http_crawler 0.3.1.6 → 0.3.1.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 781a489a3c6d8b0e6d09f79c58d09593372a36cb2faf57dc93194e4b265b2021
-  data.tar.gz: 108cb3d5763cc6ed49b71cdf4526212d4851a314db3440b0d50df4520bd52840
+  metadata.gz: 0d1fd5d70b3339e212db19695ab111c2fdef94f8f76fbf912709ab079e36bf25
+  data.tar.gz: ca1007ea22b501cc30d8209ba1ab55bc7128250cbb71859fe37e55ba470a2e13
 SHA512:
-  metadata.gz: 53c07f32f718d5a86c61988987b3a4a8c32b1b70e75a3e4735c2dc18561c94082cbf537988ca594de65282e697556db769d7256fc1a7a203a289950ae02e12cf
-  data.tar.gz: e37325d54b10229d39b118e878782cfb3827a370d48280d86615121b03a0feb8b7d2b655e41abc76bf799e2898840c1085fb3ab02e5d6a12988279b618577217
+  metadata.gz: ee63e0d0c13240c1ca443ca3788ab10dc1393e44d4e6b3f6b907e0cb45a302ac7f5eb38fc1888993b0e8e5db3ae44f0561cb7bfb732907b073d41328e7edfe3c
+  data.tar.gz: 052b85418bfad358af10778e2cbaa1fe90dbce61fea662f713ffc2a58bcc3c69e6271096dda990def0ba0dbceec023388df26b832cbdf99d4f93c6b2fa8d7e39

data/lib/http_crawler/common/string.rb

@@ -4,6 +4,6 @@ class String
   # 清除包含: 空格,回车
   #
   def del_inter
-    self.gsub(/(?:\n|\t| )/,"")
+    self.gsub(/(?:\n|\t|\r| )/,"")
   end
 end

data/lib/http_crawler/http/response.rb

@@ -57,32 +57,38 @@ module HTTP
 
     # 转换html格式
     # @return [Nokogiri::HTML::Document]
-    def html
-      return @html if defined? @html
-      self.html = self.dec
-      self.html
-    end
+    def html(data = nil)
 
-    # @return [Nokogiri::HTML::Document]
-    def html=(data)
+      if (data.blank? && defined? @html)
+        # 如果 data 为空 并且 @json 有值，直接返回 @json
+        return @html
+      end
+
+      # 如果data为空初始化为 self.dec
+      data ||= self.dec
       if (Nokogiri::HTML::Document === data)
         @html = data
       else
         @html = Nokogiri::HTML(data)
       end
       @html
+
+      return @html if defined? @html
+      self.html = self.dec
+      self.html
     end
 
+
     # 转换json格式
     # @return [Hash]
-    def json
-      return @json if defined? @json
-      self.json = self.dec
-      self.json
-    end
+    def json(data = nil)
+      if (data.blank? && defined? @json)
+        # 如果 data 为空 并且 @json 有值，直接返回 @json
+        return @json
+      end
 
-    # @return [Hash]
-    def json=(data)
+      # 如果data为空初始化为 self.dec
+      data ||= self.dec
       if (Hash === data)
         @json = data
       else
@@ -94,14 +100,14 @@ module HTTP
 
     # 通过readability 解析数据
     # @return [Readability::Document]
-    def readability
-      return @readability if defined? @readability
-      self.readability = self.dec
-      self.readability
-    end
+    def readability(data = nil)
+      if (data.blank? && defined? @readability)
+        # 如果 data 为空 并且 @readability 有值，直接返回 @readability
+        return @readability
+      end
 
-    # @return [Readability::Document]
-    def readability=(data)
+      # 如果data为空初始化为 self.dec
+      data ||= self.dec
       if (Readability::Document === data)
         @readability = data
       else
@@ -110,20 +116,20 @@ module HTTP
       @readability
     end
 
-
-    def content
-      Nokogiri::HTML(readability.content).text.del_inter
+    # 获取正文内容
+    def content(data = readability.content)
+      Nokogiri::HTML(data).text.del_inter
     end
 
     # 解析
     # 默认使用 json 的值
-    def parsing
+    def parsing(parameter = {})
       self.json
     end
 
     # 获取解析结果
-    def results
-      @results ||= parsing
+    def results(parameter = {})
+      @results ||= parsing(parameter)
     end
 
     def get_date(str)

data/lib/http_crawler/version.rb

@@ -1,3 +1,3 @@
 module HttpCrawler
-  VERSION = "0.3.1.6"
+  VERSION = "0.3.1.7"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: http_crawler
 version: !ruby/object:Gem::Version
-  version: 0.3.1.6
+  version: 0.3.1.7
 platform: ruby
 authors:
 - jagger
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2019-03-08 00:00:00.000000000 Z
+date: 2019-03-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec