html-to-markdown-wasm 2.14.1 → 2.14.3

package/README.md

@@ -17,7 +17,7 @@ Runs anywhere: Node.js, Deno, Bun, browsers, and edge runtimes.
 [![RubyGems](https://badge.fury.io/rb/html-to-markdown.svg)](https://rubygems.org/gems/html-to-markdown)
 [![NuGet](https://img.shields.io/nuget/v/Goldziher.HtmlToMarkdown.svg)](https://www.nuget.org/packages/Goldziher.HtmlToMarkdown/)
 [![Maven Central](https://img.shields.io/maven-central/v/io.github.goldziher/html-to-markdown.svg)](https://central.sonatype.com/artifact/io.github.goldziher/html-to-markdown)
-[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown)
+[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Goldziher/html-to-markdown/blob/main/LICENSE)
 
 ## Performance

package/dist/README.md

@@ -11,7 +11,7 @@ High-performance HTML → Markdown conversion powered by Rust. Shipping as a Rus
 [![Hex.pm](https://img.shields.io/hexpm/v/html_to_markdown.svg)](https://hex.pm/packages/html_to_markdown)
 [![NuGet](https://img.shields.io/nuget/v/Goldziher.HtmlToMarkdown.svg)](https://www.nuget.org/packages/Goldziher.HtmlToMarkdown/)
 [![Maven Central](https://img.shields.io/maven-central/v/io.github.goldziher/html-to-markdown.svg)](https://central.sonatype.com/artifact/io.github.goldziher/html-to-markdown)
-[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown)
+[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Goldziher/html-to-markdown/blob/main/LICENSE)
 [![Discord](https://img.shields.io/badge/Discord-Join%20our%20community-7289da)](https://discord.gg/pXxagNK2zN)
 
@@ -377,7 +377,7 @@ import (
 	"fmt"
 	"log"
 
-	"github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown"
+	"github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown"
 )
 
 func main() {

package/dist/html_to_markdown_wasm.d.ts

@@ -46,6 +46,23 @@ export class WasmInlineImageWarning {
   readonly message: string;
 }
 
+export class WasmMetadataConfig {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Create a new metadata configuration with defaults
+   *
+   * All extraction types enabled by default with 1MB structured data limit
+   */
+  constructor();
+  extract_links: boolean;
+  extract_images: boolean;
+  extract_headers: boolean;
+  extract_document: boolean;
+  extract_structured_data: boolean;
+  max_structured_data_size: number;
+}
+
 /**
  * Convert HTML to Markdown
  *
@@ -70,10 +87,56 @@ export function convertBytes(html: Uint8Array, options?: WasmConversionOptions |
 
 export function convertBytesWithInlineImages(html: Uint8Array, options?: WasmConversionOptions | null, image_config?: WasmInlineImageConfig | null): WasmHtmlExtraction;
 
+/**
+ * Convert HTML bytes to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML bytes to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ */
+export function convertBytesWithMetadata(html: Uint8Array, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function convertBytesWithOptionsHandle(html: Uint8Array, handle: WasmConversionOptionsHandle): string;
 
 export function convertWithInlineImages(html: string, options?: WasmConversionOptions | null, image_config?: WasmInlineImageConfig | null): WasmHtmlExtraction;
 
+/**
+ * Convert HTML to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithMetadata, WasmMetadataConfig } from 'html-to-markdown-wasm';
+ *
+ * const html = '<h1>Hello World</h1><a href="https://example.com">Link</a>';
+ * const config = new WasmMetadataConfig();
+ * config.extractHeaders = true;
+ * config.extractLinks = true;
+ *
+ * const result = convertWithMetadata(html, null, config);
+ * console.log(result.markdown); // # Hello World\n\n[Link](https://example.com)
+ * console.log(result.metadata.headers); // [{ level: 1, text: "Hello World", ... }]
+ * console.log(result.metadata.links); // [{ href: "https://example.com", text: "Link", ... }]
+ * ```
+ */
+export function convertWithMetadata(html: string, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function convertWithOptionsHandle(html: string, handle: WasmConversionOptionsHandle): string;
 
 export function createConversionOptionsHandle(options?: WasmConversionOptions | null): WasmConversionOptionsHandle;

package/dist/html_to_markdown_wasm_bg.js

@@ -253,6 +253,10 @@ const WasmInlineImageWarningFinalization = (typeof FinalizationRegistry === 'und
     ? { register: () => {}, unregister: () => {} }
     : new FinalizationRegistry(ptr => wasm.__wbg_wasminlineimagewarning_free(ptr >>> 0, 1));
 
