curation 2.0.1 → 2.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: c2062c7ec7fb444d27f102d26658b386a01a964ea4e03aa0d81a472316012d11
-  data.tar.gz: 634f5216e61801b3ac42c8b340d5012c101024f4d5b5e0ba155c01a89f38cafb
+  metadata.gz: 7385bf2f2c34ab16df36a865f82bb53effca4f13824122d788f22afab6d5f956
+  data.tar.gz: 63d681cb1c06c5aa34caa56deae021ae28292306df00bf85708341ad9006568a
 SHA512:
-  metadata.gz: 15008b92c6a51fdf9bd79b9f1da01d59323c13589b6e8ccd5d8754b96153173fe8a2cdffa5faa45a4d38c92b097a35012f8dff98fea6244ade8c9af22c13d1cc
-  data.tar.gz: 96a29d3c8482fce0101f91a3f24eafaec51a9115f2169831b67939edf79a47d7f5bb0c3223466de355cf8a367143a262cc54a1d5e4dd7e6bffe2d310879a9cae
+  metadata.gz: 222924c992b61f62d8347d76a2f167773a83c7b2ac613031a3dc40d17a0fbbe1f1ec8485d4caa127f0bc89133503e4489b989ef1307676f0ef125d555f4b0e6f
+  data.tar.gz: 4668058ec845b325b46dac9a22c7ce6c3047f9d536655f5297d01b63a3bc087fca5dcb6494fd6d796720359e41eb61b572e2117878d4c730d6090dabe171332a

data/Gemfile

@@ -7,3 +7,4 @@ gem 'rake', '~> 12.0'
 gem 'minitest'
 gem 'minitest-reporters'
 gem 'byebug'
+

data/Gemfile.lock

@@ -1,10 +1,11 @@
 PATH
   remote: .
   specs:
-    curation (2.0.1)
+    curation (2.0.3)
       htmlentities
       metainspector
       nokogiri
+      rails-html-sanitizer
 
 GEM
   remote: https://rubygems.org/
@@ -15,6 +16,7 @@ GEM
     base64 (0.2.0)
     builder (3.2.4)
     byebug (11.1.3)
+    crass (1.0.6)
     domain_name (0.5.20190701)
       unf (>= 0.0.5, < 1.0.0)
     faraday (2.7.11)
@@ -40,6 +42,9 @@ GEM
     htmlentities (4.3.4)
     http-cookie (1.0.5)
       domain_name (~> 0.5)
+    loofah (2.21.4)
+      crass (~> 1.0.2)
+      nokogiri (>= 1.12.0)
     metainspector (5.15.0)
       addressable (~> 2.8.4)
       faraday (~> 2.5)
@@ -52,6 +57,7 @@ GEM
       fastimage (~> 2.2)
       nesty (~> 1.0)
       nokogiri (~> 1.13)
+    mini_portile2 (2.8.9)
     minitest (5.20.0)
     minitest-reporters (1.6.1)
       ansi
@@ -59,10 +65,14 @@ GEM
       minitest (>= 5.0)
       ruby-progressbar
     nesty (1.0.2)
-    nokogiri (1.15.4-x86_64-darwin)
+    nokogiri (1.15.4)
+      mini_portile2 (~> 2.8.2)
       racc (~> 1.4)
     public_suffix (5.0.3)
     racc (1.7.3)
+    rails-html-sanitizer (1.6.0)
+      loofah (~> 2.21)
+      nokogiri (~> 1.14)
     rake (12.3.3)
     ruby-progressbar (1.13.0)
     ruby2_keywords (0.0.5)
@@ -72,6 +82,7 @@ GEM
     zlib (2.1.1)
 
 PLATFORMS
+  arm64-darwin-24
   x86_64-darwin-21
   x86_64-darwin-22
 

data/curation.gemspec

@@ -24,4 +24,5 @@ Gem::Specification.new do |spec|
   spec.add_dependency "metainspector"
   spec.add_dependency "nokogiri"
   spec.add_dependency "htmlentities"
+  spec.add_dependency "rails-html-sanitizer"
 end

data/lib/curation/finders/text.rb

@@ -26,8 +26,8 @@ module Text
     text = text.gsub('<br><br>', '<br>')
     text = text.gsub(/\s+/, ' ')
     text = clean_encoding(text)
+    text = Rails::HTML5::FullSanitizer.new.sanitize(text)
     text
-
   end
 
   def find_text

data/lib/curation/finders/title.rb

@@ -9,7 +9,7 @@ module Title
   def find_title
     find_title_with_json_ld ||
     find_title_with_metainspector ||
-    find_title_with_nokogiri ||
+    find_title_with_nokogiri ||
     ''
   end
 
@@ -29,11 +29,11 @@ module Title
     metainspector_title = metainspector.title
     # Problème avec une balise <meta property="title" content="Run 0"  />,
     # metainspector croit que c'est le titre de la page.
-    # Comme le title contient le best title, avec souvent des infos en plus sur le site, 
+    # Comme le title contient le best title, avec souvent des infos en plus sur le site,
     # on vérifie si le best title est bien contenu dans le title
-    if  metainspector_title.present? && 
+    if  metainspector_title.present? &&
         metainspector_title.present? &&
-        metainspector_best_title.present? && 
+        metainspector_best_title.present? &&
         metainspector_title.include?(metainspector_best_title)
       return metainspector_best_title
     elsif metainspector_title.present?
@@ -55,4 +55,4 @@ module Title
     end
   end
 
-end
+end

data/lib/curation/version.rb

@@ -1,3 +1,3 @@
 module Curation
-  VERSION = "2.0.1"
+  VERSION = "2.0.3"
 end

data/lib/curation.rb

@@ -10,6 +10,7 @@ require "curation/finders/title"
 require "metainspector"
 require "open-uri"
 require "htmlentities"
+require "rails-html-sanitizer"
 
 module Curation
   class Error < StandardError; end

metadata

@@ -1,14 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: curation
 version: !ruby/object:Gem::Version
-  version: 2.0.1
+  version: 2.0.3
 platform: ruby
 authors:
 - Arnaud Levy
-autorequire:
 bindir: exe
 cert_chain: []
-date: 2023-11-11 00:00:00.000000000 Z
+date: 1980-01-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: metainspector
@@ -52,6 +51,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: rails-html-sanitizer
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: When you build content curation tools, you need to extract the content
   of pages (title, text, image...). This requires different strategies and some fine
   tuning to work efficiently.
@@ -87,7 +100,6 @@ licenses:
 metadata:
   homepage_uri: https://github.com/noesya/curation
   source_code_uri: https://github.com/noesya/curation
-post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -102,8 +114,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.4.10
-signing_key:
+rubygems_version: 3.6.7
 specification_version: 4
 summary: Curation of content
 test_files: []