seotracker 0.2.1 → 0.2.2

data/Gemfile.lock

@@ -1,8 +1,8 @@
 GEM
   remote: http://rubygems.org/
   specs:
-    coderay (1.0.5)
-    domain_name (0.5.2)
+    coderay (1.0.6)
+    domain_name (0.5.3)
       unf (~> 0.0.3)
     git (1.2.5)
     jeweler (1.8.3)
@@ -21,7 +21,7 @@ GEM
       webrobots (~> 0.0, >= 0.0.9)
     method_source (0.7.1)
     mime-types (1.18)
-    minitest (2.11.4)
+    minitest (2.12.1)
     net-http-digest_auth (1.2)
     net-http-persistent (2.6)
     nokogiri (1.5.2)

data/VERSION

@@ -1 +1 @@
-0.2.1
+0.2.2

data/lib/seotracker.rb

@@ -1,6 +1,7 @@
 require 'rubygems'
 require 'mechanize'
 require 'logger'
+#require 'pry'
 
 class Seotracker
   USER_AGENT = 'Mac Safari'
@@ -20,18 +21,18 @@ class Seotracker
     end
   end
 
-  def get_position(site, word, pages = 200)
+  def get_position(site, word, region = Seotracker::Yandex::MOSCOW, pages = 200)
     pos, found, start = 0, false, 0
     while (start < pages) && !found
-      links = parse(word, start)
+      links = parse(word, start, region)
       start += RESULTS
 
       break if links == 'error'
-
       links.each do |l|
-        pos += 1
         href = l.attribute('href').value.downcase
-        unless href.rindex(site).nil?
+        # убеждаемся, что это точно ссылка. а то бывает еще адрес
+        pos += 1 if l.content.match(/.+\..+/)
+        if href.rindex(site) && same_level(href, site)
           found = true
           break
         end
@@ -43,6 +44,17 @@ class Seotracker
   def debug(message)
     @log.debug(message) if @debug
   end
+
+  protected
+
+    # убеждаемся, что сайт, для которого определяем позицию (site)
+    # и найденная ссылка (href) на одном и том же уровне
+    # например, если искали позицию ya.ru
+    # а сначала нашлась позиция maps.ya.ru то ее не считаем
+    def same_level(href, site)
+      return href.count('.') - 1 == site.count('.') if href.index('www')
+      href.count('.') == site.count('.')
+    end
 end
 
 require 'seotracker/yandex'

data/lib/seotracker/google.rb

@@ -3,7 +3,7 @@ class Seotracker::Google < Seotracker
 
   protected
 
-  def parse(word, start = 0)
+  def parse(word, start = 0, region = nil)
     page = @agent.get(SEARCH_URL, q: word, start: start)
     page.root.xpath('/html/body/div[5]/div/div/div[4]/div[2]/div[2]/div/div[2]/div/ol/li/div/h3/a')
   end

data/lib/seotracker/yandex.rb

@@ -2,33 +2,35 @@ class Seotracker::Yandex < Seotracker
   SEARCH_URL = 'http://yandex.ru/yandsearch?'
   WORDSTAT_URL = 'http://wordstat.yandex.ru/?cmd=words&page=1&&geo=&text_geo=&text='
 
+  MOSCOW = 213
+
   # 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})
+    @cookie || get_cookie
+    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
+      debug "can't get yandex wordstat: #{e.message}"
       0
     end
   end
 
   protected
 
-  def parse(word, start = 0)
+  # начинаем парсить с первой страницы, регион по умолчанию - Москва
+  def parse(word, start = 0, region = nil)
     start /= 10
-    get_cookie if @cookie.nil?
-    url = SEARCH_URL + "text=#{word}&p=#{start}"
-    page = @agent.get(url, [], nil, {'cookie' => @cookie})
-
+    @cookie || get_cookie
+    url = SEARCH_URL + "text=#{word}&p=#{start}&lr=#{region}"
+    page = @agent.get(url, [], nil, { 'cookie' => @cookie })
     begin
-      page.root.xpath('/html/body/div[3]/div/div/div[2]/ol/li/div/h2/a')
+      page.root.xpath('/html/body/div[2]/div/div/div/ol/li/div/div/span/span/a')
     rescue Exception => e
-      debug "can't parse yandex:" + e.message
+      debug "can't parse yandex: #{e.message}"
       'error'
     end
   end

data/lib/seotracker/yandex/direct.rb

@@ -1,9 +1,9 @@
 # encoding: utf-8
 
 class Seotracker::Yandex::Direct < Seotracker::Yandex
-  def special(word)
+  def special(word, region = MOSCOW)
     get_cookie
-    url = SEARCH_URL + "text=#{word}&p=0"
+    url = SEARCH_URL + "text=#{word}&p=0&lr=#{region}"
     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')
@@ -21,7 +21,7 @@ class Seotracker::Yandex::Direct < Seotracker::Yandex
 
       result
     rescue Exception => e
-      debug "can't parse yandex direct:" + e.message
+      debug "can't parse yandex direct: #{e.message}"
       'error'
     end
   end

data/seotracker.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = "seotracker"
-  s.version = "0.2.1"
+  s.version = "0.2.2"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["krivich ekaterina"]
-  s.date = "2012-04-08"
+  s.date = "2012-04-11"
   s.description = "track sites position in google or yandex by keyword"
   s.email = "kiote_the_one@mail.ru"
   s.extra_rdoc_files = [

data/spec/lib/seotracker_spec.rb

@@ -4,7 +4,7 @@ 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(:get, mock, [Seotracker::Yandex::SEARCH_URL  + "text=#{@word}&p=0&lr=213",  [], nil, {'cookie' => 'hi'}])
   mock.expect(:cookies, ['hi'])
 end
 
@@ -13,6 +13,7 @@ def common_mocker
   mock.expect(:root, mock)
   mock.expect(:attribute, mock, %w/href/)
   mock.expect(:value, "http://#{@site}")
+  mock.expect(:content, 'yandex.ru')
 end
 
 describe Seotracker do
@@ -32,7 +33,7 @@ describe Seotracker do
       # мокаем все неважное
       mock = common_mocker
       yandex_mocker(mock)
-      mock.expect(:xpath, [mock], %w\/html/body/div[3]/div/div/div[2]/ol/li/div/h2/a\)
+      mock.expect(:xpath, [mock], %w\/html/body/div[2]/div/div/div/ol/li/div/div/span/span/a\)
 
       @object.instance_variable_set(:@agent, mock)
     end
@@ -69,7 +70,6 @@ describe Seotracker::Yandex::Direct do
     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

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: seotracker
 version: !ruby/object:Gem::Version
-  version: 0.2.1
+  version: 0.2.2
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-04-08 00:00:00.000000000 Z
+date: 2012-04-11 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: minitest
-  requirement: &70224553771600 !ruby/object:Gem::Requirement
+  requirement: &70221806624540 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -21,10 +21,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70224553771600
+  version_requirements: *70221806624540
 - !ruby/object:Gem::Dependency
   name: jeweler
-  requirement: &70224553771100 !ruby/object:Gem::Requirement
+  requirement: &70221806623740 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -32,10 +32,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70224553771100
+  version_requirements: *70221806623740
 - !ruby/object:Gem::Dependency
   name: pry
-  requirement: &70224553770340 !ruby/object:Gem::Requirement
+  requirement: &70221806622960 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -43,7 +43,7 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70224553770340
+  version_requirements: *70221806622960
 description: track sites position in google or yandex by keyword
 email: kiote_the_one@mail.ru
 executables: []
@@ -80,7 +80,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: -150072496354741973
+      hash: -3070310031504885579
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements: