@kreuzberg/html-to-markdown-wasm 3.2.4 → 3.3.0

package/dist-web/html_to_markdown_wasm.d.ts

@@ -1,6 +1,11 @@
 /* tslint:disable */
 /* eslint-disable */
 
+/**
+ * The type of an inline text annotation.
+ *
+ * Uses internally tagged representation (`"annotation_type": "bold"`) for JSON serialization.
+ */
 export enum WasmAnnotationKind {
     Bold = 0,
     Italic = 1,
@@ -13,18 +18,52 @@ export enum WasmAnnotationKind {
     Link = 8,
 }
 
+/**
+ * Code block fence style in Markdown output.
+ *
+ * Determines how code blocks (`<pre><code>`) are rendered in Markdown.
+ */
 export enum WasmCodeBlockStyle {
     Indented = 0,
     Backticks = 1,
     Tildes = 2,
 }
 
+/**
+ * Main conversion options for HTML to Markdown conversion.
+ *
+ * Use [`ConversionOptions::builder()`] to construct, or [`Default::default()`] for defaults.
+ *
+ * # Example
+ *
+ * ```text
+ * use html_to_markdown_rs::ConversionOptions;
+ *
+ * let options = ConversionOptions::builder()
+ *     .heading_style(HeadingStyle::Atx)
+ *     .wrap(true)
+ *     .wrap_width(100)
+ *     .build();
+ * ```
+ */
 export class WasmConversionOptions {
     free(): void;
     [Symbol.dispose](): void;
+    /**
+     * Apply a partial update to these conversion options.
+     */
+    applyUpdate(_update: WasmConversionOptionsUpdate): void;
+    /**
+     * Create a new builder with default values.
+     */
     static builder(): WasmConversionOptionsBuilder;
     static default(): WasmConversionOptions;
-    constructor(heading_style?: WasmHeadingStyle | null, list_indent_type?: WasmListIndentType | null, list_indent_width?: number | null, bullets?: string | null, strong_em_symbol?: string | null, escape_asterisks?: boolean | null, escape_underscores?: boolean | null, escape_misc?: boolean | null, escape_ascii?: boolean | null, code_language?: string | null, autolinks?: boolean | null, default_title?: boolean | null, br_in_tables?: boolean | null, highlight_style?: WasmHighlightStyle | null, extract_metadata?: boolean | null, whitespace_mode?: WasmWhitespaceMode | null, strip_newlines?: boolean | null, wrap?: boolean | null, wrap_width?: number | null, convert_as_inline?: boolean | null, sub_symbol?: string | null, sup_symbol?: string | null, newline_style?: WasmNewlineStyle | null, code_block_style?: WasmCodeBlockStyle | null, keep_inline_images_in?: string[] | null, preprocessing?: WasmPreprocessingOptions | null, encoding?: string | null, debug?: boolean | null, strip_tags?: string[] | null, preserve_tags?: string[] | null, skip_images?: boolean | null, link_style?: WasmLinkStyle | null, output_format?: WasmOutputFormat | null, include_document_structure?: boolean | null, extract_images?: boolean | null, max_image_size?: bigint | null, capture_svg?: boolean | null, infer_dimensions?: boolean | null);
+    static from(update: WasmConversionOptionsUpdate): WasmConversionOptions;
+    /**
+     * Create from a partial update, applying to defaults.
+     */
+    static fromUpdate(update: WasmConversionOptionsUpdate): WasmConversionOptions;
+    constructor(heading_style?: WasmHeadingStyle | null, list_indent_type?: WasmListIndentType | null, list_indent_width?: number | null, bullets?: string | null, strong_em_symbol?: string | null, escape_asterisks?: boolean | null, escape_underscores?: boolean | null, escape_misc?: boolean | null, escape_ascii?: boolean | null, code_language?: string | null, autolinks?: boolean | null, default_title?: boolean | null, br_in_tables?: boolean | null, highlight_style?: WasmHighlightStyle | null, extract_metadata?: boolean | null, whitespace_mode?: WasmWhitespaceMode | null, strip_newlines?: boolean | null, wrap?: boolean | null, wrap_width?: number | null, convert_as_inline?: boolean | null, sub_symbol?: string | null, sup_symbol?: string | null, newline_style?: WasmNewlineStyle | null, code_block_style?: WasmCodeBlockStyle | null, keep_inline_images_in?: string[] | null, preprocessing?: WasmPreprocessingOptions | null, encoding?: string | null, debug?: boolean | null, strip_tags?: string[] | null, preserve_tags?: string[] | null, skip_images?: boolean | null, link_style?: WasmLinkStyle | null, output_format?: WasmOutputFormat | null, include_document_structure?: boolean | null, extract_images?: boolean | null, max_image_size?: bigint | null, capture_svg?: boolean | null, infer_dimensions?: boolean | null, exclude_selectors?: string[] | null, max_depth?: number | null);
     autolinks: boolean;
     brInTables: boolean;
     bullets: string;
@@ -39,6 +78,7 @@ export class WasmConversionOptions {
     escapeAsterisks: boolean;
     escapeMisc: boolean;
     escapeUnderscores: boolean;
+    excludeSelectors: string[];
     extractImages: boolean;
     extractMetadata: boolean;
     headingStyle: WasmHeadingStyle;
@@ -49,6 +89,8 @@ export class WasmConversionOptions {
     linkStyle: WasmLinkStyle;
     listIndentType: WasmListIndentType;
     listIndentWidth: number;
+    get maxDepth(): number | undefined;
+    set maxDepth(value: number | null | undefined);
     maxImageSize: bigint;
     newlineStyle: WasmNewlineStyle;
     outputFormat: WasmOutputFormat;
@@ -65,21 +107,51 @@ export class WasmConversionOptions {
     wrapWidth: number;
 }
 
