Class: Jekyll::L10n::Extractor

Inherits:
Object
  • Object
show all
Defined in:
lib/jekyll-l10n/extraction/extractor.rb

Overview

String Extraction Orchestrator - Finds translatable strings in generated HTML

The Extractor is the main entry point for the string extraction workflow. It scans all generated HTML files after Jekyll’s build, identifies translatable content (text nodes and configurable HTML attributes), and creates or updates GNU Gettext PO files with the extracted strings.

The extraction workflow:

  1. Scans all HTML files in Jekyll output directory (_site/)

  2. For each HTML file, extracts translatable text and attributes

  3. Normalizes text for consistent matching across builds

  4. Creates or updates page-specific PO files in _locales/ directory

  5. Optionally applies automatic translations via LibreTranslate API

Key responsibilities:

  • Load and validate extraction configuration from pages

  • Extract text and attributes from HTML with file location references

  • Create and update PO files with extracted strings

  • Log extraction statistics and progress

  • Coordinate with LibreTranslate for automatic translation

Examples:

Usage (typically invoked by PostWriteProcessor)

extractor = Extractor.new(site)
result = extractor.extract_site
# => { files_processed: 42, strings_extracted: 237, po_files_created: 3 }

See Also:

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(site) ⇒ Extractor

Initialize the string extractor

Sets up configuration and result saving infrastructure for extraction.

Parameters:

  • site (Jekyll::Site) —

    The Jekyll site object



53
54
55
56
57
58
59
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 53

def initialize(site)
  @site = site
  @source = SiteConfigAccessor.source(@site)
  @dest = SiteConfigAccessor.dest(@site)
  @config_loader = ExtractionConfigLoader.new(@site, @dest)
  @result_saver = ExtractionResultSaver.new(@site)
end

Instance Attribute Details

#site ⇒ Object (readonly)

Returns the value of attribute site.



46
47
48
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 46

def site
  @site
end

Instance Method Details

#default_stats ⇒ Object 



122
123
124
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 122

def default_stats
  { files_processed: 0, strings_extracted: 0, po_files_created: 0 }
end

#extract_site ⇒ Hash<Symbol, Integer>

Extract all translatable strings from the generated site

Main entry point for extraction. Scans all HTML files in the build output, extracts translatable strings and attributes, creates/updates PO files, and optionally translates strings via LibreTranslate API.

Examples:

result = extractor.extract_site
puts "Processed #{result[:files_processed]} files"

Returns:

  • (Hash<Symbol, Integer>) —

    Statistics hash with keys:

    • :files_processed - Number of HTML files processed

    • :strings_extracted - Total strings extracted

    • :po_files_created - Number of PO files created/updated



74
75
76
77
78
79
80
81
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 74

def extract_site
  Jekyll.logger.info 'Localization', 'Extracting translatable strings...'
  start_time = Time.now
  stats = process_all_html_files
  translate_all_compendia
  ExtractionLogger.log_summary(stats, Time.now - start_time)
  stats
end

#extract_strings_from_file(file_path, config) ⇒ Object 



126
127
128
129
130
131
132
133
134
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 126

def extract_strings_from_file(file_path, config)
  return [] unless File.exist?(file_path)

  html = FileOperations.read_utf8(file_path)
  exclude_selectors = @config_loader.extract_exclude_selectors(config)

  extractor = HtmlStringExtractor.new(config.translatable_attributes, exclude_selectors)
  extractor.extract(html, @dest, file_path)
end

#find_libretranslate_config ⇒ Object 



136
137
138
139
140
141
142
143
144
145
146
147
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 136

def find_libretranslate_config
  return nil unless @site.respond_to?(:pages)

  @site.pages.each do |page|
    next unless page.data['with_locales'] == true

    config = @config_loader.load_page_config(page.destination(''))
    return config if config.libretranslate_enabled?
  end

  nil
end

#process_all_html_files ⇒ Object 



83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 83

def process_all_html_files
  stats = { files_processed: 0, strings_extracted: 0, po_files_created: 0 }
  html_files = Dir.glob(File.join(@dest, '**', '*.html'))

  html_files.each do |file_path|
    next if @config_loader.skip_localized_page?(file_path)

    file_stats = process_file(file_path)
    stats[:files_processed] += file_stats[:files_processed]
    stats[:strings_extracted] += file_stats[:strings_extracted]
    stats[:po_files_created] += file_stats[:po_files_created]
  end

  stats
end

#process_file(file_path) ⇒ Object 



108
109
110
111
112
113
114
115
116
117
118
119
120
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 108

def process_file(file_path)
  return default_stats unless @config_loader.valid_for_extraction?(file_path)

  config = @config_loader.load_page_config(file_path)
  entries = extract_strings_from_file(file_path, config)
  return default_stats if entries.empty?

  page_path = construct_page_path(file_path)
  @result_saver.save_results(config, entries, page_path)
rescue StandardError => e
  ExtractionLogger.log_error(file_path, e)
  default_stats
end

#translate_all_compendia ⇒ Object 



99
100
101
102
103
104
105
106
 
# File 'lib/jekyll-l10n/extraction/extractor.rb', line 99

def translate_all_compendia
  config = find_libretranslate_config
  return unless config

  ErrorHandler.handle_with_logging('machine translation') do
    @result_saver.translate_compendia(config)
  end
end