textract 0.0.5 → 0.0.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 7dfaec842302697fff5fb526bc358045d63aeabb
-  data.tar.gz: 6bfa752901cf6bf183ea50370100701224a49ba8
+  metadata.gz: 2b777157b81de9ecfc3d2aff6c9a88e8413c23fc
+  data.tar.gz: ed98b845d1819d2fbb0ab9ac8172ba95b308b227
 SHA512:
-  metadata.gz: f667c4d08fbc1658d1ac98ccd5e79eae4871a51e18e425717573903d2c14f3fceb4f9d37068a410dc01c88e58cf4b6264f26ead1708d8ca867bf430daf439a5d
-  data.tar.gz: cfef813ebb6fb91e37a71c952e06a56cee13520497a198f480d7f090eebc1b598f9f6dbf67bce8de1f0fa9fbee2d7595e8e5842077d14e6b06dca34b84e0c64d
+  metadata.gz: 1ba20d4cb668c8da2ee0a53a9585433b98b0c576e67771084345b57393b5a471cc5191db60654fcdfbf335ffc28478f46397460e0c788fb18a53a703d1fda5b9
+  data.tar.gz: 15de095a7a95ffba2a8263b2c67a062a42814c4e7da32045b7be53d1dcc421ebdf5141581607bfbdf87de80fa60b8e8130b73d43bdd1d1cfe2b36451553310ad

data/lib/textract/version.rb

@@ -1,3 +1,3 @@
 module Textract
-  VERSION = "0.0.5"
+  VERSION = "0.0.6"
 end

data/lib/textract.rb

@@ -16,7 +16,7 @@ module Textract
     OpenGraph.new(html)
   end
 
-  def self.get_text_from_description(html, description, selectors)
+  def self.smart_extract(html, description, selectors)
     doc = Nokogiri::HTML html
     if selectors.nil?
       article = doc.search('article')
@@ -72,24 +72,28 @@ module Textract
     attr_reader :md5
     attr_reader :author
 
-    def initialize(url, selectors)
+    def initialize(url, selectors, format="markdown")
       @url = url
       agent = Mechanize.new
       agent.user_agent_alias = 'Mac Safari'
       @html = agent.get(url).content
       @tags = Textract.get_og_tags(@html)
 
-      @article = Textract.get_text_from_description(@html, @tags.description, selectors)
-      @text = ReverseMarkdown.convert @article.content, unknown_tags: :bypass
+      @article = Textract.smart_extract(@html, @tags.description, selectors)
+      if @article.content.nil?
+        @text = ""
+      else
+        if format == 'markdown'
+          @text = ReverseMarkdown.convert @article.content, unknown_tags: :bypass
+        else
+          @text = @article.content
+        end
+      end
       @md5 = Textract.generate_hash @text
       @author = @article.author || Textract.get_author(@html)
       @title = @tags.title || Textract.get_page_title(@html)
     end
 
-    def to_json
-      to_h.to_json
-    end
-
     def to_h
       {
         url: @url,

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: textract
 version: !ruby/object:Gem::Version
-  version: 0.0.5
+  version: 0.0.6
 platform: ruby
 authors:
 - Adam Pash
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-03-30 00:00:00.000000000 Z
+date: 2015-04-10 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: opengraph_parser