bridgetown-related-posts 0.1.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: 4e102bbfd393fdfb004ea44b684e56a9ae67528047c3a6eaf28fc2c22d08a512
+  data.tar.gz: 10ba94a0dc8e0c8815590e55cf574dff048d78f16ee1892022ab161e277bfc59
+SHA512:
+  metadata.gz: 58f762fa5b087618e12ab8682f79b916913a4b974ad7c7ee1b220f76d210c869aad5eb6e63b604cf591ea9f037d151344a31b2cd15abde7132836270284243a5
+  data.tar.gz: fd98995adb563f00740ae5c50441993c02a840ddc675992f7ed051266267197e8ad17bfc1038e824f84d49b6c326e36ae55e56b2e59ca327aa838a53310336c6

data/lib/bridgetown-related-posts/builder.rb

@@ -0,0 +1,137 @@
+# lib/bridgetown-related-posts/builder.rb
+module BridgetownRelatedPosts
+  class Builder < Bridgetown::Builder
+
+    def self.init
+      config.related_posts_limit ||= 5 # Default value
+    end
+   
+    def build
+      hook :site, :post_read do
+        BridgetownRelatedPosts.set_config(config)
+        posts_data = prepare_posts_data
+        if posts_data.empty?
+          Bridgetown.logger.warn "No posts found for RelatedPostsBuilder"
+        else
+          similarity_matrix = calculate_similarity(posts_data)
+          add_related_posts(posts_data, similarity_matrix)
+        end
+      end
+    end
+  
+    def prepare_posts_data
+      posts = site.collections["posts"].resources
+      Bridgetown.logger.info "Number of posts found: #{posts.size}"
+    
+      posts.map do |post|
+        begin
+          content = post.content
+          
+          {
+            relative_path: post.relative_path,
+            title: post.data["title"] || "Untitled",
+            content: content || "",
+          }
+        rescue => e
+          Bridgetown.logger.error "Error processing post #{post.relative_path}: #{e.message}"
+          Bridgetown.logger.error "Post data: #{post.data.inspect}"
+          nil
+        end
+      end.compact
+    end
+  
+    def calculate_similarity(posts_data)
+      idf = compute_idf(posts_data)
+      tfidf_vectors = posts_data.map { |post| compute_tfidf(post[:content], idf) }
+      
+      similarity_matrix = {}
+      posts_data.each_with_index do |post, i|
+        similarity_matrix[post[:relative_path]] = {}
+        posts_data.each_with_index do |other_post, j|
+          next if i == j
+          similarity = cosine_similarity(tfidf_vectors[i], tfidf_vectors[j])
+          similarity_matrix[post[:relative_path]][other_post[:relative_path]] = similarity
+        end
+      end
+      similarity_matrix
+    end
+  
+    def compute_idf(posts_data)
+      document_frequency = Hash.new(0)
+      posts_data.each do |post|
+        words = post[:content].downcase.split.uniq
+        words.each { |word| document_frequency[word] += 1 }
+      end
+  
+      idf = {}
+      total_docs = posts_data.size
+      document_frequency.each do |word, freq|
+        idf[word] = Math.log(total_docs.to_f / (freq + 1))
+      end
+      idf
+    end
+  
+    def compute_tfidf(text, idf)
+      words = text.downcase.split
+      word_count = words.size
+      tf = Hash.new(0)
+      words.each { |word| tf[word] += 1 }
+      
+      tfidf = {}
+      tf.each do |word, freq|
+        tfidf[word] = (freq.to_f / word_count) * (idf[word] || 0)
+      end
+      tfidf
+    end
+  
+    def cosine_similarity(vec1, vec2)
+      dot_product = 0
+      vec1.each_key do |key|
+        dot_product += vec1[key] * vec2[key] if vec2.has_key?(key)
+      end
+  
+      mag1 = Math.sqrt(vec1.values.map { |x| x ** 2 }.sum)
+      mag2 = Math.sqrt(vec2.values.map { |x| x ** 2 }.sum)
+  
+      dot_product / (mag1 * mag2)
+    end
+  
+    def add_related_posts(posts_data, similarity_matrix)
+      limit = config.related_posts&.limit || 5
+      
+      posts_data.each do |post|
+        similar_posts = similarity_matrix[post[:relative_path]]
+        if similar_posts.nil?
+          puts "No similar posts found for: #{post[:relative_path]}"
+          next
+        end
+
+        similar_posts = similar_posts
+          .sort_by { |_, score| -score }
+          .first(limit)
+          .map { |relative_path, _| site.collections["posts"].resources.find { |p| p.relative_path == relative_path } }
+          .compact
+    
+        # Find the actual post object
+        actual_post = site.collections["posts"].resources.find { |p| p.relative_path == post[:relative_path] }
+    
+        if actual_post
+          # Add related posts to the post's data
+          actual_post.data["related_posts"] = similar_posts
+    
+          Bridgetown.logger.info "Related posts for #{actual_post.data["title"]}:"
+    
+          similar_posts.each do |related_post|
+            if related_post
+              Bridgetown.logger.info "  - #{related_post.data["title"]}"
+            else
+              Bridgetown.logger.warning "  - Related post not found"
+            end
+          end
+        else
+          Bridgetown.logger.error "Post not found: #{post[:relative_path]}"
+        end
+      end
+    end
+  end
+end

data/lib/bridgetown-related-posts/config.rb

@@ -0,0 +1,12 @@
+require "bridgetown"
+require "bridgetown-related-posts/builder"
+
+Bridgetown::PluginManager.new_source_manifest(
+  origin: BridgetownRelatedPosts,
+  components: {
+    "builders" => [{
+      name: "RelatedPosts",
+      class: BridgetownRelatedPosts::Builder
+    }]
+  }
+)

data/lib/bridgetown-related-posts.rb

@@ -0,0 +1,16 @@
+require "bridgetown"
+require_relative "bridgetown-related-posts/builder"
+
+Bridgetown.initializer :"bridgetown-related-posts" do |config|
+  config.builder BridgetownRelatedPosts::Builder
+end
+
+module BridgetownRelatedPosts
+  def self.config
+    @config ||= Bridgetown::Configuration.new
+  end
+
+  def self.set_config(config)
+    @config = config
+  end
+end

metadata

@@ -0,0 +1,67 @@
+--- !ruby/object:Gem::Specification
+name: bridgetown-related-posts
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- Matthew Clarkson
+autorequire:
+bindir: bin
+cert_chain: []
+date: 2024-07-25 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: bridgetown
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.3.0
+    - - "<"
+      - !ruby/object:Gem::Version
+        version: '2.0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.3.0
+    - - "<"
+      - !ruby/object:Gem::Version
+        version: '2.0'
+description: This plugin calculates and adds related posts to your Bridgetown site
+  using TF-IDF and cosine similarity.
+email: mpclarkson@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/bridgetown-related-posts.rb
+- lib/bridgetown-related-posts/builder.rb
+- lib/bridgetown-related-posts/config.rb
+homepage: https://github.com/mpclarkson/bridgetown-related-posts/
+licenses:
+- MIT
+metadata: {}
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: 3.3.0
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.5.10
+signing_key:
+specification_version: 4
+summary: A Bridgetown plugin for automatically generating related posts based on cosine
+  similarity.
+test_files: []