html-to-markdown 2.29.0 → 3.0.0

data/lib/html_to_markdown.rb

@@ -7,205 +7,24 @@ module HtmlToMarkdown
   autoload :CLI, 'html_to_markdown/cli'
   autoload :CLIProxy, 'html_to_markdown/cli_proxy'
 
-  class Options; end # rubocop:disable Lint/EmptyClass
-
   class << self
     alias native_convert convert
-    alias native_convert_with_inline_images convert_with_inline_images
-    alias native_convert_with_inline_images_handle convert_with_inline_images_handle
-    alias native_options options
-    alias native_convert_with_options convert_with_options
-    alias native_convert_with_metadata convert_with_metadata
-    alias native_convert_with_metadata_handle convert_with_metadata_handle
-    alias native_convert_with_visitor convert_with_visitor
-    alias native_convert_with_tables convert_with_tables
   end
 
   module_function
 
-  def convert(html, options = nil, visitor = nil)
-    if visitor
-      native_convert_with_visitor(html.to_s, options, visitor)
-    else
-      native_convert(html.to_s, options)
-    end
-  end
-
-  def convert_with_options(html, options_handle)
-    native_convert_with_options(html.to_s, options_handle)
-  end
-
-  def convert_with_inline_images(html, options = nil, image_config = nil, _visitor = nil)
-    # NOTE: visitor parameter is accepted for API compatibility but not used in inline images mode
-    # The visitor pattern is only supported in the standard convert() method
-    native_convert_with_inline_images(html.to_s, options, image_config)
-  end
-
-  def convert_with_inline_images_handle(html, options_handle, image_config = nil)
-    native_convert_with_inline_images_handle(html.to_s, options_handle, image_config)
-  end
-
-  def options(options_hash = nil)
-    native_options(options_hash)
-  end
-
-  # Convert HTML to Markdown with comprehensive metadata extraction.
-  #
-  # Performs HTML-to-Markdown conversion while extracting document metadata, headers,
-  # links, images, and structured data in a single pass. Ideal for content analysis,
-  # SEO workflows, and document indexing.
-  #
-  # @param html [String] HTML string to convert. Line endings are normalized (CRLF -> LF).
-  # @param options [ConversionOptions, Hash, nil] Optional conversion configuration.
-  #   When a Hash, keys should match ConversionOptions field names (as symbols or strings).
-  #   Common options:
-  #   - :heading_style [String] "atx", "atx_closed", or "underlined" (default: "underlined")
-  #   - :list_indent_type [String] "spaces" or "tabs" (default: "spaces")
-  #   - :list_indent_width [Integer] Spaces per indent level (default: 4)
-  #   - :wrap [true, false] Enable text wrapping (default: false)
-  #   - :wrap_width [Integer] Wrap at this column width (default: 80)
-  #   See ConversionOptions documentation for complete list.
-  #
-  # @param metadata_config [Hash, nil] Optional metadata extraction configuration.
-  #   Keys should be symbols or strings. Supported keys:
-  #   - :extract_headers [true, false] Extract h1-h6 heading elements (default: true)
-  #   - :extract_links [true, false] Extract hyperlinks with type classification (default: true)
-  #   - :extract_images [true, false] Extract image elements (default: true)
-  #   - :extract_structured_data [true, false] Extract JSON-LD/Microdata/RDFa (default: true)
-  #   - :max_structured_data_size [Integer] Size limit for structured data in bytes (default: 1_000_000)
-  #
-  # @return [Array<String, Hash>] Tuple of [markdown_string, metadata_hash]
-  #   markdown_string: String - The converted Markdown output
-  #
-  #   metadata_hash: Hash with keys:
-  #   - :document [Hash] Document-level metadata:
-  #     - :title [String, nil] From <title> tag
-  #     - :description [String, nil] From <meta name="description">
-  #     - :keywords [Array<String>] From <meta name="keywords">
-  #     - :author [String, nil] From <meta name="author">
-  #     - :language [String, nil] From lang attribute (e.g., "en")
-  #     - :text_direction [String, nil] "ltr", "rtl", or "auto"
-  #     - :canonical_url [String, nil] From <link rel="canonical">
-  #     - :base_href [String, nil] From <base href="">
-  #     - :open_graph [Hash<String, String>] Open Graph properties (og:* meta tags)
-  #     - :twitter_card [Hash<String, String>] Twitter Card properties (twitter:* meta tags)
-  #     - :meta_tags [Hash<String, String>] Other meta tags
-  #
-  #   - :headers [Array<Hash>] Heading elements:
-  #     - :level [Integer] 1-6
-  #     - :text [String] Header text content
-  #     - :id [String, nil] HTML id attribute
-  #     - :depth [Integer] Tree nesting depth
-  #     - :html_offset [Integer] Byte offset in original HTML
-  #
-  #   - :links [Array<Hash>] Hyperlinks:
-  #     - :href [String] Link URL
-  #     - :text [String] Link text content
-  #     - :title [String, nil] Title attribute
-  #     - :link_type [String] "anchor", "internal", "external", "email", "phone", or "other"
-  #     - :rel [Array<String>] Rel attribute values
-  #     - :attributes [Hash<String, String>] Additional HTML attributes
-  #
-  #   - :images [Array<Hash>] Image elements:
-  #     - :src [String] Image source URL or data URI
-  #     - :alt [String, nil] Alt text for accessibility
-  #     - :title [String, nil] Title attribute
-  #     - :dimensions [Array<Integer>, nil] [width, height] if available
-  #     - :image_type [String] "data_uri", "external", "relative", or "inline_svg"
-  #     - :attributes [Hash<String, String>] Additional HTML attributes
-  #
-  #   - :structured_data [Array<Hash>] Structured data blocks:
-  #     - :data_type [String] "json_ld", "microdata", or "rdfa"
-  #     - :raw_json [String] Raw JSON content
-  #     - :schema_type [String, nil] Schema type (e.g., "Article", "Event")
-  #
-  # @raise [StandardError] If conversion fails or invalid configuration
-  #
-  # @example Basic usage
-  #   html = <<~HTML
-  #     <html lang="en">
-  #       <head>
-  #         <title>My Article</title>
-  #         <meta name="description" content="A great read">
-  #       </head>
-  #       <body>
-  #         <h1 id="intro">Introduction</h1>
-  #         <p>Visit <a href="https://example.com">our site</a></p>
-  #         <img src="photo.jpg" alt="Beautiful landscape">
-  #       </body>
-  #     </html>
-  #   HTML
-  #
-  #   markdown, metadata = HtmlToMarkdown.convert_with_metadata(html)
-  #
-  #   puts metadata[:document][:title]  # => "My Article"
-  #   puts metadata[:document][:language]  # => "en"
-  #   puts metadata[:headers].length  # => 1
-  #   puts metadata[:headers][0][:text]  # => "Introduction"
-  #   puts metadata[:links].length  # => 1
-  #   puts metadata[:images].length  # => 1
-  #
-  # @example With selective metadata extraction
-  #   config = {
-  #     extract_headers: true,
-  #     extract_links: true,
-  #     extract_images: false,      # Skip images
-  #     extract_structured_data: false  # Skip structured data
-  #   }
-  #
-  #   markdown, metadata = HtmlToMarkdown.convert_with_metadata(html, nil, config)
-  #   puts metadata[:images].empty?  # => true (not extracted)
-  #
-  # @example With conversion options
-  #   options = {
-  #     heading_style: "atx",     # Use # H1, ## H2 style
-  #     wrap: true,
-  #     wrap_width: 80
-  #   }
-  #
-  #   config = { extract_headers: true }
-  #
-  #   markdown, metadata = HtmlToMarkdown.convert_with_metadata(html, options, config)
-  #   # Markdown uses ATX-style headings and wraps at 80 characters
-  #
-  # @see #convert Simple conversion without metadata
-  # @see #convert_with_inline_images Extract inline images during conversion
-  # @see ConversionOptions Detailed conversion configuration
-  def convert_with_metadata(html, options = nil, metadata_config = nil, _visitor = nil)
-    # NOTE: visitor parameter is accepted for API compatibility but not used in metadata extraction mode
-    # The visitor pattern is only supported in the standard convert() method
-    native_convert_with_metadata(html.to_s, options, metadata_config)
-  end
-
-  def convert_with_metadata_handle(html, options_handle, metadata_config = nil)
-    native_convert_with_metadata_handle(html.to_s, options_handle, metadata_config)
-  end
-
-  # Convert HTML to Markdown with table extraction.
-  #
-  # Performs HTML-to-Markdown conversion while extracting structured table data
-  # (cells, markdown representation, header row flags) in a single pass.
-  #
-  # @param html [String] HTML string to convert.
-  # @param options [Hash, nil] Optional conversion configuration.
-  # @param metadata_config [Hash, nil] Optional metadata extraction configuration.
-  #
-  # @return [Hash] A hash with keys:
-  #   - :content [String] The converted Markdown output
-  #   - :metadata [Hash, nil] Extended metadata (if metadata extraction was configured)
-  #   - :tables [Array<Hash>] Extracted tables, each with:
-  #     - :cells [Array<Array<String>>] Table cells organized as rows x columns
-  #     - :markdown [String] Complete rendered table in Markdown format
-  #     - :is_header_row [Array<Boolean>] Per-row flag indicating header rows
-  #
-  # @raise [StandardError] If conversion fails or invalid configuration
-  #
-  # @example Basic usage
-  #   html = '<table><thead><tr><th>Name</th></tr></thead><tbody><tr><td>Alice</td></tr></tbody></table>'
-  #   result = HtmlToMarkdown.convert_with_tables(html)
-  #   puts result[:tables].length    # => 1
-  #   puts result[:tables][0][:cells] # => [["Name"], ["Alice"]]
-  def convert_with_tables(html, options = nil, metadata_config = nil)
-    native_convert_with_tables(html.to_s, options, metadata_config)
+  # Convert HTML to Markdown, returning a Hash with:
+  #   - :content [String, nil] the converted Markdown output
+  #   - :document [nil] document structure (not yet exposed)
+  #   - :metadata [Hash, nil] extracted HTML metadata
+  #   - :tables [Array<Hash>] extracted tables with :grid and :markdown
+  #   - :images [Array<Hash>] extracted inline images
+  #   - :warnings [Array<Hash>] processing warnings
+  #
+  # @param html [String] HTML string to convert
+  # @param options [Hash, nil] optional conversion options
+  # @return [Hash] conversion result
+  def convert(html, options = nil)
+    native_convert(html.to_s, options)
   end
 end

data/sig/html_to_markdown.rbs

@@ -2,36 +2,6 @@
 module HtmlToMarkdown
   VERSION: String
 
-  # Opaque handle for reusable conversion options
-  class Options
-  end
-
-  # Visitor context information passed to visitor callbacks
-  class NodeContext
-    attr_reader node_type: Symbol
-    attr_reader tag_name: String
-    attr_reader attributes: Hash[String, String]
-    attr_reader depth: Integer
-    attr_reader index_in_parent: Integer
-    attr_reader parent_tag: String | nil
-    attr_reader is_inline: bool
-  end
-
-  # Result of a visitor callback
-  type visitor_result = {
-    type: :continue,
-  } | {
-    type: :custom,
-    output: String,
-  } | {
-    type: :skip,
-  } | {
-    type: :preserve_html,
-  } | {
-    type: :error,
-    message: String,
-  }
-
   type heading_style = :underlined | :atx | :atx_closed
   type list_indent_type = :spaces | :tabs
   type highlight_style = :double_equal | :html | :bold | :none
@@ -62,7 +32,6 @@ module HtmlToMarkdown
     autolinks?: bool,
     default_title?: bool,
     br_in_tables?: bool,
-    hocr_spatial_tables?: bool,
     highlight_style?: highlight_style,
     extract_metadata?: bool,
     whitespace_mode?: whitespace_mode,
@@ -81,50 +50,14 @@ module HtmlToMarkdown
     strip_tags?: Array[String],
     preserve_tags?: Array[String],
     output_format?: output_format,
-    skip_images?: bool
-  }
-
-  type inline_image_config = {
-    max_decoded_size_bytes?: Integer,
-    filename_prefix?: String?,
+    skip_images?: bool,
+    include_document_structure?: bool,
+    extract_images?: bool,
+    max_image_size?: Integer,
     capture_svg?: bool,
     infer_dimensions?: bool
   }
 
