jekyll-l10n 1.1.1 → 1.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2a1a64737ee988c61bc5d4c37986c5e2553966bf73fc251ce807976117013236
-  data.tar.gz: b88c9b6283535514aebd9a77e2d5fe4f01407f008e5de2b8d21e0b30cb76bc85
+  metadata.gz: d59b40394ccd291d468bfd045b6b53ffd71204d6b42575f232e7159316c14561
+  data.tar.gz: c3c0bc5a22a1ce203a7623be5217f8477b9d4b57f39918ca221a10f693bbd053
 SHA512:
-  metadata.gz: 332001e7246402a4a4acc41dbc70761f6d26bc506fff76f116d3f971754b617cc39b5496ec69544edd0264777fc17030797eb9be29edf7a60fb751c86d38d1a2
-  data.tar.gz: 69b96d33ceaf4bcbe8e2ab8073e7d6fe549e1cfd5e2b09cce41ea1f32def1df650c8c7347ff8b15c230c7c237b3073ff77b015c8881b291b2f3a6c50ed020fa1
+  metadata.gz: b022fa06243f0d8150be0f53bcbd9861751634fb1d3af4d5cd16ae36bf11cbfb19fd3e86e3af63cad783fea3d21f56338ad4f14cde2fb68165ffe6b9146e12ad
+  data.tar.gz: 74a6fa21f10fe87de0c8f891f9eedb77d36d0df2d50ccd23a794cc20d3a127c183ad6edd35b6284189beb7d0472928ee0bbea8a075fbdfd1abd016586354dbbc

data/lib/jekyll-l10n/constants.rb

@@ -114,8 +114,8 @@ module Jekyll
 
       # Default behavior when LibreTranslate API returns an error
       # If true, translation stops immediately. If false, continues with remaining entries.
-      # @return [Boolean] true
-      DEFAULT_LIBRETRANSLATE_STOP_ON_ERROR = true
+      # @return [Boolean] false
+      DEFAULT_LIBRETRANSLATE_STOP_ON_ERROR = false
 
       # Default interval for logging LibreTranslate translation progress
       # Progress is logged every N entries. Set to 0 to disable.

data/lib/jekyll-l10n/extraction/extractor.rb

@@ -6,6 +6,7 @@ require_relative "logger"
 require_relative "html_string_extractor"
 require_relative "../utils/file_operations"
 require_relative "../utils/site_config_accessor"
+require_relative "../utils/error_handler"
 
 module Jekyll
   module L10n
@@ -99,7 +100,9 @@ module Jekyll
         config = find_libretranslate_config
         return unless config
 
-        @result_saver.translate_compendia(config)
+        ErrorHandler.handle_with_logging("machine translation") do
+          @result_saver.translate_compendia(config)
+        end
       end
 
       def process_file(file_path)

data/lib/jekyll-l10n/po_file/path_builder.rb

@@ -12,7 +12,8 @@ module Jekyll
     # Key responsibilities:
     # * Build paths for compendium PO files
     # * Build paths for page-specific PO files
-    # * Handle file path normalization
+    # * Handle file path normalization and validation
+    # * Prevent path traversal attacks
     #
     # @example
     #   compendium = PoPathBuilder.build('_site', '_locales', 'es', nil)
@@ -25,16 +26,35 @@ module Jekyll
       # For compendium (page_path nil): {source}/{locales_dir}/{locale}.po
       # For page-specific: {source}/{locales_dir}/{locale}/{page_path}.po
       #
+      # Validates that page_path doesn't escape the base directory using path
+      # traversal techniques (e.g., ../../../etc/passwd).
+      #
       # @param source [String] Site source directory
       # @param locales_dir [String] Locales directory name (e.g., '_locales')
       # @param locale [String] Locale code (e.g., 'es', 'fr')
       # @param page_path [String, nil] Page path for page-specific file, nil for compendium
       # @return [String] Full path to PO file
+      # @raise [Jekyll::L10n::Errors::PoFileError] if page_path attempts to escape base directory
       def self.build(source, locales_dir, locale, page_path)
         if page_path.nil?
           File.join(source, locales_dir, "#{locale}.po")
         else
-          File.join(source, locales_dir, locale, "#{page_path}.po")
+          # Build the normal path
+          normal_path = File.join(source, locales_dir, locale, "#{page_path}.po")
+
+          # Validate against path traversal using expanded paths
+          base_path = File.expand_path(File.join(source, locales_dir))
+          expanded_path = File.expand_path(normal_path)
+
+          # Security: Validate path stays within base directory
+          valid_path = expanded_path.start_with?(base_path + File::SEPARATOR) ||
+            expanded_path == base_path
+          unless valid_path
+            raise Jekyll::L10n::Errors::PoFileError,
+                  "Path traversal attempt detected in page_path: #{page_path}"
+          end
+
+          normal_path
         end
       end
     end

data/lib/jekyll-l10n/translation/libre_translator.rb

@@ -198,6 +198,18 @@ module Jekyll
       def make_api_request(text, target_locale)
         uri = URI("#{@config.libretranslate_api_url}/translate")
         Jekyll.logger.debug "LibreTranslator", "Requesting #{uri}"
+        request = build_request(uri, text, target_locale)
+        response = http_config.request(request)
+        Jekyll.logger.debug "LibreTranslator", "Response code: #{response.code}"
+        handle_api_response(response)
+      rescue Net::ReadTimeout, Net::OpenTimeout => e
+        raise TranslationError, "API timeout: #{e.message}"
+      rescue Errno::ECONNREFUSED, Errno::ETIMEDOUT, SocketError => e
+        Jekyll.logger.warn "LibreTranslator", "#{e.class.name}: #{e.message}"
+        nil
+      end
+
+      def build_request(uri, text, target_locale)
         request = Net::HTTP::Post.new(uri, api_headers)
         request.body = {
           :q      => text,
@@ -205,11 +217,7 @@ module Jekyll
           :target => target_locale,
           :format => @config.libretranslate_format,
         }.to_json
-        response = http_config.request(request)
-        Jekyll.logger.debug "LibreTranslator", "Response code: #{response.code}"
-        handle_api_response(response)
-      rescue Net::ReadTimeout, Net::OpenTimeout => e
-        raise TranslationError, "API timeout: #{e.message}"
+        request
       end
 
       def handle_api_response(response)

data/lib/jekyll-l10n/utils/site_config_accessor.rb

@@ -21,7 +21,7 @@ module Jekyll
     class SiteConfigAccessor
       # Extract localization configuration from site.
       #
-      # Accesses the localization_gettext configuration which contains the
+      # Accesses the with_locales_data configuration which contains the
       # locales, extraction settings, and other localization options.
       #
       # @param site [Jekyll::Site, Hash] Jekyll site object or hash double

data/lib/jekyll-l10n.rb

@@ -1,5 +1,12 @@
 # frozen_string_literal: true
 
+# Require runtime dependencies explicitly
+# These are declared in gemspec and must be loaded before using their classes
+require "jekyll"
+require "liquid"
+require "nokogiri"
+require "gettext/po"
+
 # # jekyll-l10n: Complete Site Localization for Jekyll
 #
 # A comprehensive localization plugin for Jekyll that provides GNU Gettext-based translation
@@ -189,9 +196,6 @@ require_relative "jekyll-l10n/jekyll/regeneration_checker"
 require_relative "jekyll-l10n/jekyll/generator"
 require_relative "jekyll-l10n/jekyll/post_write_processor"
 require_relative "jekyll-l10n/jekyll/post_write_html_reprocessor"
-require_relative "jekyll-l10n/jekyll/file_sync"
-require_relative "jekyll-l10n/jekyll/page_locator"
-require_relative "jekyll-l10n/jekyll/page_writer"
 require_relative "jekyll-l10n/jekyll/url_filter"
 require_relative "jekyll-l10n/utils/text_normalizer"
 require_relative "jekyll-l10n/utils/translation_resolver"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: jekyll-l10n
 version: !ruby/object:Gem::Version
-  version: 1.1.1
+  version: 1.2.0
 platform: ruby
 authors:
 - ReleaseBot
@@ -111,6 +111,48 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '3.13'
+- !ruby/object:Gem::Dependency
+  name: simplecov
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.22'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.22'
+- !ruby/object:Gem::Dependency
+  name: simplecov-console
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.9'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.9'
+- !ruby/object:Gem::Dependency
+  name: webmock
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.23'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.23'
 description: Jekyll plugin for jekyll site localization.
 email:
 - group_58921183_bot_c5520aeba366578e2e444dd181ff3e23@noreply.gitlab.com
@@ -132,12 +174,8 @@ files:
 - lib/jekyll-l10n/extraction/html_string_extractor.rb
 - lib/jekyll-l10n/extraction/logger.rb
 - lib/jekyll-l10n/extraction/result_saver.rb
-- lib/jekyll-l10n/jekyll/file_sync.rb
 - lib/jekyll-l10n/jekyll/generator.rb
 - lib/jekyll-l10n/jekyll/localized_page.rb
-- lib/jekyll-l10n/jekyll/localized_page_mapper.rb
-- lib/jekyll-l10n/jekyll/page_locator.rb
-- lib/jekyll-l10n/jekyll/page_writer.rb
 - lib/jekyll-l10n/jekyll/post_write_html_reprocessor.rb
 - lib/jekyll-l10n/jekyll/post_write_processor.rb
 - lib/jekyll-l10n/jekyll/regeneration_checker.rb

data/lib/jekyll-l10n/jekyll/file_sync.rb

@@ -1,110 +0,0 @@
-# frozen_string_literal: true
-
-require_relative "../translation/translator"
-require_relative "page_writer"
-require_relative "localized_page_mapper"
-require_relative "../utils/url_path_builder"
-require_relative "../utils/logger_formatter"
-
-module Jekyll
-  module L10n
-    # File Synchronization Manager - Syncs localized pages and applies translations
-    #
-    # This class manages the synchronization of generated HTML content to localized page variants
-    # and applies translations to each variant. It works in coordination with Jekyll's file
-    # writing system to ensure all locale-specific content is properly translated.
-    #
-    # The sync process:
-    # 1. Builds a map of localized pages grouped by original URL
-    # 2. For each original page, finds its localized variants
-    # 3. Reads the original HTML file content
-    # 4. Applies translations to each localized variant
-    # 5. Writes the translated HTML to disk
-    #
-    # This class is typically used during the post-write phase to handle final translation
-    # application after Jekyll has written all files to disk.
-    #
-    # @example Usage (typically internal)
-    #   file_reader = SomeFileReader.new
-    #   page_writer = PageWriter.new
-    #   syncer = LocalizationFileSync.new(site, file_reader, page_writer)
-    #   syncer.sync
-    #
-    class LocalizationFileSync
-      # Initialize the file synchronization manager
-      #
-      # @param site [Jekyll::Site] The Jekyll site object
-      # @param file_reader [Object] A file reader object with a `read(page)` method
-      # @param page_writer [PageWriter] A page writer object to write translated HTML
-      def initialize(site, file_reader, page_writer)
-        @site = site
-        @file_reader = file_reader
-        @page_writer = page_writer
-      end
-
-      # Synchronize and translate all localized pages
-      #
-      # Builds a map of localized pages, reads original HTML content, and applies
-      # translations to each locale variant. Logs the process for debugging.
-      #
-      # @return [void]
-      def sync
-        Jekyll.logger.info "Localization", "Starting file sync process"
-        localized_pages_by_url = LocalizedPageMapper.build_map(@site)
-        count = localized_pages_by_url.keys.length
-        Jekyll.logger.info "Localization",
-                           "Found #{count} pages with localized versions"
-        sync_files_and_apply_translations(localized_pages_by_url)
-      end
-
-      private
-
-      def sync_files_and_apply_translations(localized_pages_by_url)
-        @site.pages.each do |page|
-          with_locales = page.data["with_locales"]
-          is_localized = page.data["localized"]
-          log_page_check(page.url, with_locales, is_localized)
-          next unless page.data["with_locales"] == true && page.data["localized"] != true
-
-          LoggerFormatter.debug_if_enabled("FileSync",
-                                           "Processing page for translation: #{page.url}")
-          original_content = @file_reader.read(page)
-          next unless original_content
-
-          apply_translations_to_localized_pages(page, localized_pages_by_url, original_content)
-        end
-      end
-
-      def apply_translations_to_localized_pages(page, localized_pages_by_url, original_content)
-        return unless localized_pages_by_url[page.url]
-
-        localized_pages_by_url[page.url].each do |localized_page|
-          translate_and_write_localized_page(localized_page, original_content)
-        end
-      end
-
-      def translate_and_write_localized_page(localized_page, original_content)
-        content_to_translate = if localized_page.output && !localized_page.output.empty?
-                                 localized_page.output
-                               else
-                                 original_content
-                               end
-
-        localized_page.instance_variable_set(:@output, content_to_translate)
-
-        locale = localized_page.data["locale"]
-        baseurl = @site.config["baseurl"]
-
-        translator = Translator.new(localized_page)
-
-        @page_writer.translate_and_write(localized_page, translator, locale, baseurl)
-      end
-
-      def log_page_check(url, with_locales, is_localized)
-        LoggerFormatter.debug_if_enabled("FileSync",
-                                         "Checking page: #{url} (with_locales: #{with_locales}, " \
-                                         "localized: #{is_localized})")
-      end
-    end
-  end
-end

data/lib/jekyll-l10n/jekyll/localized_page_mapper.rb

@@ -1,51 +0,0 @@
-# frozen_string_literal: true
-
-require_relative "../utils/logger_formatter"
-
-module Jekyll
-  module L10n
-    # Localized Page Mapper - Indexes generated localized pages by their original URL
-    #
-    # This module builds a mapping of original URLs to their localized page variants.
-    # It scans all pages in the Jekyll site, identifies those marked as localized,
-    # and groups them by their original_url for quick lookup during processing.
-    #
-    # The map structure is: { original_url => [localized_page_1, localized_page_2, ...] }
-    # This enables efficient matching of localized variants back to their source pages.
-    #
-    # @example Usage
-    #   localized_map = LocalizedPageMapper.build_map(site)
-    #   # => { "/about/" => [<LocalizedPage locale="es">, <LocalizedPage locale="fr">] }
-    #
-    module LocalizedPageMapper
-      extend self
-
-      # Build a mapping of original URLs to localized page variants
-      #
-      # Scans all pages in the Jekyll site, finds those marked with `localized: true`,
-      # and groups them by their original_url. Returns a hash with original URLs as keys
-      # and arrays of LocalizedPage objects as values.
-      #
-      # @param site [Jekyll::Site] The Jekyll site object with all generated pages
-      # @return [Hash<String, Array<LocalizedPage>>] Map of original URL to localized variants
-      # @example
-      #   pages_map = LocalizedPageMapper.build_map(site)
-      #   spanish_pages = pages_map["/about/"]  # => [<LocalizedPage locale="es">]
-      def build_map(site)
-        localized_pages = {}
-        site.pages.each do |page|
-          next unless page.data["localized"] == true
-
-          original_url = page.data["original_url"]
-          LoggerFormatter.debug_if_enabled(
-            "LocalizedPageMapper",
-            "Found localized page: #{page.url} (original: #{original_url})"
-          )
-          localized_pages[original_url] ||= []
-          localized_pages[original_url] << page
-        end
-        localized_pages
-      end
-    end
-  end
-end

data/lib/jekyll-l10n/jekyll/page_locator.rb

@@ -1,59 +0,0 @@
-# frozen_string_literal: true
-
-module Jekyll
-  module L10n
-    # Locates original (non-localized) pages by URL.
-    #
-    # OriginalPageLocator builds an index of original pages in the site and
-    # provides fast lookup by URL. This is used during translation to find the
-    # original page configuration when processing localized variants.
-    #
-    # Key responsibilities:
-    # * Index original pages by URL
-    # * Exclude localized page variants from index
-    # * Provide O(1) lookup by URL
-    # * Lazily build index on first use
-    #
-    # @example
-    #   locator = OriginalPageLocator.new(site)
-    #   original_page = locator.find_by_url('/docs/index.html')
-    class OriginalPageLocator
-      # Initialize a new OriginalPageLocator.
-      #
-      # @param site [Jekyll::Site] Jekyll site object
-      def initialize(site)
-        @site = site
-        @index = nil
-      end
-
-      # Find an original page by URL.
-      #
-      # Builds index on first call, then uses cached index for subsequent lookups.
-      # Returns nil if page not found.
-      #
-      # @param url [String] Page URL (e.g., '/docs/index.html')
-      # @return [Jekyll::Page, nil] Original page if found, nil otherwise
-      def find_by_url(url)
-        build_index unless @index
-        @index[url]
-      end
-
-      private
-
-      # Build index of original pages by URL.
-      #
-      # Indexes all pages that are not marked as localized, allowing fast
-      # lookup by URL without iterating through site.pages each time.
-      #
-      # @return [void]
-      def build_index
-        @index = {}
-        @site.pages.each do |page|
-          next if page.data["localized"] == true
-
-          @index[page.url] = page
-        end
-      end
-    end
-  end
-end