+/**
+ * Builder for [`ConversionOptions`].
+ *
+ * All fields start with default values. Call `.build()` to produce the final options.
+ */
 export class WasmConversionOptionsBuilder {
     private constructor();
     free(): void;
     [Symbol.dispose](): void;
+    /**
+     * Build the final [`ConversionOptions`].
+     */
     build(): WasmConversionOptions;
+    /**
+     * Set the list of CSS selectors for elements to exclude entirely from output.
+     */
+    excludeSelectors(selectors: string[]): WasmConversionOptionsBuilder;
+    /**
+     * Set the list of HTML tag names whose `<img>` children are kept inline.
+     */
     keepInlineImagesIn(tags: string[]): WasmConversionOptionsBuilder;
+    /**
+     * Set the pre-processing options applied to the HTML before conversion.
+     */
     preprocessing(preprocessing: WasmPreprocessingOptions): WasmConversionOptionsBuilder;
+    /**
+     * Set the list of HTML tag names that are preserved verbatim in output.
+     */
     preserveTags(tags: string[]): WasmConversionOptionsBuilder;
+    /**
+     * Set the list of HTML tag names whose content is stripped from output.
+     */
     stripTags(tags: string[]): WasmConversionOptionsBuilder;
 }
 
+/**
+ * Partial update for `ConversionOptions`.
+ *
+ * Uses `Option<T>` fields for selective updates. Bindings use this to construct
+ * options from language-native types. Prefer [`ConversionOptionsBuilder`] for Rust code.
+ */
 export class WasmConversionOptionsUpdate {
     free(): void;
     [Symbol.dispose](): void;
-    constructor(heading_style?: WasmHeadingStyle | null, list_indent_type?: WasmListIndentType | null, list_indent_width?: number | null, bullets?: string | null, strong_em_symbol?: string | null, escape_asterisks?: boolean | null, escape_underscores?: boolean | null, escape_misc?: boolean | null, escape_ascii?: boolean | null, code_language?: string | null, autolinks?: boolean | null, default_title?: boolean | null, br_in_tables?: boolean | null, highlight_style?: WasmHighlightStyle | null, extract_metadata?: boolean | null, whitespace_mode?: WasmWhitespaceMode | null, strip_newlines?: boolean | null, wrap?: boolean | null, wrap_width?: number | null, convert_as_inline?: boolean | null, sub_symbol?: string | null, sup_symbol?: string | null, newline_style?: WasmNewlineStyle | null, code_block_style?: WasmCodeBlockStyle | null, keep_inline_images_in?: string[] | null, preprocessing?: WasmPreprocessingOptionsUpdate | null, encoding?: string | null, debug?: boolean | null, strip_tags?: string[] | null, preserve_tags?: string[] | null, skip_images?: boolean | null, link_style?: WasmLinkStyle | null, output_format?: WasmOutputFormat | null, include_document_structure?: boolean | null, extract_images?: boolean | null, max_image_size?: bigint | null, capture_svg?: boolean | null, infer_dimensions?: boolean | null);
+    constructor(heading_style?: WasmHeadingStyle | null, list_indent_type?: WasmListIndentType | null, list_indent_width?: number | null, bullets?: string | null, strong_em_symbol?: string | null, escape_asterisks?: boolean | null, escape_underscores?: boolean | null, escape_misc?: boolean | null, escape_ascii?: boolean | null, code_language?: string | null, autolinks?: boolean | null, default_title?: boolean | null, br_in_tables?: boolean | null, highlight_style?: WasmHighlightStyle | null, extract_metadata?: boolean | null, whitespace_mode?: WasmWhitespaceMode | null, strip_newlines?: boolean | null, wrap?: boolean | null, wrap_width?: number | null, convert_as_inline?: boolean | null, sub_symbol?: string | null, sup_symbol?: string | null, newline_style?: WasmNewlineStyle | null, code_block_style?: WasmCodeBlockStyle | null, keep_inline_images_in?: string[] | null, preprocessing?: WasmPreprocessingOptionsUpdate | null, encoding?: string | null, debug?: boolean | null, strip_tags?: string[] | null, preserve_tags?: string[] | null, skip_images?: boolean | null, link_style?: WasmLinkStyle | null, output_format?: WasmOutputFormat | null, include_document_structure?: boolean | null, extract_images?: boolean | null, max_image_size?: bigint | null, capture_svg?: boolean | null, infer_dimensions?: boolean | null, max_depth?: number | null, exclude_selectors?: string[] | null);
     get autolinks(): boolean | undefined;
     set autolinks(value: boolean | null | undefined);
     get brInTables(): boolean | undefined;
@@ -108,6 +180,8 @@ export class WasmConversionOptionsUpdate {
     set escapeMisc(value: boolean | null | undefined);
     get escapeUnderscores(): boolean | undefined;
     set escapeUnderscores(value: boolean | null | undefined);
+    get excludeSelectors(): string[] | undefined;
+    set excludeSelectors(value: string[] | null | undefined);
     get extractImages(): boolean | undefined;
     set extractImages(value: boolean | null | undefined);
     get extractMetadata(): boolean | undefined;
@@ -128,6 +202,8 @@ export class WasmConversionOptionsUpdate {
     set listIndentType(value: WasmListIndentType | null | undefined);
     get listIndentWidth(): number | undefined;
     set listIndentWidth(value: number | null | undefined);
+    get maxDepth(): number | undefined;
+    set maxDepth(value: number | null | undefined);
     get maxImageSize(): bigint | undefined;
     set maxImageSize(value: bigint | null | undefined);
     get newlineStyle(): WasmNewlineStyle | undefined;
@@ -158,6 +234,22 @@ export class WasmConversionOptionsUpdate {
     set wrapWidth(value: number | null | undefined);
 }
 
+/**
+ * The primary result of HTML conversion and extraction.
+ *
+ * Contains the converted text output, optional structured document tree,
+ * metadata, extracted tables, images, and processing warnings.
+ *
+ * # Example
+ *
+ * ```text
+ * use html_to_markdown_rs::{convert, ConversionOptions};
+ *
+ * let result = convert("<h1>Hello</h1><p>World</p>", None)?;
+ * assert!(result.content.is_some());
+ * assert!(result.warnings.is_empty());
+ * ```
+ */
 export class WasmConversionResult {
     free(): void;
     [Symbol.dispose](): void;
@@ -172,6 +264,26 @@ export class WasmConversionResult {
     warnings: WasmProcessingWarning[];
 }
 
+/**
+ * Document-level metadata extracted from `<head>` and top-level elements.
+ *
+ * Contains all metadata typically used by search engines, social media platforms,
+ * and browsers for document indexing and presentation.
+ *
+ * # Examples
+ *
+ * ```
+ * # use html_to_markdown_rs::metadata::DocumentMetadata;
+ * let doc = DocumentMetadata {
+ *     title: Some("My Article".to_string()),
+ *     description: Some("A great article about Rust".to_string()),
+ *     keywords: vec!["rust".to_string(), "programming".to_string()],
+ *     ..Default::default()
+ * };
+ *
+ * assert_eq!(doc.title, Some("My Article".to_string()));
+ * ```
+ */
 export class WasmDocumentMetadata {
     free(): void;
     [Symbol.dispose](): void;
@@ -196,6 +308,9 @@ export class WasmDocumentMetadata {
     twitterCard: any;
 }
 
+/**
+ * A single node in the document tree.
+ */
 export class WasmDocumentNode {
     free(): void;
     [Symbol.dispose](): void;
@@ -210,6 +325,11 @@ export class WasmDocumentNode {
     set parent(value: number | null | undefined);
 }
 
+/**
+ * A structured document tree representing the semantic content of an HTML document.
+ *
+ * Uses a flat node array with index-based parent/child references for efficient traversal.
+ */
 export class WasmDocumentStructure {
     free(): void;
     [Symbol.dispose](): void;
@@ -219,6 +339,9 @@ export class WasmDocumentStructure {
     set sourceFormat(value: string | null | undefined);
 }
 
+/**
+ * A single cell in a table grid.
+ */
 export class WasmGridCell {
     free(): void;
     [Symbol.dispose](): void;
@@ -231,9 +354,61 @@ export class WasmGridCell {
     rowSpan: number;
 }
 
+/**
+ * Header element metadata with hierarchy tracking.
+ *
+ * Captures heading elements (h1-h6) with their text content, identifiers,
+ * and position in the document structure.
+ *
+ * # Examples
+ *
+ * ```
+ * # use html_to_markdown_rs::metadata::HeaderMetadata;
+ * let header = HeaderMetadata {
+ *     level: 1,
+ *     text: "Main Title".to_string(),
+ *     id: Some("main-title".to_string()),
+ *     depth: 0,
+ *     html_offset: 145,
+ * };
+ *
+ * assert_eq!(header.level, 1);
+ * assert!(header.is_valid());
+ * ```
+ */
 export class WasmHeaderMetadata {
     free(): void;
     [Symbol.dispose](): void;
+    /**
+     * Validate that the header level is within valid range (1-6).
+     *
+     * # Returns
+     *
+     * `true` if level is 1-6, `false` otherwise.
+     *
+     * # Examples
+     *
+     * ```
+     * # use html_to_markdown_rs::metadata::HeaderMetadata;
+     * let valid = HeaderMetadata {
+     *     level: 3,
+     *     text: "Title".to_string(),
+     *     id: None,
+     *     depth: 2,
+     *     html_offset: 100,
+     * };
+     * assert!(valid.is_valid());
+     *
+     * let invalid = HeaderMetadata {
+     *     level: 7,  // Invalid
+     *     text: "Title".to_string(),
+     *     id: None,
+     *     depth: 2,
+     *     html_offset: 100,
+     * };
+     * assert!(!invalid.is_valid());
+     * ```
+     */
     isValid(): boolean;
     constructor(level: number, text: string, depth: number, html_offset: number, id?: string | null);
     depth: number;
@@ -244,12 +419,22 @@ export class WasmHeaderMetadata {
     text: string;
 }
 
+/**
+ * Heading style options for Markdown output.
+ *
+ * Controls how headings (h1-h6) are rendered in the output Markdown.
+ */
 export enum WasmHeadingStyle {
     Underlined = 0,
     Atx = 1,
     AtxClosed = 2,
 }
 
+/**
+ * Highlight rendering style for `<mark>` elements.
+ *
+ * Controls how highlighted text is rendered in Markdown output.
+ */
 export enum WasmHighlightStyle {
     DoubleEqual = 0,
     Html = 1,
@@ -257,6 +442,27 @@ export enum WasmHighlightStyle {
     None = 3,
 }
 
+/**
+ * Comprehensive metadata extraction result from HTML document.
+ *
+ * Contains all extracted metadata types in a single structure,
+ * suitable for serialization and transmission across language boundaries.
+ *
+ * # Examples
+ *
+ * ```
+ * # use html_to_markdown_rs::metadata::HtmlMetadata;
+ * let metadata = HtmlMetadata {
+ *     document: Default::default(),
+ *     headers: Vec::new(),
+ *     links: Vec::new(),
+ *     images: Vec::new(),
+ *     structured_data: Vec::new(),
+ * };
+ *
+ * assert!(metadata.headers.is_empty());
+ * ```
+ */
 export class WasmHtmlMetadata {
     free(): void;
     [Symbol.dispose](): void;
@@ -268,21 +474,48 @@ export class WasmHtmlMetadata {
     structuredData: WasmStructuredData[];
 }
 
+/**
+ * Image metadata with source and dimensions.
+ *
+ * Captures `<img>` elements and inline `<svg>` elements with metadata
+ * for image analysis and optimization.
+ *
+ * # Examples
+ *
+ * ```
+ * # use html_to_markdown_rs::metadata::{ImageMetadata, ImageType};
+ * let img = ImageMetadata {
+ *     src: "https://example.com/image.jpg".to_string(),
+ *     alt: Some("An example image".to_string()),
+ *     title: Some("Example".to_string()),
+ *     dimensions: Some((800, 600)),
+ *     image_type: ImageType::External,
+ *     attributes: Default::default(),
+ * };
+ *
+ * assert_eq!(img.image_type, ImageType::External);
+ * ```
+ */
 export class WasmImageMetadata {
     free(): void;
     [Symbol.dispose](): void;
-    constructor(src: string, image_type: WasmImageType, attributes: any, alt?: string | null, title?: string | null, dimensions?: string | null);
+    constructor(src: string, image_type: WasmImageType, attributes: any, alt?: string | null, title?: string | null, dimensions?: Uint32Array | null);
     get alt(): string | undefined;
     set alt(value: string | null | undefined);
     attributes: any;
-    get dimensions(): string | undefined;
-    set dimensions(value: string | null | undefined);
+    get dimensions(): Uint32Array | undefined;
+    set dimensions(value: Uint32Array | null | undefined);
     imageType: WasmImageType;
     src: string;
     get title(): string | undefined;
     set title(value: string | null | undefined);
 }
 
+/**
+ * Image source classification for proper handling and processing.
+ *
+ * Determines whether an image is embedded (data URI), inline SVG, external, or relative.
+ */
 export enum WasmImageType {
     DataUri = 0,
     InlineSvg = 1,
@@ -290,9 +523,52 @@ export enum WasmImageType {
     Relative = 3,
 }
 
+/**
+ * Hyperlink metadata with categorization and attributes.
+ *
+ * Represents `<a>` elements with parsed href values, text content, and link type classification.
+ *
+ * # Examples
+ *
+ * ```
+ * # use html_to_markdown_rs::metadata::{LinkMetadata, LinkType};
+ * let link = LinkMetadata {
+ *     href: "https://example.com".to_string(),
+ *     text: "Example".to_string(),
+ *     title: Some("Visit Example".to_string()),
+ *     link_type: LinkType::External,
+ *     rel: vec!["nofollow".to_string()],
+ *     attributes: Default::default(),
+ * };
+ *
+ * assert_eq!(link.link_type, LinkType::External);
+ * assert_eq!(link.text, "Example");
+ * ```
+ */
 export class WasmLinkMetadata {
     free(): void;
     [Symbol.dispose](): void;
+    /**
+     * Classify a link based on href value.
+     *
+     * # Arguments
+     *
+     * * `href` - The href attribute value
+     *
+     * # Returns
+     *
+     * Appropriate [`LinkType`] based on protocol and content.
+     *
+     * # Examples
+     *
+     * ```
+     * # use html_to_markdown_rs::metadata::{LinkMetadata, LinkType};
+     * assert_eq!(LinkMetadata::classify_link("#section"), LinkType::Anchor);
+     * assert_eq!(LinkMetadata::classify_link("mailto:test@example.com"), LinkType::Email);
+     * assert_eq!(LinkMetadata::classify_link("tel:+1234567890"), LinkType::Phone);
+     * assert_eq!(LinkMetadata::classify_link("https://example.com"), LinkType::External);
+     * ```
+     */
     static classifyLink(href: string): WasmLinkType;
     constructor(href: string, text: string, link_type: WasmLinkType, rel: string[], attributes: any, title?: string | null);
     attributes: any;
@@ -304,11 +580,22 @@ export class WasmLinkMetadata {
     set title(value: string | null | undefined);
 }
 
+/**
+ * Link rendering style in Markdown output.
+ *
+ * Controls whether links and images use inline `[text](url)` syntax or
+ * reference-style `[text][1]` syntax with definitions collected at the end.
+ */
 export enum WasmLinkStyle {
     Inline = 0,
     Reference = 1,
 }
 
+/**
+ * Link classification based on href value and document context.
+ *
+ * Used to categorize links during extraction for filtering and analysis.
+ */
 export enum WasmLinkType {
     Anchor = 0,
     Internal = 1,
@@ -318,48 +605,31 @@ export enum WasmLinkType {
     Other = 5,
 }
 
+/**
+ * List indentation character type.
+ *
+ * Controls whether list items are indented with spaces or tabs.
+ */
 export enum WasmListIndentType {
     Spaces = 0,
     Tabs = 1,
 }
 
-export class WasmMetadataConfig {
-    free(): void;
-    [Symbol.dispose](): void;
-    anyEnabled(): boolean;
-    static default(): WasmMetadataConfig;
-    constructor(extract_document?: boolean | null, extract_headers?: boolean | null, extract_links?: boolean | null, extract_images?: boolean | null, extract_structured_data?: boolean | null, max_structured_data_size?: number | null);
-    extractDocument: boolean;
-    extractHeaders: boolean;
-    extractImages: boolean;
-    extractLinks: boolean;
-    extractStructuredData: boolean;
-    maxStructuredDataSize: number;
-}
-
-export class WasmMetadataConfigUpdate {
-    free(): void;
-    [Symbol.dispose](): void;
-    constructor(extract_document?: boolean | null, extract_headers?: boolean | null, extract_links?: boolean | null, extract_images?: boolean | null, extract_structured_data?: boolean | null, max_structured_data_size?: number | null);
-    get extractDocument(): boolean | undefined;
-    set extractDocument(value: boolean | null | undefined);
-    get extractHeaders(): boolean | undefined;
-    set extractHeaders(value: boolean | null | undefined);
-    get extractImages(): boolean | undefined;
-    set extractImages(value: boolean | null | undefined);
-    get extractLinks(): boolean | undefined;
-    set extractLinks(value: boolean | null | undefined);
-    get extractStructuredData(): boolean | undefined;
-    set extractStructuredData(value: boolean | null | undefined);
-    get maxStructuredDataSize(): number | undefined;
-    set maxStructuredDataSize(value: number | null | undefined);
-}
-
+/**
+ * Line break syntax in Markdown output.
+ *
+ * Controls how soft line breaks (from `<br>` or line breaks in source) are rendered.
+ */
 export enum WasmNewlineStyle {
     Spaces = 0,
     Backslash = 1,
 }
 
+/**
+ * The semantic content type of a document node.
+ *
+ * Uses internally tagged representation (`"node_type": "heading"`) for JSON serialization.
+ */
 export enum WasmNodeContent {
     Heading = 0,
     Paragraph = 1,
@@ -376,16 +646,168 @@ export enum WasmNodeContent {
     Group = 12,
 }
 
+/**
+ * Context information passed to all visitor methods.
+ *
+ * Provides comprehensive metadata about the current node being visited,
+ * including its type, attributes, position in the DOM tree, and parent context.
+ */
+export class WasmNodeContext {
+    free(): void;
+    [Symbol.dispose](): void;
+    constructor(node_type: WasmNodeType, tag_name: string, attributes: any, depth: number, index_in_parent: number, is_inline: boolean, parent_tag?: string | null);
+    attributes: any;
+    depth: number;
+    indexInParent: number;
+    isInline: boolean;
+    nodeType: WasmNodeType;
+    get parentTag(): string | undefined;
+    set parentTag(value: string | null | undefined);
+    tagName: string;
+}
+
+/**
+ * Node type enumeration covering all HTML element types.
+ *
+ * This enum categorizes all HTML elements that the converter recognizes,
+ * providing a coarse-grained classification for visitor dispatch.
+ */
+export enum WasmNodeType {
+    Text = 0,
+    Element = 1,
+    Heading = 2,
+    Paragraph = 3,
+    Div = 4,
+    Blockquote = 5,
+    Pre = 6,
+    Hr = 7,
+    List = 8,
+    ListItem = 9,
+    DefinitionList = 10,
+    DefinitionTerm = 11,
+    DefinitionDescription = 12,
+    Table = 13,
+    TableRow = 14,
+    TableCell = 15,
+    TableHeader = 16,
+    TableBody = 17,
+    TableHead = 18,
+    TableFoot = 19,
+    Link = 20,
+    Image = 21,
+    Strong = 22,
+    Em = 23,
+    Code = 24,
+    Strikethrough = 25,
+    Underline = 26,
+    Subscript = 27,
+    Superscript = 28,
+    Mark = 29,
+    Small = 30,
+    Br = 31,
+    Span = 32,
+    Article = 33,
+    Section = 34,
+    Nav = 35,
+    Aside = 36,
+    Header = 37,
+    Footer = 38,
+    Main = 39,
+    Figure = 40,
+    Figcaption = 41,
+    Time = 42,
+    Details = 43,
+    Summary = 44,
+    Form = 45,
+    Input = 46,
+    Select = 47,
+    Option = 48,
+    Button = 49,
+    Textarea = 50,
+    Label = 51,
+    Fieldset = 52,
+    Legend = 53,
+    Audio = 54,
+    Video = 55,
+    Picture = 56,
+    Source = 57,
+    Iframe = 58,
+    Svg = 59,
+    Canvas = 60,
+    Ruby = 61,
+    Rt = 62,
+    Rp = 63,
+    Abbr = 64,
+    Kbd = 65,
+    Samp = 66,
+    Var = 67,
+    Cite = 68,
+    Q = 69,
+    Del = 70,
+    Ins = 71,
+    Data = 72,
+    Meter = 73,
+    Progress = 74,
+    Output = 75,
+    Template = 76,
+    Slot = 77,
+    Html = 78,
+    Head = 79,
+    Body = 80,
+    Title = 81,
+    Meta = 82,
+    LinkTag = 83,
+    Style = 84,
+    Script = 85,
+    Base = 86,
+    Custom = 87,
+}
+
+/**
+ * Output format for conversion.
+ *
+ * Specifies the target markup language format for the conversion output.
+ */
 export enum WasmOutputFormat {
     Markdown = 0,
     Djot = 1,
     Plain = 2,
 }
 
+/**
+ * HTML preprocessing options for document cleanup before conversion.
+ */
 export class WasmPreprocessingOptions {
     free(): void;
     [Symbol.dispose](): void;
+    /**
+     * Apply a partial update to these preprocessing options.
+     *
+     * Any specified fields in the update will override the current values.
+     * Unspecified fields (None) are left unchanged.
+     *
+     * # Arguments
+     *
+     * * `update` - Partial preprocessing options update
+     */
+    applyUpdate(_update: WasmPreprocessingOptionsUpdate): void;
     static default(): WasmPreprocessingOptions;
+    static from(update: WasmPreprocessingOptionsUpdate): WasmPreprocessingOptions;
+    /**
+     * Create new preprocessing options from a partial update.
+     *
+     * Creates a new `PreprocessingOptions` struct with defaults, then applies the update.
+     * Fields not specified in the update keep their default values.
+     *
+     * # Arguments
+     *
+     * * `update` - Partial preprocessing options update
+     *
+     * # Returns
+     *
+     * New `PreprocessingOptions` with specified updates applied to defaults
+     */
+    static fromUpdate(update: WasmPreprocessingOptionsUpdate): WasmPreprocessingOptions;
     constructor(enabled?: boolean | null, preset?: WasmPreprocessingPreset | null, remove_navigation?: boolean | null, remove_forms?: boolean | null);
     enabled: boolean;
     preset: WasmPreprocessingPreset;
@@ -393,6 +815,13 @@ export class WasmPreprocessingOptions {
     removeNavigation: boolean;
 }
 
+/**
+ * Partial update for `PreprocessingOptions`.
+ *
+ * This struct uses `Option<T>` to represent optional fields that can be selectively updated.
+ * Only specified fields (Some values) will override existing options; None values leave the
+ * corresponding fields unchanged when applied via [`PreprocessingOptions::apply_update`].
+ */
 export class WasmPreprocessingOptionsUpdate {
     free(): void;
     [Symbol.dispose](): void;
@@ -407,12 +836,20 @@ export class WasmPreprocessingOptionsUpdate {
     set removeNavigation(value: boolean | null | undefined);
 }
 
+/**
+ * HTML preprocessing aggressiveness level.
+ *
+ * Controls the extent of cleanup performed before conversion. Higher levels remove more elements.
+ */
 export enum WasmPreprocessingPreset {
     Minimal = 0,
     Standard = 1,
     Aggressive = 2,
 }
 
+/**
+ * A non-fatal warning generated during HTML processing.
+ */
 export class WasmProcessingWarning {
     free(): void;
     [Symbol.dispose](): void;
@@ -421,6 +858,25 @@ export class WasmProcessingWarning {
     message: string;
 }
 
+/**
+ * Structured data block (JSON-LD, Microdata, or RDFa).
+ *
+ * Represents machine-readable structured data found in the document.
+ * JSON-LD blocks are collected as raw JSON strings for flexibility.
+ *
+ * # Examples
+ *
+ * ```
+ * # use html_to_markdown_rs::metadata::{StructuredData, StructuredDataType};
+ * let schema = StructuredData {
+ *     data_type: StructuredDataType::JsonLd,
+ *     raw_json: r#"{"@context":"https://schema.org","@type":"Article"}"#.to_string(),
+ *     schema_type: Some("Article".to_string()),
+ * };
+ *
+ * assert_eq!(schema.data_type, StructuredDataType::JsonLd);
+ * ```
+ */
 export class WasmStructuredData {
     free(): void;
     [Symbol.dispose](): void;
@@ -431,12 +887,20 @@ export class WasmStructuredData {
     set schemaType(value: string | null | undefined);
 }
 
+/**
+ * Structured data format type.
+ *
+ * Identifies the schema/format used for structured data markup.
+ */
 export enum WasmStructuredDataType {
     JsonLd = 0,
     Microdata = 1,
     RDFa = 2,
 }
 
+/**
+ * A top-level extracted table with both structured data and markdown representation.
+ */
 export class WasmTableData {
     free(): void;
     [Symbol.dispose](): void;
@@ -445,6 +909,9 @@ export class WasmTableData {
     markdown: string;
 }
 
+/**
+ * A structured table grid with cell-level data including spans.
+ */
 export class WasmTableGrid {
     free(): void;
     [Symbol.dispose](): void;
@@ -454,6 +921,11 @@ export class WasmTableGrid {
     rows: number;
 }
 
+/**
+ * An inline text annotation with byte-range offsets.
+ *
+ * Annotations describe formatting (bold, italic, etc.) and links within a node's text content.
+ */
 export class WasmTextAnnotation {
     free(): void;
     [Symbol.dispose](): void;
@@ -463,26 +935,55 @@ export class WasmTextAnnotation {
     start: number;
 }
 
+/**
+ * Text directionality of document content.
+ *
+ * Corresponds to the HTML `dir` attribute and `bdi` element directionality.
+ */
 export enum WasmTextDirection {
     LeftToRight = 0,
     RightToLeft = 1,
     Auto = 2,
 }
 
+/**
+ * Result of a visitor callback.
+ *
+ * Allows visitors to control the conversion flow by either proceeding
+ * with default behavior, providing custom output, skipping elements,
+ * preserving HTML, or signaling errors.
+ */
+export enum WasmVisitResult {
+    Continue = 0,
+    Custom = 1,
+    Skip = 2,
+    PreserveHtml = 3,
+    Error = 4,
+}
+
+/**
+ * Categories of processing warnings.
+ */
 export enum WasmWarningKind {
     ImageExtractionFailed = 0,
     EncodingFallback = 1,
     TruncatedInput = 2,
     MalformedHtml = 3,
     SanitizationApplied = 4,
+    DepthLimitExceeded = 5,
 }
 
+/**
+ * Whitespace handling strategy during conversion.
+ *
+ * Determines how sequences of whitespace characters (spaces, tabs, newlines) are processed.
+ */
 export enum WasmWhitespaceMode {
     Normalized = 0,
     Strict = 1,
 }
 
-export function convert(html: string, options?: WasmConversionOptions | null): WasmConversionResult;
+export function convert(html: string, options?: WasmConversionOptions | null, visitor?: any | null): WasmConversionResult;
 
 export type InitInput = RequestInfo | URL | Response | BufferSource | WebAssembly.Module;
 
@@ -500,14 +1001,15 @@ export interface InitOutput {
     readonly __wbg_wasmhtmlmetadata_free: (a: number, b: number) => void;
     readonly __wbg_wasmimagemetadata_free: (a: number, b: number) => void;
     readonly __wbg_wasmlinkmetadata_free: (a: number, b: number) => void;
-    readonly __wbg_wasmmetadataconfig_free: (a: number, b: number) => void;
-    readonly __wbg_wasmmetadataconfigupdate_free: (a: number, b: number) => void;
+    readonly __wbg_wasmnodecontext_free: (a: number, b: number) => void;
     readonly __wbg_wasmpreprocessingoptions_free: (a: number, b: number) => void;
     readonly __wbg_wasmprocessingwarning_free: (a: number, b: number) => void;
     readonly __wbg_wasmstructureddata_free: (a: number, b: number) => void;
     readonly __wbg_wasmtabledata_free: (a: number, b: number) => void;
     readonly __wbg_wasmtablegrid_free: (a: number, b: number) => void;
-    readonly convert: (a: number, b: number, c: number, d: number) => void;
+    readonly __wbg_wasmtextannotation_free: (a: number, b: number) => void;
+    readonly convert: (a: number, b: number, c: number, d: number, e: number) => void;
+    readonly wasmconversionoptions_applyUpdate: (a: number, b: number) => void;
     readonly wasmconversionoptions_autolinks: (a: number) => number;
     readonly wasmconversionoptions_brInTables: (a: number) => number;
     readonly wasmconversionoptions_builder: () => number;
@@ -524,8 +1026,11 @@ export interface InitOutput {
     readonly wasmconversionoptions_escapeAsterisks: (a: number) => number;
     readonly wasmconversionoptions_escapeMisc: (a: number) => number;
     readonly wasmconversionoptions_escapeUnderscores: (a: number) => number;
+    readonly wasmconversionoptions_excludeSelectors: (a: number, b: number) => void;
     readonly wasmconversionoptions_extractImages: (a: number) => number;
     readonly wasmconversionoptions_extractMetadata: (a: number) => number;
+    readonly wasmconversionoptions_from: (a: number) => number;
+    readonly wasmconversionoptions_fromUpdate: (a: number) => number;
     readonly wasmconversionoptions_headingStyle: (a: number) => number;
     readonly wasmconversionoptions_highlightStyle: (a: number) => number;
     readonly wasmconversionoptions_includeDocumentStructure: (a: number) => number;
@@ -534,8 +1039,9 @@ export interface InitOutput {
     readonly wasmconversionoptions_linkStyle: (a: number) => number;
     readonly wasmconversionoptions_listIndentType: (a: number) => number;
     readonly wasmconversionoptions_listIndentWidth: (a: number) => number;
+    readonly wasmconversionoptions_maxDepth: (a: number) => number;
     readonly wasmconversionoptions_maxImageSize: (a: number) => bigint;
-    readonly wasmconversionoptions_new: (a: number, b: number, c: number, d: number, e: number, f: number, g: number, h: number, i: number, j: number, k: number, l: number, m: number, n: number, o: number, p: number, q: number, r: number, s: number, t: number, u: number, v: number, w: number, x: number, y: number, z: number, a1: number, b1: number, c1: number, d1: number, e1: number, f1: number, g1: number, h1: number, i1: number, j1: number, k1: number, l1: number, m1: number, n1: number, o1: number, p1: number, q1: number, r1: number, s1: number, t1: bigint, u1: number, v1: number) => number;
+    readonly wasmconversionoptions_new: (a: number, b: number, c: number, d: number, e: number, f: number, g: number, h: number, i: number, j: number, k: number, l: number, m: number, n: number, o: number, p: number, q: number, r: number, s: number, t: number, u: number, v: number, w: number, x: number, y: number, z: number, a1: number, b1: number, c1: number, d1: number, e1: number, f1: number, g1: number, h1: number, i1: number, j1: number, k1: number, l1: number, m1: number, n1: number, o1: number, p1: number, q1: number, r1: number, s1: number, t1: bigint, u1: number, v1: number, w1: number, x1: number, y1: number) => number;
     readonly wasmconversionoptions_newlineStyle: (a: number) => number;
     readonly wasmconversionoptions_outputFormat: (a: number) => number;
     readonly wasmconversionoptions_preprocessing: (a: number) => number;
@@ -554,6 +1060,7 @@ export interface InitOutput {
     readonly wasmconversionoptions_set_escapeAsterisks: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_escapeMisc: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_escapeUnderscores: (a: number, b: number) => void;
+    readonly wasmconversionoptions_set_excludeSelectors: (a: number, b: number, c: number) => void;
     readonly wasmconversionoptions_set_extractImages: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_extractMetadata: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_headingStyle: (a: number, b: number) => void;
@@ -564,6 +1071,7 @@ export interface InitOutput {
     readonly wasmconversionoptions_set_linkStyle: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_listIndentType: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_listIndentWidth: (a: number, b: number) => void;
+    readonly wasmconversionoptions_set_maxDepth: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_maxImageSize: (a: number, b: bigint) => void;
     readonly wasmconversionoptions_set_newlineStyle: (a: number, b: number) => void;
     readonly wasmconversionoptions_set_outputFormat: (a: number, b: number) => void;
@@ -588,6 +1096,7 @@ export interface InitOutput {
     readonly wasmconversionoptions_wrap: (a: number) => number;
     readonly wasmconversionoptions_wrapWidth: (a: number) => number;
     readonly wasmconversionoptionsbuilder_build: (a: number) => number;
+    readonly wasmconversionoptionsbuilder_excludeSelectors: (a: number, b: number, c: number) => number;
     readonly wasmconversionoptionsbuilder_keepInlineImagesIn: (a: number, b: number, c: number) => number;
     readonly wasmconversionoptionsbuilder_preprocessing: (a: number, b: number) => number;
     readonly wasmconversionoptionsbuilder_preserveTags: (a: number, b: number, c: number) => number;
@@ -606,6 +1115,7 @@ export interface InitOutput {
     readonly wasmconversionoptionsupdate_escapeAsterisks: (a: number) => number;
     readonly wasmconversionoptionsupdate_escapeMisc: (a: number) => number;
     readonly wasmconversionoptionsupdate_escapeUnderscores: (a: number) => number;
+    readonly wasmconversionoptionsupdate_excludeSelectors: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_extractImages: (a: number) => number;
     readonly wasmconversionoptionsupdate_extractMetadata: (a: number) => number;
     readonly wasmconversionoptionsupdate_headingStyle: (a: number) => number;
@@ -616,8 +1126,9 @@ export interface InitOutput {
     readonly wasmconversionoptionsupdate_linkStyle: (a: number) => number;
     readonly wasmconversionoptionsupdate_listIndentType: (a: number) => number;
     readonly wasmconversionoptionsupdate_listIndentWidth: (a: number) => number;
+    readonly wasmconversionoptionsupdate_maxDepth: (a: number) => number;
     readonly wasmconversionoptionsupdate_maxImageSize: (a: number, b: number) => void;
-    readonly wasmconversionoptionsupdate_new: (a: number, b: number, c: number, d: number, e: number, f: number, g: number, h: number, i: number, j: number, k: number, l: number, m: number, n: number, o: number, p: number, q: number, r: number, s: number, t: number, u: number, v: number, w: number, x: number, y: number, z: number, a1: number, b1: number, c1: number, d1: number, e1: number, f1: number, g1: number, h1: number, i1: number, j1: number, k1: number, l1: number, m1: number, n1: number, o1: number, p1: number, q1: number, r1: number, s1: number, t1: bigint, u1: number, v1: number) => number;
+    readonly wasmconversionoptionsupdate_new: (a: number, b: number, c: number, d: number, e: number, f: number, g: number, h: number, i: number, j: number, k: number, l: number, m: number, n: number, o: number, p: number, q: number, r: number, s: number, t: number, u: number, v: number, w: number, x: number, y: number, z: number, a1: number, b1: number, c1: number, d1: number, e1: number, f1: number, g1: number, h1: number, i1: number, j1: number, k1: number, l1: number, m1: number, n1: number, o1: number, p1: number, q1: number, r1: number, s1: number, t1: bigint, u1: number, v1: number, w1: number, x1: number, y1: number) => number;
     readonly wasmconversionoptionsupdate_newlineStyle: (a: number) => number;
     readonly wasmconversionoptionsupdate_outputFormat: (a: number) => number;
     readonly wasmconversionoptionsupdate_preprocessing: (a: number) => number;
@@ -636,6 +1147,7 @@ export interface InitOutput {
     readonly wasmconversionoptionsupdate_set_escapeAsterisks: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_escapeMisc: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_escapeUnderscores: (a: number, b: number) => void;
+    readonly wasmconversionoptionsupdate_set_excludeSelectors: (a: number, b: number, c: number) => void;
     readonly wasmconversionoptionsupdate_set_extractImages: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_extractMetadata: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_headingStyle: (a: number, b: number) => void;
@@ -646,6 +1158,7 @@ export interface InitOutput {
     readonly wasmconversionoptionsupdate_set_linkStyle: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_listIndentType: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_listIndentWidth: (a: number, b: number) => void;
+    readonly wasmconversionoptionsupdate_set_maxDepth: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_maxImageSize: (a: number, b: number, c: bigint) => void;
     readonly wasmconversionoptionsupdate_set_newlineStyle: (a: number, b: number) => void;
     readonly wasmconversionoptionsupdate_set_outputFormat: (a: number, b: number) => void;
@@ -784,34 +1297,23 @@ export interface InitOutput {
     readonly wasmlinkmetadata_set_title: (a: number, b: number, c: number) => void;
     readonly wasmlinkmetadata_text: (a: number, b: number) => void;
     readonly wasmlinkmetadata_title: (a: number, b: number) => void;
-    readonly wasmmetadataconfig_anyEnabled: (a: number) => number;
-    readonly wasmmetadataconfig_default: () => number;
-    readonly wasmmetadataconfig_extractDocument: (a: number) => number;
-    readonly wasmmetadataconfig_extractHeaders: (a: number) => number;
-    readonly wasmmetadataconfig_extractImages: (a: number) => number;
-    readonly wasmmetadataconfig_extractLinks: (a: number) => number;
-    readonly wasmmetadataconfig_extractStructuredData: (a: number) => number;
-    readonly wasmmetadataconfig_maxStructuredDataSize: (a: number) => number;
-    readonly wasmmetadataconfig_new: (a: number, b: number, c: number, d: number, e: number, f: number) => number;
-    readonly wasmmetadataconfig_set_extractDocument: (a: number, b: number) => void;
-    readonly wasmmetadataconfig_set_extractHeaders: (a: number, b: number) => void;
-    readonly wasmmetadataconfig_set_extractImages: (a: number, b: number) => void;
-    readonly wasmmetadataconfig_set_extractLinks: (a: number, b: number) => void;
-    readonly wasmmetadataconfig_set_extractStructuredData: (a: number, b: number) => void;
-    readonly wasmmetadataconfig_set_maxStructuredDataSize: (a: number, b: number) => void;
-    readonly wasmmetadataconfigupdate_extractDocument: (a: number) => number;
-    readonly wasmmetadataconfigupdate_extractHeaders: (a: number) => number;
-    readonly wasmmetadataconfigupdate_extractImages: (a: number) => number;
-    readonly wasmmetadataconfigupdate_extractLinks: (a: number) => number;
-    readonly wasmmetadataconfigupdate_extractStructuredData: (a: number) => number;
-    readonly wasmmetadataconfigupdate_new: (a: number, b: number, c: number, d: number, e: number, f: number) => number;
-    readonly wasmmetadataconfigupdate_set_extractDocument: (a: number, b: number) => void;
-    readonly wasmmetadataconfigupdate_set_extractHeaders: (a: number, b: number) => void;
-    readonly wasmmetadataconfigupdate_set_extractImages: (a: number, b: number) => void;
-    readonly wasmmetadataconfigupdate_set_extractLinks: (a: number, b: number) => void;
-    readonly wasmmetadataconfigupdate_set_extractStructuredData: (a: number, b: number) => void;
+    readonly wasmnodecontext_attributes: (a: number) => number;
+    readonly wasmnodecontext_indexInParent: (a: number) => number;
+    readonly wasmnodecontext_isInline: (a: number) => number;
+    readonly wasmnodecontext_new: (a: number, b: number, c: number, d: number, e: number, f: number, g: number, h: number, i: number) => number;
+    readonly wasmnodecontext_nodeType: (a: number) => number;
+    readonly wasmnodecontext_parentTag: (a: number, b: number) => void;
+    readonly wasmnodecontext_set_attributes: (a: number, b: number) => void;
+    readonly wasmnodecontext_set_indexInParent: (a: number, b: number) => void;
+    readonly wasmnodecontext_set_isInline: (a: number, b: number) => void;
+    readonly wasmnodecontext_set_nodeType: (a: number, b: number) => void;
+    readonly wasmnodecontext_set_parentTag: (a: number, b: number, c: number) => void;
+    readonly wasmnodecontext_set_tagName: (a: number, b: number, c: number) => void;
+    readonly wasmnodecontext_tagName: (a: number, b: number) => void;
+    readonly wasmpreprocessingoptions_applyUpdate: (a: number, b: number) => void;
     readonly wasmpreprocessingoptions_default: () => number;
     readonly wasmpreprocessingoptions_enabled: (a: number) => number;
+    readonly wasmpreprocessingoptions_from: (a: number) => number;
     readonly wasmpreprocessingoptions_new: (a: number, b: number, c: number, d: number) => number;
     readonly wasmpreprocessingoptions_preset: (a: number) => number;
     readonly wasmpreprocessingoptions_removeForms: (a: number) => number;
@@ -854,22 +1356,23 @@ export interface InitOutput {
     readonly wasmtextannotation_new: (a: number, b: number, c: number) => number;
     readonly wasmtextannotation_set_end: (a: number, b: number) => void;
     readonly wasmtextannotation_set_kind: (a: number, b: number) => void;
+    readonly wasmtextannotation_set_start: (a: number, b: number) => void;
+    readonly wasmtextannotation_start: (a: number) => number;
+    readonly wasmpreprocessingoptions_fromUpdate: (a: number) => number;
     readonly wasmheadermetadata_set_depth: (a: number, b: number) => void;
+    readonly wasmnodecontext_set_depth: (a: number, b: number) => void;
     readonly wasmtablegrid_set_cols: (a: number, b: number) => void;
     readonly wasmtablegrid_set_rows: (a: number, b: number) => void;
-    readonly wasmtextannotation_set_start: (a: number, b: number) => void;
-    readonly wasmmetadataconfigupdate_set_maxStructuredDataSize: (a: number, b: number) => void;
     readonly wasmheadermetadata_depth: (a: number) => number;
+    readonly wasmnodecontext_depth: (a: number) => number;
     readonly wasmtablegrid_cols: (a: number) => number;
     readonly wasmtablegrid_rows: (a: number) => number;
-    readonly wasmtextannotation_start: (a: number) => number;
-    readonly wasmmetadataconfigupdate_maxStructuredDataSize: (a: number) => number;
     readonly __wbg_wasmpreprocessingoptionsupdate_free: (a: number, b: number) => void;
-    readonly __wbg_wasmtextannotation_free: (a: number, b: number) => void;
     readonly __wbindgen_export: (a: number, b: number) => number;
     readonly __wbindgen_export2: (a: number, b: number, c: number, d: number) => number;
+    readonly __wbindgen_export3: (a: number) => void;
     readonly __wbindgen_add_to_stack_pointer: (a: number) => number;
-    readonly __wbindgen_export3: (a: number, b: number, c: number) => void;
+    readonly __wbindgen_export4: (a: number, b: number, c: number) => void;
 }
 
 export type SyncInitInput = BufferSource | WebAssembly.Module;