-  type inline_image_format = "png" | "jpeg" | "gif" | "bmp" | "webp" | "svg" | String
-
-  type inline_image_source = "img_data_uri" | "svg_element"
-
-  type inline_image = {
-    data: String,
-    format: inline_image_format,
-    filename: String?,
-    description: String?,
-    dimensions: [Integer, Integer]?,
-    source: inline_image_source,
-    attributes: Hash[String, String]
-  }
-
-  type inline_image_warning = {
-    index: Integer,
-    message: String
-  }
-
-  type html_extraction = {
-    markdown: String,
-    inline_images: Array[inline_image],
-    warnings: Array[inline_image_warning]
-  }
-
-  type metadata_config = {
-    extract_document?: bool,
-    extract_headers?: bool,
-    extract_links?: bool,
-    extract_images?: bool,
-    extract_structured_data?: bool,
-    max_structured_data_size?: Integer
-  }
-
   type text_direction = "ltr" | "rtl" | "auto" | nil
 
   type document_metadata = {
@@ -185,314 +118,20 @@ module HtmlToMarkdown
     structured_data: Array[structured_data]
   }
 
-  type table_data = {
-    cells: Array[Array[String]],
-    markdown: String,
-    is_header_row: Array[bool]
-  }
-
-  type table_extraction_result = {
-    content: String,
-    metadata: extended_metadata?,
-    tables: Array[table_data]
-  }
-
-  # Native methods (implemented in Rust via Magnus/rb-sys)
-  # These are aliased from the Rust extension and available as both module and instance methods
+  # Native method (implemented in Rust via Magnus/rb-sys)
   private
 
-  def self.native_convert: (String html, conversion_options? options) -> String
-  def self.native_options: (conversion_options? options_hash) -> Options
-  def self.native_convert_with_options: (String html, Options options_handle) -> String
-  def self.native_convert_with_inline_images_handle: (
-    String html,
-    Options options_handle,
-    inline_image_config? image_config
-  ) -> html_extraction
-  def self.native_convert_with_inline_images: (
-    String html,
-    conversion_options? options,
-    inline_image_config? image_config
-  ) -> html_extraction
-  def self.native_convert_with_metadata_handle: (
-    String html,
-    Options options_handle,
-    metadata_config? metadata_config
-  ) -> [String, extended_metadata]
-  def self.native_convert_with_metadata: (
-    String html,
-    conversion_options? options,
-    metadata_config? metadata_config
-  ) -> [String, extended_metadata]
-  def self.native_convert_with_visitor: (
-    String html,
-    conversion_options? options,
-    visitor? visitor
-  ) -> String
-  def self.native_convert_with_tables: (
-    String html,
-    conversion_options? options,
-    metadata_config? metadata_config
-  ) -> table_extraction_result
-
-  def native_convert: (String html, conversion_options? options) -> String
-  def native_options: (conversion_options? options_hash) -> Options
-  def native_convert_with_options: (String html, Options options_handle) -> String
-  def native_convert_with_inline_images_handle: (
-    String html,
-    Options options_handle,
-    inline_image_config? image_config
-  ) -> html_extraction
-  def native_convert_with_inline_images: (
-    String html,
-    conversion_options? options,
-    inline_image_config? image_config
-  ) -> html_extraction
-  def native_convert_with_metadata_handle: (
-    String html,
-    Options options_handle,
-    metadata_config? metadata_config
-  ) -> [String, extended_metadata]
-  def native_convert_with_metadata: (
-    String html,
-    conversion_options? options,
-    metadata_config? metadata_config
-  ) -> [String, extended_metadata]
-  def native_convert_with_visitor: (
-    String html,
-    conversion_options? options,
-    visitor? visitor
-  ) -> String
-  def native_convert_with_tables: (
-    String html,
-    conversion_options? options,
-    metadata_config? metadata_config
-  ) -> table_extraction_result
-
-  # Visitor interface for customizing conversion behavior
-  type visitor = Object
+  def self.native_convert: (String html, conversion_options? options) -> Hash[String, untyped]
+  def native_convert: (String html, conversion_options? options) -> Hash[String, untyped]
 
   public
 
-  # Convert HTML to Markdown with optional configuration and visitor
-  #
-  # The optional visitor parameter allows customization of conversion behavior for specific elements.
-  # When both options and visitor are provided, the visitor can override default conversions.
-  #
-  # Args:
-  #   html: HTML string to convert
-  #   options: Optional conversion configuration
-  #   visitor: Optional visitor object for customizing conversion
-  #
-  # Returns:
-  #   markdown: String - Converted markdown output
-  #
-  # Example:
-  #   markdown = HtmlToMarkdown.convert(html, { wrap: true }, my_visitor)
-  def self.convert: (String html, ?conversion_options options, ?visitor visitor) -> String
-
-  # Create a reusable options handle for performance
-  def self.options: (?conversion_options options_hash) -> Options
-
-  # Convert HTML using a pre-built options handle
-  def self.convert_with_options: (String html, Options options_handle) -> String
-  def self.convert_with_inline_images_handle: (
-    String html,
-    Options options_handle,
-    ?inline_image_config image_config
-  ) -> html_extraction
-
-  # Convert HTML with inline image extraction
-  #
-  # Optionally accepts a visitor for customizing conversion behavior.
-  #
-  # Args:
-  #   html: HTML string to convert
-  #   options: Optional conversion configuration
-  #   image_config: Optional inline image extraction configuration
-  #   visitor: Optional visitor object for customizing conversion
-  #
-  # Returns:
-  #   html_extraction: Hash containing markdown, inline_images array, and warnings array
-  #
-  # Example:
-  #   result = HtmlToMarkdown.convert_with_inline_images(html, { wrap: true }, image_config, my_visitor)
-  def self.convert_with_inline_images: (
-    String html,
-    ?conversion_options options,
-    ?inline_image_config image_config,
-    ?visitor visitor
-  ) -> html_extraction
-
-  # Convert HTML to Markdown with a custom visitor (deprecated)
-  #
-  # DEPRECATED: Use convert() with the optional visitor parameter instead.
-  # This method is maintained for backward compatibility.
-  #
-  # All convert functions now accept optional visitors:
-  #   - convert(html, options, visitor)
-  #   - convert_with_inline_images(html, options, image_config, visitor)
-  #   - convert_with_metadata(html, options, metadata_config, visitor)
-  #
-  # The visitor object can implement any of the following methods:
-  #   - visit_element_start(ctx) -> visitor_result
-  #   - visit_element_end(ctx, output) -> visitor_result
-  #   - visit_text(ctx, text) -> visitor_result
-  #   - visit_link(ctx, href, text, title) -> visitor_result
-  #   - visit_image(ctx, src, alt, title) -> visitor_result
-  #   - visit_heading(ctx, level, text, id) -> visitor_result
-  #   - visit_code_block(ctx, lang, code) -> visitor_result
-  #   - visit_code_inline(ctx, code) -> visitor_result
-  #   - visit_list_item(ctx, ordered, marker, text) -> visitor_result
-  #   - visit_list_start(ctx, ordered) -> visitor_result
-  #   - visit_list_end(ctx, ordered, output) -> visitor_result
-  #   - visit_table_start(ctx) -> visitor_result
-  #   - visit_table_row(ctx, cells, is_header) -> visitor_result
-  #   - visit_table_end(ctx, output) -> visitor_result
-  #   - visit_blockquote(ctx, content, depth) -> visitor_result
-  #   - visit_strong(ctx, text) -> visitor_result
-  #   - visit_emphasis(ctx, text) -> visitor_result
-  #   - visit_strikethrough(ctx, text) -> visitor_result
-  #   - visit_underline(ctx, text) -> visitor_result
-  #   - visit_subscript(ctx, text) -> visitor_result
-  #   - visit_superscript(ctx, text) -> visitor_result
-  #   - visit_mark(ctx, text) -> visitor_result
-  #   - visit_line_break(ctx) -> visitor_result
-  #   - visit_horizontal_rule(ctx) -> visitor_result
-  #   - visit_custom_element(ctx, tag_name, html) -> visitor_result
-  #   - visit_definition_list_start(ctx) -> visitor_result
-  #   - visit_definition_term(ctx, text) -> visitor_result
-  #   - visit_definition_description(ctx, text) -> visitor_result
-  #   - visit_definition_list_end(ctx, output) -> visitor_result
-  #   - visit_form(ctx, action, method) -> visitor_result
