/table/tbody/tr[2]/td[3]')
+    begin
+      res.first.content
+    rescue Exception => e
+      debug "can't get yandex wordstat:" + e.message
+      0
+    end
+  end
 
   protected
 
   def parse(word, start = 0)
     start /= 10
+    get_cookie if @cookie.nil?
+    url = SEARCH_URL + "text=#{word}&p=#{start}"
+    page = @agent.get(url, [], nil, {'cookie' => @cookie})
 
-    # у яндекса хитрая проверка на роботов
-    if start == 0
-      @agent.get('http://kiks.yandex.ru/su/')
-      @cookie = @agent.cookies.first
+    begin
+      page.root.xpath('/html/body/div[3]/div/div/div[2]/ol/li/div/h2/a')
+    rescue Exception => e
+      debug "can't parse yandex:" + e.message
+      'error'
     end
+  end
 
-    #agent = Mechanize.new
-    #agent.user_agent_alias = Seotracker::USER_AGENT
-    ##agent.log = Logger.new('log.txt')
-
-    url = SEARCH_URL + "text=#{word}&p=#{start}"
-
-    page = @agent.get(url, [], nil, {'cookie' => @cookie})
-    page.root.xpath('/html/body/div[3]/div/div/div[2]/ol/li/div/h2/a')
+  def get_cookie
+    # у яндекса хитрая проверка на роботов
+    @agent.get('http://kiks.yandex.ru/su/')
+    @cookie = @agent.cookies.first
   end
-end
+end

data/lib/seotracker.rb

@@ -11,6 +11,13 @@ class Seotracker
     @agent.user_agent_alias = Seotracker::USER_AGENT
     @agent.agent.http.verify_mode = OpenSSL::SSL::VERIFY_NONE
     @agent.agent.http.retry_change_requests = true
+
+    @debug = true
+
+    if @debug
+      @log = Logger.new('log.txt')
+      #@agent.log = @log если надо логировать mechnize
+    end
   end
 
   def get_position(site, word)
@@ -18,6 +25,9 @@ class Seotracker
     while (start < 200) && !found
       links = parse(word, start)
       start += RESULTS
+
+      break if links == 'error'
+
       links.each do |l|
         pos += 1
         href = l.attribute('href').value.downcase
@@ -29,7 +39,12 @@ class Seotracker
     end
     found ? pos : 0
   end
+
+  def debug(message)
+    @log.debug(message) if @debug
+  end
 end
 
 require 'seotracker/yandex'
+require 'seotracker/yandex/direct'
 require 'seotracker/google'

data/seotracker.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = "seotracker"
-  s.version = "0.1.0"
+  s.version = "0.2.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["krivich ekaterina"]
-  s.date = "2012-04-03"
+  s.date = "2012-04-06"
   s.description = "track sites position in google or yandex by keyword"
   s.email = "kiote_the_one@mail.ru"
   s.extra_rdoc_files = [
@@ -26,6 +26,7 @@ Gem::Specification.new do |s|
     "lib/seotracker.rb",
     "lib/seotracker/google.rb",
     "lib/seotracker/yandex.rb",
+    "lib/seotracker/yandex/direct.rb",
     "seotracker.gemspec",
     "spec/lib/seotracker_spec.rb",
     "spec/spec_helper.rb"

data/spec/lib/seotracker_spec.rb

@@ -2,15 +2,23 @@
 $LOAD_PATH << File.expand_path('../../', __FILE__)
 require 'spec_helper'
 
+def yandex_mocker(mock)
+  mock.expect(:get, mock, ['http://kiks.yandex.ru/su/'])
+  mock.expect(:get, mock, [Seotracker::Yandex::SEARCH_URL  + "text=#{@word}&p=0",  [], nil, {'cookie' => 'hi'}])
+  mock.expect(:cookies, ['hi'])
+end
+
+def common_mocker
+  mock = MiniTest::Mock.new
+  mock.expect(:root, mock)
+  mock.expect(:attribute, mock, %w/href/)
+  mock.expect(:value, "http://#{@site}")
+end
+
 describe Seotracker do
   # общие моки, тут от природной лени несколько объектов используют один мок
   # что может несколько напрягать в дальнейшем
-  def common_mocker
-    mock = MiniTest::Mock.new
-    mock.expect(:root, mock)
-    mock.expect(:attribute, mock, %w/href/)
-    mock.expect(:value, "http://#{@site}")
-  end
+
 
   before do
     @site = 'serialmaniak.ru'
@@ -23,9 +31,7 @@ describe Seotracker do
 
       # мокаем все неважное
       mock = common_mocker
-      mock.expect(:get, mock, [Seotracker::Yandex::SEARCH_URL  + "text=#{@word}&p=0",  [], nil, {'cookie' => 'hi'}])
-      mock.expect(:get, mock, ['http://kiks.yandex.ru/su/'])
-      mock.expect(:cookies, ['hi'])
+      yandex_mocker(mock)
       mock.expect(:xpath, [mock], %w\/html/body/div[3]/div/div/div[2]/ol/li/div/h2/a\)
 
       @object.instance_variable_set(:@agent, mock)
@@ -52,4 +58,23 @@ describe Seotracker do
       @object.get_position(@site, @word).must_be :>, 0
     end
   end
+end
+
+describe Seotracker::Yandex::Direct do
+  before do
+    @client = Seotracker::Yandex::Direct.new
+    @word = 'окна'
+
+    mock =common_mocker
+    yandex_mocker(mock)
+    mock.expect(:xpath, [mock, mock, mock], %w\/html/body/div[3]/div/div/div/div/div[2]/div/h2/a\)
+    mock.expect(:xpath, [mock, mock, mock], %w\/html/body/div[3]/div/div/div/div/div/div/div/div/span\)
+    mock.expect(:content, 'yandex.ru')
+
+    @client.instance_variable_set(:@agent, mock)
+  end
+
+  it 'should get valid special' do
+    @client.special(@word).count.must_equal 3
+  end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: seotracker
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-04-03 00:00:00.000000000 Z
+date: 2012-04-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: minitest
-  requirement: &70225702846080 !ruby/object:Gem::Requirement
+  requirement: &70312546860700 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -21,10 +21,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70225702846080
+  version_requirements: *70312546860700
 - !ruby/object:Gem::Dependency
   name: mechanize
-  requirement: &70225702845080 !ruby/object:Gem::Requirement
+  requirement: &70312546859880 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -32,10 +32,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70225702845080
+  version_requirements: *70312546859880
 - !ruby/object:Gem::Dependency
   name: jeweler
-  requirement: &70225702844120 !ruby/object:Gem::Requirement
+  requirement: &70312546859300 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -43,7 +43,7 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70225702844120
+  version_requirements: *70312546859300
 description: track sites position in google or yandex by keyword
 email: kiote_the_one@mail.ru
 executables: []
@@ -61,6 +61,7 @@ files:
 - lib/seotracker.rb
 - lib/seotracker/google.rb
 - lib/seotracker/yandex.rb
+- lib/seotracker/yandex/direct.rb
 - seotracker.gemspec
 - spec/lib/seotracker_spec.rb
 - spec/spec_helper.rb
@@ -79,7 +80,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: -195465089122756196
+      hash: -713848669405107154
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements: