curation 1.7 → 1.10

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e3e4d559a2583393e2d58cba1bfe30ef8870733ce013c00780b29706b9bfce68
-  data.tar.gz: 5802c61d564cfc2568f9173f6b270e3e288f4dbdfcfec624b02205bd01a81aaf
+  metadata.gz: a23e967e5d017ce61f9719647f45c20e1336aff067bc2323f64641dfac695f75
+  data.tar.gz: 49427d9325a27034c1969d71875dcd4eacff4f5c3ac9625ccc6cdb554c4c2df4
 SHA512:
-  metadata.gz: 7cd290b2f9b5d9188c9db0bd07a9f88ce564faac927c3590af5c88cf11be0a848dd209fd7c0dd48e5af103bc8f05822911e7ffbe94c9a35dab0732e9e7d70a21
-  data.tar.gz: 0bcd9304b78f9ddb362ef7809831fe5065454adc64aa117df302d8a0ec384d92d7aac64e2dc1e8ebc3caf12267d78dbf7150f9fe12254a2b1d31509644e1e62e
+  metadata.gz: b01f4209b09d6ec09917096159b98be71b31fd0952524a12c7310ba29f88ee8c888bc2927015ef2b1fb529cfe46964c08c743c818807518dee6bfa3cc32f6767
+  data.tar.gz: 564d14e3afaa17f00ac7b034917c7b28612b54b6f71126a36de4d6d43def7b8c19814475ee5b13651bbea7e7eed752b873f918af10baf169bd153d66b8d05c7d

data/Gemfile.lock

@@ -1,64 +1,88 @@
 PATH
   remote: .
   specs:
-    curation (1.7)
-      metainspector
+    curation (1.10)
+      htmlentities
+      metainspector (~> 5.12)
       nokogiri
 
 GEM
   remote: https://rubygems.org/
   specs:
-    addressable (2.7.0)
+    addressable (2.8.0)
       public_suffix (>= 2.0.2, < 5.0)
     ansi (1.5.0)
     builder (3.2.4)
     byebug (11.1.3)
     domain_name (0.5.20190701)
       unf (>= 0.0.5, < 1.0.0)
-    faraday (1.0.1)
-      multipart-post (>= 1.2, < 3)
-    faraday-cookie_jar (0.0.6)
-      faraday (>= 0.7.4)
+    faraday (1.10.0)
+      faraday-em_http (~> 1.0)
+      faraday-em_synchrony (~> 1.0)
+      faraday-excon (~> 1.1)
+      faraday-httpclient (~> 1.0)
+      faraday-multipart (~> 1.0)
+      faraday-net_http (~> 1.0)
+      faraday-net_http_persistent (~> 1.0)
+      faraday-patron (~> 1.0)
+      faraday-rack (~> 1.0)
+      faraday-retry (~> 1.0)
+      ruby2_keywords (>= 0.0.4)
+    faraday-cookie_jar (0.0.7)
+      faraday (>= 0.8.0)
       http-cookie (~> 1.0.0)
+    faraday-em_http (1.0.0)
+    faraday-em_synchrony (1.0.0)
     faraday-encoding (0.0.5)
       faraday
-    faraday-http-cache (2.2.0)
+    faraday-excon (1.1.0)
+    faraday-http-cache (2.4.0)
       faraday (>= 0.8)
-    faraday_middleware (1.0.0)
+    faraday-httpclient (1.0.1)
+    faraday-multipart (1.0.4)
+      multipart-post (~> 2)
+    faraday-net_http (1.0.1)
+    faraday-net_http_persistent (1.2.0)
+    faraday-patron (1.0.0)
+    faraday-rack (1.0.0)
+    faraday-retry (1.0.3)
+    faraday_middleware (1.2.0)
       faraday (~> 1.0)
-    fastimage (2.1.7)
-    http-cookie (1.0.3)
+    fastimage (2.2.6)
+    htmlentities (4.3.4)
+    http-cookie (1.0.5)
       domain_name (~> 0.5)
-    metainspector (5.10.1)
-      addressable (~> 2.7.0)
-      faraday (~> 1.0.0)
-      faraday-cookie_jar (~> 0.0.6)
-      faraday-encoding (~> 0.0.5)
-      faraday-http-cache (~> 2.2.0)
-      faraday_middleware (~> 1.0.0)
-      fastimage (~> 2.1.7)
-      nesty (~> 1.0.2)
-      nokogiri (~> 1.10.9)
-    mini_portile2 (2.4.0)
-    minitest (5.14.1)
-    minitest-reporters (1.4.2)
+    metainspector (5.12.1)
+      addressable (~> 2.7)
+      faraday (>= 1.4, < 3.0)
+      faraday-cookie_jar (~> 0.0)
+      faraday-encoding (~> 0.0)
+      faraday-http-cache (~> 2.2)
+      faraday_middleware (~> 1.0)
+      fastimage (~> 2.2)
+      nesty (~> 1.0)
+      nokogiri (~> 1.11)
+    minitest (5.15.0)
+    minitest-reporters (1.5.0)
       ansi
       builder
       minitest (>= 5.0)
       ruby-progressbar
-    multipart-post (2.1.1)
+    multipart-post (2.2.0)
     nesty (1.0.2)
-    nokogiri (1.10.10)
-      mini_portile2 (~> 2.4.0)
-    public_suffix (4.0.5)
+    nokogiri (1.13.6-x86_64-darwin)
+      racc (~> 1.4)
+    public_suffix (4.0.7)
+    racc (1.6.0)
     rake (12.3.3)
-    ruby-progressbar (1.10.1)
+    ruby-progressbar (1.11.0)
+    ruby2_keywords (0.0.5)
     unf (0.1.4)
       unf_ext
-    unf_ext (0.0.7.7)
+    unf_ext (0.0.8.2)
 
 PLATFORMS
-  ruby
+  x86_64-darwin-21
 
 DEPENDENCIES
   byebug
@@ -68,4 +92,4 @@ DEPENDENCIES
   rake (~> 12.0)
 
 BUNDLED WITH
-   2.1.4
+   2.3.12

data/curation.gemspec

@@ -21,6 +21,7 @@ Gem::Specification.new do |spec|
   spec.bindir        = "exe"
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ["lib"]
-  spec.add_dependency "metainspector"
+  spec.add_dependency "metainspector", '~> 5.12'
   spec.add_dependency "nokogiri"
+  spec.add_dependency "htmlentities"
 end

data/lib/curation/version.rb

@@ -1,3 +1,3 @@
 module Curation
-  VERSION = "1.7"
+  VERSION = "1.10"
 end

data/lib/curation.rb

@@ -1,6 +1,7 @@
 require "curation/version"
 require "metainspector"
 require "open-uri"
+require "htmlentities"
 
 module Curation
   class Error < StandardError; end
@@ -20,7 +21,7 @@ module Curation
     ]
 
     def initialize(url, html = nil)
-      @url = url
+      @url = url.to_s.gsub('http://', 'https://')
       @html = html
     end
 
@@ -37,6 +38,7 @@ module Curation
     end
 
     def text
+      # require 'byebug'; byebug
       @text ||= find_text
     end
 
@@ -49,6 +51,8 @@ module Curation
     def find_title
       if json_ld.any?
         json_ld.each do |ld|
+          # require 'byebug'; byebug
+          ld = ld.first if ld.is_a?(Array)
           return ld['headline'] if ld.has_key? 'headline'
         end
       end
@@ -70,6 +74,7 @@ module Curation
     def find_image
       if json_ld.any?
         json_ld.each do |ld|
+          ld = ld.first if ld.is_a?(Array)
           if ld.has_key? 'image'
             image_data = ld['image']
             return image_data if image_data.is_a? String
@@ -96,6 +101,14 @@ module Curation
     end
 