-  #   - visit_input(ctx, input_type, name, value) -> visitor_result
-  #   - visit_button(ctx, text) -> visitor_result
-  #   - visit_audio(ctx, src) -> visitor_result
-  #   - visit_video(ctx, src) -> visitor_result
-  #   - visit_iframe(ctx, src) -> visitor_result
-  #   - visit_details(ctx, open) -> visitor_result
-  #   - visit_summary(ctx, text) -> visitor_result
-  #   - visit_figure_start(ctx) -> visitor_result
-  #   - visit_figcaption(ctx, text) -> visitor_result
-  #   - visit_figure_end(ctx, output) -> visitor_result
-  #
-  # Each method should return a Hash with at least :type key:
-  #   { type: :continue } - Continue with default behavior
-  #   { type: :custom, output: "..." } - Replace with custom markdown
-  #   { type: :skip } - Skip this element entirely
-  #   { type: :preserve_html } - Keep original HTML
-  #   { type: :error, message: "..." } - Stop conversion with error
-  #
-  # Args:
-  #   html: HTML string to convert
-  #   options: Optional conversion configuration
-  #   visitor: Visitor object that responds to visitor callback methods
-  #
-  # Returns:
-  #   markdown: String - Converted markdown output
-  #
-  # Example:
-  #   class MyVisitor
-  #     def visit_link(ctx, href, text, title = nil)
-  #       { type: :custom, output: "[#{text}](#{href})" }
-  #     end
-  #   end
-  #
-  #   HtmlToMarkdown.convert_with_visitor(html, visitor: MyVisitor.new)
-  def self.convert_with_visitor: (String html, ?conversion_options options, visitor: visitor) -> String
-
-  # Convert HTML to Markdown with metadata extraction
-  #
-  # Extracts comprehensive metadata (headers, links, images, structured data) during conversion.
-  # Optionally accepts a visitor for customizing conversion behavior.
-  #
-  # Args:
-  #   html: HTML string to convert
-  #   options: Optional conversion configuration
-  #   metadata_config: Optional metadata extraction configuration
-  #   visitor: Optional visitor object for customizing conversion
-  #
-  # Returns:
-  #   Array containing:
-  #   - [0] markdown: String - Converted markdown output
-  #   - [1] metadata: Hash - Extracted metadata with document, headers, links, images, structured_data
-  #
-  # The metadata hash contains:
-  #   - document: Document-level metadata (title, description, lang, etc.)
-  #   - headers: List of header elements with hierarchy
-  #   - links: List of extracted hyperlinks with classification
-  #   - images: List of extracted images with metadata
-  #   - structured_data: List of JSON-LD, Microdata, or RDFa blocks
-  #
-  # Example:
-  #   html = '<html lang="en"><head><title>Test</title></head><body><h1>Hello</h1></body></html>'
-  #   markdown, metadata = HtmlToMarkdown.convert_with_metadata(html)
-  #   puts "Title: #{metadata['document']['title']}"
-  #   puts "Headers: #{metadata['headers'].length}"
-  #
-  # Example with visitor:
-  #   markdown, metadata = HtmlToMarkdown.convert_with_metadata(html, options, metadata_config, my_visitor)
-  def self.convert_with_metadata: (
-    String html,
-    ?conversion_options options,
-    ?metadata_config metadata_config,
-    ?visitor visitor
-  ) -> [String, extended_metadata]
-  def self.convert_with_metadata_handle: (
-    String html,
-    Options options_handle,
-    ?metadata_config metadata_config
-  ) -> [String, extended_metadata]
-
-  # Convert HTML and extract tables as structured data
-  #
-  # Args:
-  #   html: HTML string to convert
-  #   options: Optional conversion configuration
-  #   metadata_config: Optional metadata extraction configuration
-  #
-  # Returns:
-  #   table_extraction_result: Hash containing content, metadata, and tables array
+  # Convert HTML to Markdown, returning a Hash with content, metadata, tables, images, and warnings.
   #
   # Example:
-  #   result = HtmlToMarkdown.convert_with_tables(html)
-  #   puts result[:tables].length
-  def self.convert_with_tables: (
-    String html,
-    ?conversion_options options,
-    ?metadata_config metadata_config
-  ) -> table_extraction_result
+  #   result = HtmlToMarkdown.convert(html)
+  def self.convert: (String html, ?conversion_options options) -> Hash[String, untyped]
 
-  # Instance method versions (created by module_function)
-  def convert: (String html, ?conversion_options options, ?visitor visitor) -> String
-  def options: (?conversion_options options_hash) -> Options
-  def convert_with_options: (String html, Options options_handle) -> String
-  def convert_with_inline_images_handle: (
-    String html,
-    Options options_handle,
-    ?inline_image_config image_config
-  ) -> html_extraction
-  def convert_with_inline_images: (
-    String html,
-    ?conversion_options options,
-    ?inline_image_config image_config,
-    ?visitor visitor
-  ) -> html_extraction
-  def convert_with_visitor: (String html, ?conversion_options options, visitor: visitor) -> String
-  def convert_with_metadata: (
-    String html,
-    ?conversion_options options,
-    ?metadata_config metadata_config,
-    ?visitor visitor
-  ) -> [String, extended_metadata]
-  def convert_with_metadata_handle: (
-    String html,
-    Options options_handle,
-    ?metadata_config metadata_config
-  ) -> [String, extended_metadata]
-  def convert_with_tables: (
-    String html,
-    ?conversion_options options,
-    ?metadata_config metadata_config
-  ) -> table_extraction_result
+  # Instance method version (created by module_function)
+  def convert: (String html, ?conversion_options options) -> Hash[String, untyped]
 end

data/vendor/Cargo.toml

@@ -3,7 +3,7 @@ members = ["html-to-markdown-rs"]
 resolver = "2"
 
 [workspace.package]
-version = "2.29.0"
+version = "3.0.0"
 edition = "2024"
 rust-version = "1.85"
 authors = ["Na'aman Hirschfeld <naaman@kreuzberg.dev>"]
@@ -13,18 +13,21 @@ homepage = "https://kreuzberg.dev"
 
 [workspace.dependencies]
 ahash = { version = "0.8", features = ["std", "compile-time-rng"], default-features = false }
-async-trait = "0.1"
 base64 = "0.22"
 clap = { version = "4.6", features = ["derive"] }
 clap_complete = "4.6"
-clap_mangen = "0.2"
+clap_mangen = "0.3"
 encoding_rs = "0.8"
-ext-php-rs = "0.15.6"
+ext-php-rs = "0.15.7"
 html5ever = "0.39.0"
 once_cell = "1.21"
 pyo3 = { version = "0.28.2", features = ["abi3-py310"] }
+rayon = "1.11"
 regex = "1.12"
 serde = { version = "1.0", features = ["derive"] }
 serde_json = "1.0"
+tempfile = "3.27"
 thiserror = "2.0"
 tl = { package = "astral-tl", version = "0.7.11" }
+toml = "1.1"
+which = "8"