Class: BridgetownRelatedPosts::Builder

Inherits:
Bridgetown::Builder
  • Object
show all
Defined in:
lib/bridgetown-related-posts/builder.rb

Class Method Summary collapse

Instance Method Summary collapse

Class Method Details

.init ⇒ Object 



5
6
7
 
# File 'lib/bridgetown-related-posts/builder.rb', line 5

def self.init
  config.related_posts_limit ||= 5 # Default value
end

Instance Method Details



99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
 
# File 'lib/bridgetown-related-posts/builder.rb', line 99

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

#build ⇒ Object 



9
10
11
12
13
14
15
16
17
18
19
20
 
# File 'lib/bridgetown-related-posts/builder.rb', line 9

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

#calculate_similarity(posts_data) ⇒ Object 



43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
 
# File 'lib/bridgetown-related-posts/builder.rb', line 43

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

#compute_idf(posts_data) ⇒ Object 



59
60
61
62
63
64
65
66
67
68
69
70
71
72
 
# File 'lib/bridgetown-related-posts/builder.rb', line 59

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

#compute_tfidf(text, idf) ⇒ Object 



74
75
76
77
78
79
80
81
82
83
84
85
 
# File 'lib/bridgetown-related-posts/builder.rb', line 74

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

#cosine_similarity(vec1, vec2) ⇒ Object 



87
88
89
90
91
92
93
94
95
96
97
 
# File 'lib/bridgetown-related-posts/builder.rb', line 87

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

#prepare_posts_data ⇒ Object 



22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
 
# File 'lib/bridgetown-related-posts/builder.rb', line 22

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