     def find_text
+      text = find_text_with_json_ld || find_text_with_nokogiri
+      text.to_s.gsub!('<br><br>', '<br>')
+      # require 'byebug'; byebug
+      text = clean_encoding text
+      text
+    end
+
+    def find_text_with_json_ld
       if json_ld.any?
         json_ld.each do |ld|
           next unless ['NewsArticle', 'ReportageNewsArticle'].include? ld['@type']
@@ -103,6 +116,10 @@ module Curation
           return ld['articleBody'] if ld.has_key? 'articleBody'
         end
       end
+      nil
+    end
+
+    def find_text_with_nokogiri
       h = nokogiri.dup
       BLACKLIST.each do |tag|
         h.css(tag).remove
@@ -110,7 +127,6 @@ module Curation
       nodes = h.css('p')
       nodes.xpath('//style').remove
       text = nodes.to_html
-      text.gsub!('<br><br>', '<br>')
       text
     end
 
@@ -125,6 +141,7 @@ module Curation
       return Date.parse metatags['pubdate'] rescue nil
       return Date.parse nokogiri.css('meta[property="article:published"]').first['content'] rescue nil
       return Date.parse nokogiri.css('meta[property="article:published_time"]').first['content'] rescue nil
+      return Date.parse nokogiri.css('meta[property="og:article:published_time"]').first['content'] rescue nil
       chunks = html.split('DisplayDate')
       if chunks.count > 1
         value = chunks[1]
@@ -156,12 +173,11 @@ module Curation
         begin
           options = nokogiri.css('[type="application/ld+json"]')
           options.each do |option|
-            string = option.inner_text
-            hash = JSON.parse(string)
-            @json_ld << hash
+            @json_ld << json_ld_from_object(option)
           end
           # Some sites have tables in tables
           @json_ld.flatten!
+          # require 'byebug'; byebug
         rescue
           puts 'Curation::Page json_ld error'
         end
@@ -169,6 +185,12 @@ module Curation
       @json_ld
     end
 
+    def json_ld_from_object(object)
+      JSON.parse object.inner_text
+    rescue
+      {}
+    end
+
     def file
       @file ||= URI.open url, 'User-Agent' => "Mozilla/5.0"
     rescue
@@ -188,9 +210,13 @@ module Curation
 
     def nokogiri
       unless @nokogiri
-        file.rewind
-        @nokogiri = Nokogiri::HTML file
-        file.rewind
+        if file.nil?
+          @nokogiri = metainspector.parsed
+        else
+          file.rewind
+          @nokogiri = Nokogiri::HTML file
+          file.rewind
+        end
       end
       @nokogiri
     rescue
@@ -212,5 +238,26 @@ module Curation
     rescue
       puts 'Curation::Page metatags error'
     end
+
+    # r&Atilde;&copy;forme -> réforme
+    def clean_encoding(text)
+      clean_text = HTMLEntities.new.decode text
+      double_encoding = false
+      [
+        'é', # é
+        'è', # è
+        'î', # î
+        'ê', # ê
+      ].each do |string|
+        # require 'byebug'; byebug
+        double_encoding = true if clean_text.include? string
+      end
+      if double_encoding
+        clean_text.encode('iso-8859-1', undef: :replace)
+                  .force_encoding('utf-8')
+      else
+        text
+      end
+    end
   end
 end

metadata

@@ -1,17 +1,31 @@
 --- !ruby/object:Gem::Specification
 name: curation
 version: !ruby/object:Gem::Version
-  version: '1.7'
+  version: '1.10'
 platform: ruby
 authors:
 - Arnaud Levy
-autorequire: 
+autorequire:
 bindir: exe
 cert_chain: []
-date: 2020-07-25 00:00:00.000000000 Z
+date: 2022-06-08 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: metainspector
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.12'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.12'
+- !ruby/object:Gem::Dependency
+  name: nokogiri
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -25,7 +39,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: nokogiri
+  name: htmlentities
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -65,7 +79,7 @@ licenses:
 metadata:
   homepage_uri: https://github.com/arnaudlevy/curation
   source_code_uri: https://github.com/arnaudlevy/curation
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -80,8 +94,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.0.3
-signing_key: 
+rubygems_version: 3.1.6
+signing_key:
 specification_version: 4
 summary: Curation of content
 test_files: []