jekyll-uj-powertools 1.2.7 → 1.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e1eb2312be53f2c304f73706fa54b5c46ad7dc7d3f5ec430da54c41068923501
-  data.tar.gz: 3f5529e3c48c5dd86ec556927e5eaa8eabf80d496437df3ca2f1416bfff0e841
+  metadata.gz: 2db2b27f78f8588283c589c8d6c06e342ba6a11a1b1b55669ea81b722b5fe64b
+  data.tar.gz: 47b62725a184ea99d3f25daec3620e57539dbd8fb42159bdd81bb1f68176e5f9
 SHA512:
-  metadata.gz: 42c17aed83e227c2ace9e65c73079ca439e49cae48183a5bdef771e556c043f0cf67f166af174a0505f174f6a36f89aed4bf186b9741f23dad3fe03fbc599bf1
-  data.tar.gz: d60994e315edf47d64e04e4d5718380293f0aa9f6719b9c3e714f034f451c2b277f3e06f5e6207d291302e70ceb200aba5bc9029e45c8967bda23b152a46bd76
+  metadata.gz: 81e5f82be027417b008c9a6d3cbc2bb86bc5f992a60447fb42dc94011da6abb4cb5374756bca80141c36cad263e92e16c2a7ee18a94c4c5122173e9c1464a5ae
+  data.tar.gz: 6a3ea9d448ba3a543cf70623e6eb67ab72dc25eecbd761d9959e66cf6a9db92342271307713a95ea6bb268b761e2f23e0565d107c3c4fa8f6528543b3118f036

data/Gemfile

@@ -5,6 +5,7 @@ source "https://rubygems.org"
 # Specify your gem's dependencies in jekyll-uj-powertools.gemspec
 gemspec
 
+# Require Jekyll version if specified in the environment variable
 if ENV["JEKYLL_VERSION"]
   gem "jekyll", "~> #{ENV["JEKYLL_VERSION"]}"
 end

data/README.md

@@ -69,6 +69,13 @@ Convert a string to title case.
 {{ "hello world" | uj_title_case }}
 ```
 
+### `uj.cache_breaker` Variable
+Use the `uj.cache_breaker` variable to append a cache-busting query parameter to your assets.
+
+```liquid
+<link rel="stylesheet" href="{{ "/assets/css/style.css" | prepend: site.baseurl }}?v={{ uj.cache_breaker }}">
+```
+
 These examples show how you can use the features of `jekyll-uj-powertools` in your Jekyll site.
 
 ## 🔧 Development

data/jekyll-uj-powertools.gemspec

@@ -32,6 +32,9 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "rake"
   spec.add_development_dependency "rspec"
 
+  # Translation and HTML manipulation requires Nokogiri
+  spec.add_runtime_dependency 'nokogiri', '>= 1.17'
+
   # Ruby version
   spec.required_ruby_version = ">= 2.0.0"
 end

data/lib/generators/inject-data.rb

@@ -0,0 +1,41 @@
+module Jekyll
+  class InjectData < Generator
+    safe true
+    priority :low
+
+    def generate(site)
+      # Define a global variable accessible in templates
+      # site.config['cache_breaker'] = Time.now.to_i.to_s
+
+      # Process pages
+      site.pages.each do |page|
+        inject_data(page, site)
+      end
+
+      # Process documents in all collections
+      site.collections.each do |_, collection|
+        collection.docs.each do |document|
+          inject_data(document, site)
+        end
+      end
+    end
+
+    private
+
+    def inject_data(item, site)
+      # Inject a random number into the item's data
+      item.data['random_id'] = rand(100) # Random number between 0 and 99
+
+      return unless item.data['layout'] # Skip items without layouts
+
+      # Find the layout file by its name
+      layout_name = item.data['layout']
+      layout = site.layouts[layout_name]
+
+      if layout && layout.data
+        # Merge layout front matter into item's "layout_data"
+        item.data['layout_data'] = layout.data
+      end
+    end
+  end
+end

data/lib/generators/translate-pages.rb

@@ -0,0 +1,209 @@
+require 'json'
+require 'net/http'
+require 'fileutils'
+require 'nokogiri'
+require 'digest'
+
+module Jekyll
+  class TranslatePages < Generator
+    safe true
+    priority :low
+
+    # Variables
+    # Translation path
+    CACHE_DIR = '.temp/translations'
+    # Re-translate pages older than this many days
+    RECHECK_DAYS = 30
+
+    def generate(site)
+      target_langs = site.config.dig('translation', 'languages') || []
+
+      # Log
+      puts "🔁 Starting translation process for supported languages: #{target_langs.join(', ')}"
+      puts "📂 Cache directory: #{CACHE_DIR}"
+      # puts "🔍 All environment variables:"
+      # ENV.each { |k, v| puts "   #{k}=#{v}" }
+      puts "🔍 UJ_ environment variables:"
+      ENV.select { |k, _| k.start_with?('UJ_') }.each { |k, v| puts "   #{k}=#{v}" }
+
+      # Skip if site config translation is disabled
+      unless site.config.dig('translation', 'enabled')
+        puts "🚫 Translation is disabled in _config.yml (translation.enabled: false)"
+        return
+      end
+
+      # Ensure OpenAI API key is set
+      unless ENV['OPENAI_API_KEY'] && !ENV['OPENAI_API_KEY'].strip.empty?
+        puts "❌ OPENAI_API_KEY not found in environment. Exiting translation process."
+        return
+      end
+
+      # Quit if no languages are configured
+      if target_langs.empty?
+        puts "🚫 No target languages configured in _config.yml (translation.languages). Exiting translation process."
+        return
+      end
+
+      # Keep track of skipped files
+      skipped_files = []
+
+      # Loop through all pages in the site
+      site.pages.clone.each do |page|
+        next unless page.output_ext == '.html'
+        original_content = page.content
+        original_hash = Digest::SHA256.hexdigest(original_content)
+        page_path = page.path.sub(/^_?site\//, '')
+
+        # Skip if page.translation.enabled is false
+        if page.data['translation'] && page.data['translation']['enabled'] == false
+          skipped_files << "#{page_path} (translation disabled)"
+          next
+        end
+
+        # Skip if page.redirect.url is set
+        if page.data['redirect'] && page.data['redirect']['url']
+          skipped_files << "#{page_path} (redirect set)"
+          next
+        end
+
+        target_langs.each do |lang|
+          translated_path = File.join(CACHE_DIR, lang, page_path)
+          meta_path = "#{translated_path}.meta.json"
+
+          # @TODO: Remove this
+          # Unless its pages/legal/terms.md, QUIT
+          if page_path != 'pages/legal/terms.md'
+            skipped_files << "#{page_path} (only 'pages/legal/terms.md' is processed)"
+            next
+          end
+
+          # Log
+          puts "🌐 Processing page '#{page_path}' for language '#{lang}'"
+
+          # Either read cached translation or generate a new one
+          translated = read_or_translate(original_content, original_hash, lang, translated_path, meta_path)
+
+          next unless translated # skip this lang if translation failed
+
+          # Build new page with translated content
+          new_page = page.dup
+          new_page.data = page.data.dup
+          new_page.data['lang'] = lang
+          new_page.data['permalink'] = "/#{lang}#{page.url}"
+          new_page.content = rewrite_links(translated, lang)
+
+          site.pages << new_page
+          puts "✅ Added translated page: /#{lang}#{page.url}"
+        end
+      end
+
+      # Log skipped files at the end
+      if skipped_files.any?
+        puts "\n🚫 Skipped files:"
+        skipped_files.each { |f| puts " - #{f}" }
+      end
+
+      puts "🎉 Translation process complete."
+    end
+
+    private
+
+    # Return cached translation or generate new one via API
+    def read_or_translate(content, hash, lang, path, meta_path)
+      if File.exist?(path) && File.exist?(meta_path)
+        meta = JSON.parse(File.read(meta_path)) rescue {}
+        last_hash = meta['hash']
+        last_time = Time.at(meta['timestamp'].to_i) rescue Time.at(0)
+
+        age_days = ((Time.now - last_time) / (60 * 60 * 24)).round
+        puts "📅 Cache age: #{age_days}/#{RECHECK_DAYS} days"
+
+        # Determine whether we should re-check based on age
+        recheck_due_to_age = RECHECK_DAYS && RECHECK_DAYS > 0 && age_days >= RECHECK_DAYS
+
+        # Re-translate if hash changed or translation is too old (only if RECHECK_DAYS is truthy)
+        if last_hash == hash && !recheck_due_to_age
+          puts "📦 Using cached translation at: #{path}"
+          return File.read(path)
+        else
+          puts "🔁 Cache stale or hash changed, regenerating translation..."
+        end
+      else
+        puts "📭 No cache found, generating translation..."
+      end
+
+      begin
+        result = translate_with_api(content, lang)
+      rescue => e
+        puts "❌ Skipping translation for '#{lang}' due to error: #{e.message}"
+        return nil
+      end
+
+      FileUtils.mkdir_p(File.dirname(path))
+      File.write(path, result)
+      File.write(meta_path, {
+        timestamp: Time.now.to_i,
+        hash: hash
+      }.to_json)
+
+      puts "📝 Cached translation and metadata written to: #{path}"
+
+      # Log before/after content
+      puts "\n--- BEFORE CONTENT (#{lang}) ---\n#{content[0..500]}..."
+      puts "\n--- AFTER TRANSLATION (#{lang}) ---\n#{result[0..500]}..."
+
+      result
+    end
+
+    # Perform translation via OpenAI API
+    def translate_with_api(content, lang)
+      system_prompt = "You are a professional translator. Translate the provided HTML content, preserving all original formatting, HTML structure, metadata, and links. Do not explain anything — just return the translated HTML. Translate to #{lang}."
+      user_message = content
+
+      uri = URI('https://api.openai.com/v1/chat/completions')
+
+      res = Net::HTTP.post(
+        uri,
+        {
+          model: 'gpt-4',
+          messages: [
+            { role: 'system', content: system_prompt },
+            { role: 'user', content: user_message }
+          ],
+          temperature: 0.3
+        }.to_json,
+        {
+          'Authorization' => "Bearer #{ENV['OPENAI_API_KEY']}",
+          'Content-Type' => 'application/json'
+        }
+      )
+
+      if res.code.to_i != 200
+        raise "HTTP #{res.code}: #{res.body}"
+      end
+
+      json = JSON.parse(res.body)
+      result = json.dig('choices', 0, 'message', 'content')
+
+      if result.nil? || result.strip.empty?
+        raise "Translation returned empty or invalid content"
+      end
+
+      puts "🔤 Translation complete."
+      result
+    end
+
+    # Rewrite internal links to language-prefixed versions
+    def rewrite_links(html, lang)
+      doc = Nokogiri::HTML::DocumentFragment.parse(html)
+      doc.css('a[href^="/"]').each do |a|
+        href = a['href']
+        next if href.start_with?("/#{lang}") || href.include?('.') || href.start_with?('//')
+        new_href = "/#{lang}#{href}"
+        puts "🔗 Rewriting link: #{href} -> #{new_href}"
+        a['href'] = new_href
+      end
+      doc.to_html
+    end
+  end
+end

data/lib/jekyll-uj-powertools/version.rb

@@ -1,5 +1,5 @@
 module Jekyll
   module UJPowertools
-    VERSION = "1.2.7"
+    VERSION = "1.3.0"
   end
 end

data/lib/jekyll-uj-powertools.rb

@@ -62,57 +62,9 @@ module Jekyll
     end
   end
 
-  # Set a global cache buster timestamp
-  # class CacheBreakerGenerator < Jekyll::Generator
-  #   safe true
-  #   priority :highest
-
-  #   def generate(site)
-  #     # Define a global variable accessible in templates
-  #     site.config['cache_breaker2'] = Time.now.to_i.to_s
-  #   end
-  # end
-
-  # Inject data into pages and documents
-  class InjectData < Generator
-    safe true
-    priority :low
-
-    def generate(site)
-      # Define a global variable accessible in templates
-      # site.config['cache_breaker'] = Time.now.to_i.to_s
-
-      # Process pages
-      site.pages.each do |page|
-        inject_data(page, site)
-      end
-
-      # Process documents in all collections
-      site.collections.each do |_, collection|
-        collection.docs.each do |document|
-          inject_data(document, site)
-        end
-      end
-    end
-
-    private
-
-    def inject_data(item, site)
-      # Inject a random number into the item's data
-      item.data['random_id'] = rand(100) # Random number between 0 and 99
-
-      return unless item.data['layout'] # Skip items without layouts
-
-      # Find the layout file by its name
-      layout_name = item.data['layout']
-      layout = site.layouts[layout_name]
-
-      if layout && layout.data
-        # Merge layout front matter into item's "layout_data"
-        item.data['layout_data'] = layout.data
-      end
-    end
-  end
+  # Load Generators
+  require_relative "generators/inject-data"
+  require_relative "generators/translate-pages"
 end
 
 # Register the filter
@@ -121,6 +73,5 @@ Liquid::Template.register_filter(Jekyll::UJPowertools)
 # Register hook
 Jekyll::Hooks.register :site, :pre_render do |site|
   site.config['uj'] ||= {}
-  site.config['uj']['cacheBreaker'] = Jekyll::UJPowertools.cache_timestamp
   site.config['uj']['cache_breaker'] = Jekyll::UJPowertools.cache_timestamp
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: jekyll-uj-powertools
 version: !ruby/object:Gem::Version
-  version: 1.2.7
+  version: 1.3.0
 platform: ruby
 authors:
 - ITW Creative Works
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-05-01 00:00:00.000000000 Z
+date: 2025-06-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: jekyll
@@ -72,6 +72,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: nokogiri
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.17'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.17'
 description: jekyll-uj-powertools provides a powerful set of utilities for Jekyll,
   including functions to remove ads from strings and escape JSON characters.
 email:
@@ -85,6 +99,8 @@ files:
 - README.md
 - Rakefile
 - jekyll-uj-powertools.gemspec
+- lib/generators/inject-data.rb
+- lib/generators/translate-pages.rb
 - lib/jekyll-uj-powertools.rb
 - lib/jekyll-uj-powertools/version.rb
 - spec/jekyll-uj-powertools_spec.rb