web_stat 0.3.2 → 0.3.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: eff1d2b80ff6bf4b61f82a100319a6b37ee13cb8a2ad96a11fd336c8a7164398
-  data.tar.gz: f4bf36491dbe6ae32b1cbf718ceb1fe7e91adce60848f4296d6c1319ea5c4eea
+  metadata.gz: 5304507f8ce01a6a98717756927366005db0eceb89b3d22f29e4283c6eedb4d7
+  data.tar.gz: f561221644f98831867c2d853201a0312dcb75674ca4ae247b319b4f2a1ba085
 SHA512:
-  metadata.gz: f38fe9d4f0ad495107c5ac42d2d240b42e810ae4f7ad0db5738b705f7e03b24846875cbc2139a334ebfc7d33ff1df953dec768f69f13a0932cd01a2c6f221753
-  data.tar.gz: 5f30259af5a84e2eb43645766aac49d151367a82c9e5643a0ddf3a5f15dbcbba59884def134cfaa19eaca1a932e77c28d65bfd389cecdd0262812cad808ff5e3
+  metadata.gz: bec01871973e80dc46b9246a264791213b3bdd4cb20693fc0d51cc28c2d9bba28b40af65322dc2b0c950e9bb5aedd288ff8fd771adfbd5b2aee1564b98888cb0
+  data.tar.gz: 5cfce031ad779f347fae362ccff2deadb3711e4f9fcc06cafcd232b86e8c18aff63e40c430a330cfdf00d398aa5257457c5dc9f3bee492478359d3801e217f08

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    web_stat (0.3.2)
+    web_stat (0.3.6)
       bundler (>= 2.0.2)
       cld (>= 0.8.0)
       mechanize (>= 2.7)

data/lib/web_stat/config/web_stat.yml

@@ -1,15 +1,21 @@
-# Minimum number of characters to detect meta title
-min_length_of_meta_title: 10
-# Split regular expression for titles 
-regex_to_sprit_title: '\||-|:|｜|：|〜|\~| – '
-# User Agent
-user_agent: "web_stat gem agent"
-# Eyecatch image xpaths
-eyecatch_image_xpaths:
-  - '/html/head/meta[@property="twitter:image"]/@content'
-  - '/html/head/meta[@property="og:image"]/@content'
-  - '//img[@class="attachment-post-thumbnail"]/@src'
-  - '//div[@id="content"]//img/@src'
-  - '//img/@src'
-userdic: ""
-use_chromedirver: true
+development: &development
+  # Minimum number of characters to detect meta title
+  min_length_of_meta_title: 10
+  # Split regular expression for titles 
+  regex_to_sprit_title: '\||-|:|｜|：|〜|\~| – '
+  # User Agent
+  user_agent: "web_stat gem agent"
+  # Eyecatch image xpaths
+  eyecatch_image_xpaths:
+    - '/html/head/meta[@property="twitter:image"]/@content'
+    - '/html/head/meta[@property="og:image"]/@content'
+    - '//img[@class="attachment-post-thumbnail"]/@src'
+    - '//div[@id="content"]//img/@src'
+    - '//img/@src'
+  userdic: ""
+  use_chromedirver: false
+test:
+  <<: *development
+production:
+  <<: *development
+  use_chromedirver: true

data/lib/web_stat/configure.rb

@@ -3,31 +3,37 @@ module WebStat
   class Configure
     DEFAULT_CONFIG_FILE_PATH = 'config/web_stat.yml'
     
-    # Get yaml
-    def self.get
-      YAML.load_file(self.get_configure_path)
-    end
-    
-    # Get configure path
-    def self.get_configure_path
-      if File.exists?(self.get_custom_configure_path)
-        self.get_custom_configure_path
-      else
-        self.get_default_configure_path
+    class << self
+      # Get yaml
+      def get
+        if defined? Rails
+          YAML.load_file(get_configure_path)[Rails.env]
+        else
+          YAML.load_file(get_configure_path)["production"]
+        end
       end
-    end
-    
-    # Get default configure path
-    def self.get_default_configure_path
-      File.join(File.expand_path("../", __FILE__), DEFAULT_CONFIG_FILE_PATH)
-    end
-    
-    # Get custom configure path
-    def self.get_custom_configure_path
-      if defined? Rails
-	File.join(Rails.root, DEFAULT_CONFIG_FILE_PATH)
-      else
-        File.join(Bundler.root, DEFAULT_CONFIG_FILE_PATH)
+      
+      # Get configure path
+      def get_configure_path
+        if File.exists?(get_custom_configure_path)
+          get_custom_configure_path
+        else
+          get_default_configure_path
+        end
+      end
+      
+      # Get default configure path
+      def get_default_configure_path
+        File.join(File.expand_path("../", __FILE__), DEFAULT_CONFIG_FILE_PATH)
+      end
+      
+      # Get custom configure path
+      def get_custom_configure_path
+        if defined? Rails
+  	      File.join(Rails.root, DEFAULT_CONFIG_FILE_PATH)
+        else
+          File.join(Bundler.root, DEFAULT_CONFIG_FILE_PATH)
+        end
       end
     end
   end

data/lib/web_stat/fetch.rb

@@ -47,6 +47,9 @@ module WebStat
           break
         end
       end
+      if path.nil? || path.empty?
+        path = @nokogiri.at('body').xpath('//img').first.attr('src')
+      end
       if ! path.nil? && path.match(/^\//)
         "#{URI.parse(@url).scheme}://#{URI.parse(@url).host}#{path}"
       else
@@ -57,7 +60,7 @@ module WebStat
     # Get local path to save url
     # @param [String] url
     def save_local_path(url)
-      return nil if url.nil?
+      return nil if url.nil? || ! url.match(%{^http})
       tmp_file = "/tmp/#{Digest::SHA1.hexdigest(url)}"
       agent = Mechanize.new { |_agent| _agent.user_agent = WebStat::Configure.get["user_agent"] }
       image = agent.get(url)
@@ -96,7 +99,7 @@ module WebStat
     # Get the informations of @url
     # @param [Hash] Specify a dictionary for each language code. example ) {"ja": /***/**.dic, "other": /***/***.dic}
     def stat(userdics: nil)
-      clean_content = content.scrub('').gsub(/[\n\t\r　]/, "").gsub(/\s{2,}/, "\s")
+      clean_content = content.scrub('').gsub(/[\n\t\r　]/, "").gsub(/\s{2,}/, "\s").gsub(URI.regexp, "")
       language_code = CLD.detect_language(clean_content)[:code]
       if userdics && userdics.has_key?(language_code) && File.exists?(userdics[language_code])
         tag = WebStat::Tag.new("#{title} #{content}", userdic: userdics[language_code])

data/lib/web_stat/version.rb

@@ -1,3 +1,3 @@
 module WebStat
-  VERSION = "0.3.2"
+  VERSION = "0.3.7"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: web_stat
 version: !ruby/object:Gem::Version
-  version: 0.3.2
+  version: 0.3.7
 platform: ruby
 authors:
 - yusuke abe
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2020-06-13 00:00:00.000000000 Z
+date: 2020-06-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler