@kreuzberg/html-to-markdown-wasm 3.6.0-rc.1 → 3.6.0-rc.2

package/README.md

@@ -126,6 +126,25 @@ const markdown = result.content;
 console.log(markdown);
 ```
 
+## Architecture
+
+The converter routes each input through one of three tiers based on a fast prescan of the byte stream:
+
+1. **Tier-1 — single-pass byte scanner.** Handles 110+ HTML tags directly. Bails on any construct it cannot prove byte-equivalent to Tier-2.
+2. **Tier-2 — DOM walker.** Picks up Tier-1 bails and inputs the classifier rejected up front.
+3. **Tier-3 — standards-conformant parser.** Engaged for malformed HTML requiring full HTML5 repair.
+
+The dispatcher is invisible to the caller. Output is byte-identical across tiers — enforced by a 116-snapshot oracle.
+
+## Capabilities
+
+- **16 languages, one Rust core.** Rust, Python, Node.js, WASM, Java, Go, C#, PHP, Ruby, Elixir, R, Dart, Kotlin (Android), Swift, Zig, C ABI.
+- **CommonMark-compatible Markdown** with GFM-style tables.
+- **Djot output**: set `output_format = "djot"` (see Djot Output Format section below).
+- **Real-HTML robust**: unclosed tags, CDATA, custom elements, malformed entities, nested tables, mixed encodings handled without losing content.
+- **Metadata extraction**, **visitor API**, **inline images**, **configurable preprocessing presets**.
+- **Per-group regression gates in CI**: every PR runs the bench harness against per-group thresholds.
+
 ## API Reference
 
 ### Core Function

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kreuzberg/html-to-markdown-wasm",
-  "version": "3.6.0-rc.1",
+  "version": "3.6.0-rc.2",
   "private": false,
   "description": "High-performance HTML to Markdown converter",
   "license": "MIT",

package/pkg/bundler/README.md

@@ -126,6 +126,25 @@ const markdown = result.content;
 console.log(markdown);
 ```
 
+## Architecture
+
+The converter routes each input through one of three tiers based on a fast prescan of the byte stream:
+
+1. **Tier-1 — single-pass byte scanner.** Handles 110+ HTML tags directly. Bails on any construct it cannot prove byte-equivalent to Tier-2.
+2. **Tier-2 — DOM walker.** Picks up Tier-1 bails and inputs the classifier rejected up front.
+3. **Tier-3 — standards-conformant parser.** Engaged for malformed HTML requiring full HTML5 repair.
+
+The dispatcher is invisible to the caller. Output is byte-identical across tiers — enforced by a 116-snapshot oracle.
+
+## Capabilities
+
+- **16 languages, one Rust core.** Rust, Python, Node.js, WASM, Java, Go, C#, PHP, Ruby, Elixir, R, Dart, Kotlin (Android), Swift, Zig, C ABI.
+- **CommonMark-compatible Markdown** with GFM-style tables.
+- **Djot output**: set `output_format = "djot"` (see Djot Output Format section below).
+- **Real-HTML robust**: unclosed tags, CDATA, custom elements, malformed entities, nested tables, mixed encodings handled without losing content.
+- **Metadata extraction**, **visitor API**, **inline images**, **configurable preprocessing presets**.
+- **Per-group regression gates in CI**: every PR runs the bench harness against per-group thresholds.
+
 ## API Reference
 
 ### Core Function

package/pkg/bundler/html_to_markdown_wasm.d.ts

@@ -40,7 +40,7 @@ export class WasmConversionOptions {
     free(): void;
     [Symbol.dispose](): void;
     static default(): WasmConversionOptions;
-    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptions | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, excludeSelectors?: string[] | null, maxDepth?: number | null);
+    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptions | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, excludeSelectors?: string[] | null, tierStrategy?: WasmTierStrategy | null, maxDepth?: number | null);
     autolinks: boolean;
     brInTables: boolean;
     bullets: string;
@@ -87,6 +87,8 @@ export class WasmConversionOptions {
     strongEmSymbol: string;
     subSymbol: string;
     supSymbol: string;
+    get tierStrategy(): string;
+    set tierStrategy(value: WasmTierStrategy);
     get urlEscapeStyle(): string;
     set urlEscapeStyle(value: WasmUrlEscapeStyle);
     get visitor(): WasmVisitorHandle | undefined;
@@ -107,7 +109,7 @@ export class WasmConversionOptionsUpdate {
     free(): void;
     [Symbol.dispose](): void;
     static default(): WasmConversionOptionsUpdate;
-    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptionsUpdate | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, maxDepth?: number | null, excludeSelectors?: string[] | null);
+    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptionsUpdate | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, maxDepth?: number | null, excludeSelectors?: string[] | null, tierStrategy?: WasmTierStrategy | null);
     get autolinks(): boolean | undefined;
     set autolinks(value: boolean | null | undefined);
     get brInTables(): boolean | undefined;
@@ -184,6 +186,8 @@ export class WasmConversionOptionsUpdate {
     set subSymbol(value: string | null | undefined);
     get supSymbol(): string | undefined;
     set supSymbol(value: string | null | undefined);
+    get tierStrategy(): string | undefined;
+    set tierStrategy(value: WasmTierStrategy | null | undefined);
     get urlEscapeStyle(): string | undefined;
     set urlEscapeStyle(value: WasmUrlEscapeStyle | null | undefined);
     get visitor(): WasmVisitorHandle | undefined;
@@ -830,6 +834,15 @@ export enum WasmTextDirection {
     Auto = 2,
 }
 
+/**
+ * Controls which conversion tier is used.
+ */
+export enum WasmTierStrategy {
+    Auto = 0,
+    Tier2 = 1,
+    Tier1 = 2,
+}
+
 /**
  * URL encoding strategy for link and image destinations.
  *

package/pkg/bundler/html_to_markdown_wasm.js

@@ -5,5 +5,5 @@ import { __wbg_set_wasm } from "./html_to_markdown_wasm_bg.js";
 __wbg_set_wasm(wasm);
 
 export {
-    WasmAnnotationKind, WasmCodeBlockStyle, WasmConversionOptions, WasmConversionOptionsUpdate, WasmConversionResult, WasmDocumentMetadata, WasmDocumentNode, WasmDocumentStructure, WasmGridCell, WasmHeaderMetadata, WasmHeadingStyle, WasmHighlightStyle, WasmHtmlMetadata, WasmImageMetadata, WasmImageType, WasmLinkMetadata, WasmLinkStyle, WasmLinkType, WasmListIndentType, WasmNewlineStyle, WasmNodeContent, WasmNodeContext, WasmNodeType, WasmOutputFormat, WasmPreprocessingOptions, WasmPreprocessingOptionsUpdate, WasmPreprocessingPreset, WasmProcessingWarning, WasmStructuredData, WasmStructuredDataType, WasmTableData, WasmTableGrid, WasmTextAnnotation, WasmTextDirection, WasmUrlEscapeStyle, WasmVisitResult, WasmVisitorHandle, WasmWarningKind, WasmWhitespaceMode, convert
+    WasmAnnotationKind, WasmCodeBlockStyle, WasmConversionOptions, WasmConversionOptionsUpdate, WasmConversionResult, WasmDocumentMetadata, WasmDocumentNode, WasmDocumentStructure, WasmGridCell, WasmHeaderMetadata, WasmHeadingStyle, WasmHighlightStyle, WasmHtmlMetadata, WasmImageMetadata, WasmImageType, WasmLinkMetadata, WasmLinkStyle, WasmLinkType, WasmListIndentType, WasmNewlineStyle, WasmNodeContent, WasmNodeContext, WasmNodeType, WasmOutputFormat, WasmPreprocessingOptions, WasmPreprocessingOptionsUpdate, WasmPreprocessingPreset, WasmProcessingWarning, WasmStructuredData, WasmStructuredDataType, WasmTableData, WasmTableGrid, WasmTextAnnotation, WasmTextDirection, WasmTierStrategy, WasmUrlEscapeStyle, WasmVisitResult, WasmVisitorHandle, WasmWarningKind, WasmWhitespaceMode, convert
 } from "./html_to_markdown_wasm_bg.js";

package/pkg/bundler/html_to_markdown_wasm_bg.js

@@ -512,9 +512,10 @@ export class WasmConversionOptions {
      * @param {boolean | null} [captureSvg]
      * @param {boolean | null} [inferDimensions]
      * @param {string[] | null} [excludeSelectors]
+     * @param {WasmTierStrategy | null} [tierStrategy]
      * @param {number | null} [maxDepth]
      */
-    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, excludeSelectors, maxDepth) {
+    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, excludeSelectors, tierStrategy, maxDepth) {
         var ptr0 = isLikeNone(bullets) ? 0 : passStringToWasm0(bullets, wasm.__wbindgen_export, wasm.__wbindgen_export2);
         var len0 = WASM_VECTOR_LEN;
         var ptr1 = isLikeNone(strongEmSymbol) ? 0 : passStringToWasm0(strongEmSymbol, wasm.__wbindgen_export, wasm.__wbindgen_export2);
@@ -540,7 +541,7 @@ export class WasmConversionOptions {
         var len9 = WASM_VECTOR_LEN;
         var ptr10 = isLikeNone(excludeSelectors) ? 0 : passArrayJsValueToWasm0(excludeSelectors, wasm.__wbindgen_export);
         var len10 = WASM_VECTOR_LEN;
-        const ret = wasm.wasmconversionoptions_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, ptr10, len10, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0);
+        const ret = wasm.wasmconversionoptions_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, ptr10, len10, isLikeNone(tierStrategy) ? 3 : tierStrategy, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0);
         this.__wbg_ptr = ret;
         WasmConversionOptionsFinalization.register(this, this.__wbg_ptr, this);
         return this;
@@ -856,6 +857,12 @@ export class WasmConversionOptions {
         const len0 = WASM_VECTOR_LEN;
         wasm.wasmconversionoptions_set_supSymbol(this.__wbg_ptr, ptr0, len0);
     }
+    /**
+     * @param {WasmTierStrategy} value
+     */
+    set tierStrategy(value) {
+        wasm.wasmconversionoptions_set_tierStrategy(this.__wbg_ptr, value);
+    }
     /**
      * @param {WasmUrlEscapeStyle} value
      */
@@ -978,6 +985,25 @@ export class WasmConversionOptions {
             wasm.__wbindgen_export4(deferred1_0, deferred1_1, 1);
         }
     }
+    /**
+     * @returns {string}
+     */
+    get tierStrategy() {
+        let deferred1_0;
+        let deferred1_1;
+        try {
+            const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+            wasm.wasmconversionoptions_tierStrategy(retptr, this.__wbg_ptr);
+            var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+            var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+            deferred1_0 = r0;
+            deferred1_1 = r1;
+            return getStringFromWasm0(r0, r1);
+        } finally {
+            wasm.__wbindgen_add_to_stack_pointer(16);
+            wasm.__wbindgen_export4(deferred1_0, deferred1_1, 1);
+        }
+    }
     /**
      * @returns {string}
      */
@@ -1436,8 +1462,9 @@ export class WasmConversionOptionsUpdate {
      * @param {boolean | null} [inferDimensions]
      * @param {number | null} [maxDepth]
      * @param {string[] | null} [excludeSelectors]
+     * @param {WasmTierStrategy | null} [tierStrategy]
      */
-    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, maxDepth, excludeSelectors) {
+    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, maxDepth, excludeSelectors, tierStrategy) {
         var ptr0 = isLikeNone(bullets) ? 0 : passStringToWasm0(bullets, wasm.__wbindgen_export, wasm.__wbindgen_export2);
         var len0 = WASM_VECTOR_LEN;
         var ptr1 = isLikeNone(strongEmSymbol) ? 0 : passStringToWasm0(strongEmSymbol, wasm.__wbindgen_export, wasm.__wbindgen_export2);
@@ -1463,7 +1490,7 @@ export class WasmConversionOptionsUpdate {
         var len9 = WASM_VECTOR_LEN;
         var ptr10 = isLikeNone(excludeSelectors) ? 0 : passArrayJsValueToWasm0(excludeSelectors, wasm.__wbindgen_export);
         var len10 = WASM_VECTOR_LEN;
-        const ret = wasm.wasmconversionoptionsupdate_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0, ptr10, len10);
+        const ret = wasm.wasmconversionoptionsupdate_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0, ptr10, len10, isLikeNone(tierStrategy) ? 3 : tierStrategy);
         this.__wbg_ptr = ret;
         WasmConversionOptionsUpdateFinalization.register(this, this.__wbg_ptr, this);
         return this;
@@ -1785,6 +1812,12 @@ export class WasmConversionOptionsUpdate {
         var len0 = WASM_VECTOR_LEN;
         wasm.wasmconversionoptionsupdate_set_supSymbol(this.__wbg_ptr, ptr0, len0);
     }
+    /**
+     * @param {WasmTierStrategy | null} [value]
+     */
+    set tierStrategy(value) {
+        wasm.wasmconversionoptionsupdate_set_tierStrategy(this.__wbg_ptr, isLikeNone(value) ? 3 : value);
+    }
     /**
      * @param {WasmUrlEscapeStyle | null} [value]
      */
@@ -1910,6 +1943,25 @@ export class WasmConversionOptionsUpdate {
             wasm.__wbindgen_add_to_stack_pointer(16);
         }
     }
+    /**
+     * @returns {string | undefined}
+     */
+    get tierStrategy() {
+        try {
+            const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+            wasm.wasmconversionoptionsupdate_tierStrategy(retptr, this.__wbg_ptr);
+            var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+            var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+            let v1;
+            if (r0 !== 0) {
+                v1 = getStringFromWasm0(r0, r1).slice();
+                wasm.__wbindgen_export4(r0, r1 * 1, 1);
+            }
+            return v1;
+        } finally {
+            wasm.__wbindgen_add_to_stack_pointer(16);
+        }
+    }
     /**
      * @returns {string | undefined}
      */
@@ -5233,6 +5285,16 @@ export const WasmTextDirection = Object.freeze({
     Auto: 2, "2": "Auto",
 });
 
+/**
+ * Controls which conversion tier is used.
+ * @enum {0 | 1 | 2}
+ */
+export const WasmTierStrategy = Object.freeze({
+    Auto: 0, "0": "Auto",
+    Tier2: 1, "1": "Tier2",
+    Tier1: 2, "2": "Tier1",
+});
+
 /**
  * URL encoding strategy for link and image destinations.
  *

package/pkg/bundler/html_to_markdown_wasm_bg.wasm.d.ts

@@ -59,7 +59,7 @@ export const wasmconversionoptions_listIndentType: (a: number, b: number) => voi
 export const wasmconversionoptions_listIndentWidth: (a: number) => number;
 export const wasmconversionoptions_maxDepth: (a: number) => number;
 export const wasmconversionoptions_maxImageSize: (a: number) => bigint;
-export const 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: number, u1: number, v1: bigint, w1: number, x1: number, y1: number, z1: number, a2: number) => number;
+export const 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: number, u1: number, v1: bigint, w1: number, x1: number, y1: number, z1: number, a2: number, b2: number) => number;
 export const wasmconversionoptions_newlineStyle: (a: number, b: number) => void;
 export const wasmconversionoptions_outputFormat: (a: number, b: number) => void;
 export const wasmconversionoptions_preprocessing: (a: number) => number;
@@ -102,6 +102,7 @@ export const wasmconversionoptions_set_stripTags: (a: number, b: number, c: numb
 export const wasmconversionoptions_set_strongEmSymbol: (a: number, b: number, c: number) => void;
 export const wasmconversionoptions_set_subSymbol: (a: number, b: number, c: number) => void;
 export const wasmconversionoptions_set_supSymbol: (a: number, b: number, c: number) => void;
+export const wasmconversionoptions_set_tierStrategy: (a: number, b: number) => void;
 export const wasmconversionoptions_set_urlEscapeStyle: (a: number, b: number) => void;
 export const wasmconversionoptions_set_visitor: (a: number, b: number) => void;
 export const wasmconversionoptions_set_whitespaceMode: (a: number, b: number) => void;
@@ -113,6 +114,7 @@ export const wasmconversionoptions_stripTags: (a: number, b: number) => void;
 export const wasmconversionoptions_strongEmSymbol: (a: number, b: number) => void;
 export const wasmconversionoptions_subSymbol: (a: number, b: number) => void;
 export const wasmconversionoptions_supSymbol: (a: number, b: number) => void;
+export const wasmconversionoptions_tierStrategy: (a: number, b: number) => void;
 export const wasmconversionoptions_urlEscapeStyle: (a: number, b: number) => void;
 export const wasmconversionoptions_visitor: (a: number) => number;
 export const wasmconversionoptions_whitespaceMode: (a: number, b: number) => void;
@@ -147,7 +149,7 @@ export const wasmconversionoptionsupdate_listIndentType: (a: number, b: number)
 export const wasmconversionoptionsupdate_listIndentWidth: (a: number) => number;
 export const wasmconversionoptionsupdate_maxDepth: (a: number) => number;
 export const wasmconversionoptionsupdate_maxImageSize: (a: number, b: number) => void;
-export const 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: number, u1: number, v1: bigint, w1: number, x1: number, y1: number, z1: number, a2: number) => number;
+export const 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: number, u1: number, v1: bigint, w1: number, x1: number, y1: number, z1: number, a2: number, b2: number) => number;
 export const wasmconversionoptionsupdate_newlineStyle: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_outputFormat: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_preprocessing: (a: number) => number;
@@ -190,6 +192,7 @@ export const wasmconversionoptionsupdate_set_stripTags: (a: number, b: number, c
 export const wasmconversionoptionsupdate_set_strongEmSymbol: (a: number, b: number, c: number) => void;
 export const wasmconversionoptionsupdate_set_subSymbol: (a: number, b: number, c: number) => void;
 export const wasmconversionoptionsupdate_set_supSymbol: (a: number, b: number, c: number) => void;
+export const wasmconversionoptionsupdate_set_tierStrategy: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_set_urlEscapeStyle: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_set_visitor: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_set_whitespaceMode: (a: number, b: number) => void;
@@ -201,6 +204,7 @@ export const wasmconversionoptionsupdate_stripTags: (a: number, b: number) => vo
 export const wasmconversionoptionsupdate_strongEmSymbol: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_subSymbol: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_supSymbol: (a: number, b: number) => void;
+export const wasmconversionoptionsupdate_tierStrategy: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_urlEscapeStyle: (a: number, b: number) => void;
 export const wasmconversionoptionsupdate_visitor: (a: number) => number;
 export const wasmconversionoptionsupdate_whitespaceMode: (a: number, b: number) => void;

package/pkg/bundler/package.json

@@ -2,7 +2,7 @@
   "name": "html-to-markdown-wasm",
   "type": "module",
   "description": "High-performance HTML to Markdown converter",
-  "version": "3.6.0-rc.1",
+  "version": "3.6.0-rc.2",
   "license": "MIT",
   "repository": {
     "type": "git",

package/pkg/deno/README.md

@@ -126,6 +126,25 @@ const markdown = result.content;
 console.log(markdown);
 ```
 
+## Architecture
+
+The converter routes each input through one of three tiers based on a fast prescan of the byte stream:
+
+1. **Tier-1 — single-pass byte scanner.** Handles 110+ HTML tags directly. Bails on any construct it cannot prove byte-equivalent to Tier-2.
+2. **Tier-2 — DOM walker.** Picks up Tier-1 bails and inputs the classifier rejected up front.
+3. **Tier-3 — standards-conformant parser.** Engaged for malformed HTML requiring full HTML5 repair.
+
+The dispatcher is invisible to the caller. Output is byte-identical across tiers — enforced by a 116-snapshot oracle.
+
+## Capabilities
+
+- **16 languages, one Rust core.** Rust, Python, Node.js, WASM, Java, Go, C#, PHP, Ruby, Elixir, R, Dart, Kotlin (Android), Swift, Zig, C ABI.
+- **CommonMark-compatible Markdown** with GFM-style tables.
+- **Djot output**: set `output_format = "djot"` (see Djot Output Format section below).
+- **Real-HTML robust**: unclosed tags, CDATA, custom elements, malformed entities, nested tables, mixed encodings handled without losing content.
+- **Metadata extraction**, **visitor API**, **inline images**, **configurable preprocessing presets**.
+- **Per-group regression gates in CI**: every PR runs the bench harness against per-group thresholds.
+
 ## API Reference
 
 ### Core Function

package/pkg/deno/html_to_markdown_wasm.d.ts

@@ -40,7 +40,7 @@ export class WasmConversionOptions {
     free(): void;
     [Symbol.dispose](): void;
     static default(): WasmConversionOptions;
-    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptions | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, excludeSelectors?: string[] | null, maxDepth?: number | null);
+    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptions | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, excludeSelectors?: string[] | null, tierStrategy?: WasmTierStrategy | null, maxDepth?: number | null);
     autolinks: boolean;
     brInTables: boolean;
     bullets: string;
@@ -87,6 +87,8 @@ export class WasmConversionOptions {
     strongEmSymbol: string;
     subSymbol: string;
     supSymbol: string;
+    get tierStrategy(): string;
+    set tierStrategy(value: WasmTierStrategy);
     get urlEscapeStyle(): string;
     set urlEscapeStyle(value: WasmUrlEscapeStyle);
     get visitor(): WasmVisitorHandle | undefined;
@@ -107,7 +109,7 @@ export class WasmConversionOptionsUpdate {
     free(): void;
     [Symbol.dispose](): void;
     static default(): WasmConversionOptionsUpdate;
-    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptionsUpdate | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, maxDepth?: number | null, excludeSelectors?: string[] | null);
+    constructor(headingStyle?: WasmHeadingStyle | null, listIndentType?: WasmListIndentType | null, listIndentWidth?: number | null, bullets?: string | null, strongEmSymbol?: string | null, escapeAsterisks?: boolean | null, escapeUnderscores?: boolean | null, escapeMisc?: boolean | null, escapeAscii?: boolean | null, codeLanguage?: string | null, autolinks?: boolean | null, defaultTitle?: boolean | null, brInTables?: boolean | null, compactTables?: boolean | null, highlightStyle?: WasmHighlightStyle | null, extractMetadata?: boolean | null, whitespaceMode?: WasmWhitespaceMode | null, stripNewlines?: boolean | null, wrap?: boolean | null, wrapWidth?: number | null, convertAsInline?: boolean | null, subSymbol?: string | null, supSymbol?: string | null, newlineStyle?: WasmNewlineStyle | null, codeBlockStyle?: WasmCodeBlockStyle | null, keepInlineImagesIn?: string[] | null, preprocessing?: WasmPreprocessingOptionsUpdate | null, encoding?: string | null, debug?: boolean | null, stripTags?: string[] | null, preserveTags?: string[] | null, skipImages?: boolean | null, urlEscapeStyle?: WasmUrlEscapeStyle | null, linkStyle?: WasmLinkStyle | null, outputFormat?: WasmOutputFormat | null, includeDocumentStructure?: boolean | null, extractImages?: boolean | null, maxImageSize?: bigint | null, captureSvg?: boolean | null, inferDimensions?: boolean | null, maxDepth?: number | null, excludeSelectors?: string[] | null, tierStrategy?: WasmTierStrategy | null);
     get autolinks(): boolean | undefined;
     set autolinks(value: boolean | null | undefined);
     get brInTables(): boolean | undefined;
@@ -184,6 +186,8 @@ export class WasmConversionOptionsUpdate {
     set subSymbol(value: string | null | undefined);
     get supSymbol(): string | undefined;
     set supSymbol(value: string | null | undefined);
+    get tierStrategy(): string | undefined;
+    set tierStrategy(value: WasmTierStrategy | null | undefined);
     get urlEscapeStyle(): string | undefined;
     set urlEscapeStyle(value: WasmUrlEscapeStyle | null | undefined);
     get visitor(): WasmVisitorHandle | undefined;
@@ -830,6 +834,15 @@ export enum WasmTextDirection {
     Auto = 2,
 }
 
+/**
+ * Controls which conversion tier is used.
+ */
+export enum WasmTierStrategy {
+    Auto = 0,
+    Tier2 = 1,
+    Tier1 = 2,
+}
+
 /**
  * URL encoding strategy for link and image destinations.
  *

package/pkg/deno/html_to_markdown_wasm.js

@@ -514,9 +514,10 @@ export class WasmConversionOptions {
      * @param {boolean | null} [captureSvg]
      * @param {boolean | null} [inferDimensions]
      * @param {string[] | null} [excludeSelectors]
+     * @param {WasmTierStrategy | null} [tierStrategy]
      * @param {number | null} [maxDepth]
      */
-    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, excludeSelectors, maxDepth) {
+    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, excludeSelectors, tierStrategy, maxDepth) {
         var ptr0 = isLikeNone(bullets) ? 0 : passStringToWasm0(bullets, wasm.__wbindgen_export, wasm.__wbindgen_export2);
         var len0 = WASM_VECTOR_LEN;
         var ptr1 = isLikeNone(strongEmSymbol) ? 0 : passStringToWasm0(strongEmSymbol, wasm.__wbindgen_export, wasm.__wbindgen_export2);
@@ -542,7 +543,7 @@ export class WasmConversionOptions {
         var len9 = WASM_VECTOR_LEN;
         var ptr10 = isLikeNone(excludeSelectors) ? 0 : passArrayJsValueToWasm0(excludeSelectors, wasm.__wbindgen_export);
         var len10 = WASM_VECTOR_LEN;
-        const ret = wasm.wasmconversionoptions_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, ptr10, len10, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0);
+        const ret = wasm.wasmconversionoptions_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, ptr10, len10, isLikeNone(tierStrategy) ? 3 : tierStrategy, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0);
         this.__wbg_ptr = ret;
         WasmConversionOptionsFinalization.register(this, this.__wbg_ptr, this);
         return this;
@@ -858,6 +859,12 @@ export class WasmConversionOptions {
         const len0 = WASM_VECTOR_LEN;
         wasm.wasmconversionoptions_set_supSymbol(this.__wbg_ptr, ptr0, len0);
     }
+    /**
+     * @param {WasmTierStrategy} value
+     */
+    set tierStrategy(value) {
+        wasm.wasmconversionoptions_set_tierStrategy(this.__wbg_ptr, value);
+    }
     /**
      * @param {WasmUrlEscapeStyle} value
      */
@@ -980,6 +987,25 @@ export class WasmConversionOptions {
             wasm.__wbindgen_export4(deferred1_0, deferred1_1, 1);
         }
     }
+    /**
+     * @returns {string}
+     */
+    get tierStrategy() {
+        let deferred1_0;
+        let deferred1_1;
+        try {
+            const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+            wasm.wasmconversionoptions_tierStrategy(retptr, this.__wbg_ptr);
+            var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+            var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+            deferred1_0 = r0;
+            deferred1_1 = r1;
+            return getStringFromWasm0(r0, r1);
+        } finally {
+            wasm.__wbindgen_add_to_stack_pointer(16);
+            wasm.__wbindgen_export4(deferred1_0, deferred1_1, 1);
+        }
+    }
     /**
      * @returns {string}
      */
@@ -1438,8 +1464,9 @@ export class WasmConversionOptionsUpdate {
      * @param {boolean | null} [inferDimensions]
      * @param {number | null} [maxDepth]
      * @param {string[] | null} [excludeSelectors]
+     * @param {WasmTierStrategy | null} [tierStrategy]
      */
-    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, maxDepth, excludeSelectors) {
+    constructor(headingStyle, listIndentType, listIndentWidth, bullets, strongEmSymbol, escapeAsterisks, escapeUnderscores, escapeMisc, escapeAscii, codeLanguage, autolinks, defaultTitle, brInTables, compactTables, highlightStyle, extractMetadata, whitespaceMode, stripNewlines, wrap, wrapWidth, convertAsInline, subSymbol, supSymbol, newlineStyle, codeBlockStyle, keepInlineImagesIn, preprocessing, encoding, debug, stripTags, preserveTags, skipImages, urlEscapeStyle, linkStyle, outputFormat, includeDocumentStructure, extractImages, maxImageSize, captureSvg, inferDimensions, maxDepth, excludeSelectors, tierStrategy) {
         var ptr0 = isLikeNone(bullets) ? 0 : passStringToWasm0(bullets, wasm.__wbindgen_export, wasm.__wbindgen_export2);
         var len0 = WASM_VECTOR_LEN;
         var ptr1 = isLikeNone(strongEmSymbol) ? 0 : passStringToWasm0(strongEmSymbol, wasm.__wbindgen_export, wasm.__wbindgen_export2);
@@ -1465,7 +1492,7 @@ export class WasmConversionOptionsUpdate {
         var len9 = WASM_VECTOR_LEN;
         var ptr10 = isLikeNone(excludeSelectors) ? 0 : passArrayJsValueToWasm0(excludeSelectors, wasm.__wbindgen_export);
         var len10 = WASM_VECTOR_LEN;
-        const ret = wasm.wasmconversionoptionsupdate_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0, ptr10, len10);
+        const ret = wasm.wasmconversionoptionsupdate_new(isLikeNone(headingStyle) ? 3 : headingStyle, isLikeNone(listIndentType) ? 2 : listIndentType, isLikeNone(listIndentWidth) ? Number.MAX_SAFE_INTEGER : (listIndentWidth) >>> 0, ptr0, len0, ptr1, len1, isLikeNone(escapeAsterisks) ? 0xFFFFFF : escapeAsterisks ? 1 : 0, isLikeNone(escapeUnderscores) ? 0xFFFFFF : escapeUnderscores ? 1 : 0, isLikeNone(escapeMisc) ? 0xFFFFFF : escapeMisc ? 1 : 0, isLikeNone(escapeAscii) ? 0xFFFFFF : escapeAscii ? 1 : 0, ptr2, len2, isLikeNone(autolinks) ? 0xFFFFFF : autolinks ? 1 : 0, isLikeNone(defaultTitle) ? 0xFFFFFF : defaultTitle ? 1 : 0, isLikeNone(brInTables) ? 0xFFFFFF : brInTables ? 1 : 0, isLikeNone(compactTables) ? 0xFFFFFF : compactTables ? 1 : 0, isLikeNone(highlightStyle) ? 4 : highlightStyle, isLikeNone(extractMetadata) ? 0xFFFFFF : extractMetadata ? 1 : 0, isLikeNone(whitespaceMode) ? 2 : whitespaceMode, isLikeNone(stripNewlines) ? 0xFFFFFF : stripNewlines ? 1 : 0, isLikeNone(wrap) ? 0xFFFFFF : wrap ? 1 : 0, isLikeNone(wrapWidth) ? Number.MAX_SAFE_INTEGER : (wrapWidth) >>> 0, isLikeNone(convertAsInline) ? 0xFFFFFF : convertAsInline ? 1 : 0, ptr3, len3, ptr4, len4, isLikeNone(newlineStyle) ? 2 : newlineStyle, isLikeNone(codeBlockStyle) ? 3 : codeBlockStyle, ptr5, len5, ptr6, ptr7, len7, isLikeNone(debug) ? 0xFFFFFF : debug ? 1 : 0, ptr8, len8, ptr9, len9, isLikeNone(skipImages) ? 0xFFFFFF : skipImages ? 1 : 0, isLikeNone(urlEscapeStyle) ? 2 : urlEscapeStyle, isLikeNone(linkStyle) ? 2 : linkStyle, isLikeNone(outputFormat) ? 3 : outputFormat, isLikeNone(includeDocumentStructure) ? 0xFFFFFF : includeDocumentStructure ? 1 : 0, isLikeNone(extractImages) ? 0xFFFFFF : extractImages ? 1 : 0, !isLikeNone(maxImageSize), isLikeNone(maxImageSize) ? BigInt(0) : maxImageSize, isLikeNone(captureSvg) ? 0xFFFFFF : captureSvg ? 1 : 0, isLikeNone(inferDimensions) ? 0xFFFFFF : inferDimensions ? 1 : 0, isLikeNone(maxDepth) ? Number.MAX_SAFE_INTEGER : (maxDepth) >>> 0, ptr10, len10, isLikeNone(tierStrategy) ? 3 : tierStrategy);
         this.__wbg_ptr = ret;
         WasmConversionOptionsUpdateFinalization.register(this, this.__wbg_ptr, this);
         return this;
@@ -1787,6 +1814,12 @@ export class WasmConversionOptionsUpdate {
         var len0 = WASM_VECTOR_LEN;
         wasm.wasmconversionoptionsupdate_set_supSymbol(this.__wbg_ptr, ptr0, len0);
     }
+    /**
+     * @param {WasmTierStrategy | null} [value]
+     */
+    set tierStrategy(value) {
+        wasm.wasmconversionoptionsupdate_set_tierStrategy(this.__wbg_ptr, isLikeNone(value) ? 3 : value);
+    }
     /**
      * @param {WasmUrlEscapeStyle | null} [value]
      */
@@ -1912,6 +1945,25 @@ export class WasmConversionOptionsUpdate {
             wasm.__wbindgen_add_to_stack_pointer(16);
         }
     }
+    /**
+     * @returns {string | undefined}
+     */
+    get tierStrategy() {
+        try {
+            const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+            wasm.wasmconversionoptionsupdate_tierStrategy(retptr, this.__wbg_ptr);
+            var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+            var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+            let v1;
+            if (r0 !== 0) {
+                v1 = getStringFromWasm0(r0, r1).slice();
+                wasm.__wbindgen_export4(r0, r1 * 1, 1);
+            }
+            return v1;
+        } finally {
+            wasm.__wbindgen_add_to_stack_pointer(16);
+        }
+    }
     /**
      * @returns {string | undefined}
      */
@@ -5235,6 +5287,16 @@ export const WasmTextDirection = Object.freeze({
     Auto: 2, "2": "Auto",
 });
 
+/**
+ * Controls which conversion tier is used.
+ * @enum {0 | 1 | 2}
+ */
+export const WasmTierStrategy = Object.freeze({
+    Auto: 0, "0": "Auto",
+    Tier2: 1, "1": "Tier2",
+    Tier1: 2, "2": "Tier1",
+});
+
 /**
  * URL encoding strategy for link and image destinations.
  *