data/lib/jekyll-l10n/jekyll/page_writer.rb

@@ -1,120 +0,0 @@
-# frozen_string_literal: true
-
-require_relative "../utils/url_path_builder"
-require_relative "../translation/html_translator"
-require_relative "../utils/file_operations"
-require_relative "../utils/logger_formatter"
-
-module Jekyll
-  module L10n
-    # Writes localized pages with metadata to disk.
-    #
-    # LocalizedPageWriter translates page output and updates locale metadata
-    # (html lang attribute) before writing localized pages to the build output
-    # directory. It ensures proper directory structure and updates locale information
-    # in the HTML tag.
-    #
-    # Key responsibilities:
-    # * Apply translations to localized page output
-    # * Update HTML lang attribute to target locale
-    # * Create necessary directory structure
-    # * Write localized HTML to disk with UTF-8 encoding
-    # * Clean up auto-inserted meta charset tags
-    # * Handle parse errors gracefully
-    #
-    # @example
-    #   writer = LocalizedPageWriter.new('_site')
-    #   writer.translate_and_write(page, translator, 'es', '/baseurl')
-    #   # Localized page written to disk with translations and lang attribute updated
-    class LocalizedPageWriter
-      # Initialize a new LocalizedPageWriter.
-      #
-      # @param dest [String] Destination build directory
-      def initialize(dest)
-        @dest = dest
-      end
-
-      # Translate page content and write to disk.
-      #
-      # Applies translator, updates HTML lang attribute with locale, ensures
-      # output directory exists, and writes translated HTML to file.
-      #
-      # @param localized_page [Jekyll::Page] Localized page to write
-      # @param translator [Object] Translator object with translate method
-      # @param locale [String] Target locale code
-      # @param _baseurl [String] Base URL (passed for compatibility, not used)
-      # @return [void]
-      def translate_and_write(localized_page, translator, locale, _baseurl)
-        log_debug_info(localized_page, locale, "start")
-        translator.translate
-
-        log_debug_info(localized_page, locale, "after translate")
-        localized_page.output = fix_locale_metadata(localized_page.output, locale)
-
-        log_debug_info(localized_page, locale, "after fix_locale")
-        write_localized_page(localized_page)
-      end
-
-      private
-
-      def log_debug_info(localized_page, locale, phase)
-        LoggerFormatter.debug_if_enabled("PageWriter",
-                                         "#{phase}: URL=#{localized_page.url}, locale=#{locale}, " \
-                                         "output_size=#{localized_page.output&.length || 0}")
-      end
-
-      def write_localized_page(localized_page)
-        localized_file_path = UrlPathBuilder.url_to_file_path(localized_page.url)
-        localized_file = File.join(@dest, localized_file_path)
-        FileOperations.ensure_directory(localized_file)
-
-        LoggerFormatter.debug_if_enabled("PageWriter", "Writing to #{localized_file_path}")
-        FileOperations.write_utf8(localized_file, localized_page.output)
-      end
-
-      def fix_locale_metadata(html, locale)
-        return html unless html && locale
-
-        doc = parse_html(html)
-        update_html_lang_attribute(doc, locale)
-        result = serialize_html(doc)
-        cleanup_auto_inserted_meta_tag(result)
-      rescue StandardError => e
-        Jekyll.logger.error "Localization",
-                            "Failed to parse HTML for locale #{locale}: #{e.message}"
-        html
-      end
-
-      def parse_html(html)
-        # CRITICAL: Nokogiri::HTML auto-inserts <meta http-equiv="Content-Type">
-        # We parse with HTML to access the <html> tag, then remove the auto-inserted
-        # meta tag using regex post-processing.
-        # See: spec/regression/nokogiri_meta_tag_spec.rb
-        Nokogiri::HTML(html)
-      end
-
-      def update_html_lang_attribute(doc, locale)
-        html_tag = doc.at("html")
-
-        if html_tag
-          html_tag["lang"] = locale
-        else
-          Jekyll.logger.warn("Localization",
-                             "No <html> tag found for locale #{locale}, skipping lang attribute")
-        end
-      end
-
-      def serialize_html(doc)
-        doc.to_html
-      end
-
-      def cleanup_auto_inserted_meta_tag(result)
-        # Remove the auto-inserted meta tag by libxml2 during HTML serialization
-        # Matches: <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
-        # See: spec/regression/nokogiri_meta_tag_spec.rb
-        pattern = %r!<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">\n?!
-        result.gsub(pattern, "")
-      end
-    end
-  end
-end