+const WasmMetadataConfigFinalization = (typeof FinalizationRegistry === 'undefined')
+    ? { register: () => {}, unregister: () => {} }
+    : new FinalizationRegistry(ptr => wasm.__wbg_wasmmetadataconfig_free(ptr >>> 0, 1));
+
 export class WasmConversionOptionsHandle {
     static __wrap(ptr) {
         ptr = ptr >>> 0;
@@ -598,6 +602,112 @@ export class WasmInlineImageWarning {
 }
 if (Symbol.dispose) WasmInlineImageWarning.prototype[Symbol.dispose] = WasmInlineImageWarning.prototype.free;
 
+/**
+ * Metadata extraction configuration
+ */
+export class WasmMetadataConfig {
+    __destroy_into_raw() {
+        const ptr = this.__wbg_ptr;
+        this.__wbg_ptr = 0;
+        WasmMetadataConfigFinalization.unregister(this);
+        return ptr;
+    }
+    free() {
+        const ptr = this.__destroy_into_raw();
+        wasm.__wbg_wasmmetadataconfig_free(ptr, 0);
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_links() {
+        const ret = wasm.wasmmetadataconfig_extract_links(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_images() {
+        const ret = wasm.wasmmetadataconfig_extract_images(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_headers() {
+        const ret = wasm.wasmmetadataconfig_extract_headers(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_document() {
+        const ret = wasm.wasmmetadataconfig_extract_document(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_links(value) {
+        wasm.wasmmetadataconfig_set_extract_links(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_images(value) {
+        wasm.wasmmetadataconfig_set_extract_images(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_headers(value) {
+        wasm.wasmmetadataconfig_set_extract_headers(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_document(value) {
+        wasm.wasmmetadataconfig_set_extract_document(this.__wbg_ptr, value);
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_structured_data() {
+        const ret = wasm.wasmmetadataconfig_extract_structured_data(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {number}
+     */
+    get max_structured_data_size() {
+        const ret = wasm.wasmmetadataconfig_max_structured_data_size(this.__wbg_ptr);
+        return ret >>> 0;
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_structured_data(value) {
+        wasm.wasmmetadataconfig_set_extract_structured_data(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {number} value
+     */
+    set max_structured_data_size(value) {
+        wasm.wasmmetadataconfig_set_max_structured_data_size(this.__wbg_ptr, value);
+    }
+    /**
+     * Create a new metadata configuration with defaults
+     *
+     * All extraction types enabled by default with 1MB structured data limit
+     */
+    constructor() {
+        const ret = wasm.wasmmetadataconfig_new();
+        this.__wbg_ptr = ret >>> 0;
+        WasmMetadataConfigFinalization.register(this, this.__wbg_ptr, this);
+        return this;
+    }
+}
+if (Symbol.dispose) WasmMetadataConfig.prototype[Symbol.dispose] = WasmMetadataConfig.prototype.free;
+
 /**
  * Convert HTML to Markdown
  *
@@ -703,6 +813,44 @@ export function convertBytesWithInlineImages(html, options, image_config) {
     }
 }
 
+/**
+ * Convert HTML bytes to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML bytes to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ * @param {Uint8Array} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+export function convertBytesWithMetadata(html, options, metadata_config) {
+    try {
+        const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+        let ptr0 = 0;
+        if (!isLikeNone(metadata_config)) {
+            _assertClass(metadata_config, WasmMetadataConfig);
+            ptr0 = metadata_config.__destroy_into_raw();
+        }
+        wasm.convertBytesWithMetadata(retptr, addHeapObject(html), addHeapObject(options), ptr0);
+        var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+        var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+        var r2 = getDataViewMemory0().getInt32(retptr + 4 * 2, true);
+        if (r2) {
+            throw takeObject(r1);
+        }
+        return takeObject(r0);
+    } finally {
+        wasm.__wbindgen_add_to_stack_pointer(16);
+    }
+}
+
 /**
  * @param {Uint8Array} html
  * @param {WasmConversionOptionsHandle} handle
@@ -763,6 +911,62 @@ export function convertWithInlineImages(html, options, image_config) {
     }
 }
 
+/**
+ * Convert HTML to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithMetadata, WasmMetadataConfig } from 'html-to-markdown-wasm';
+ *
+ * const html = '<h1>Hello World</h1><a href="https://example.com">Link</a>';
+ * const config = new WasmMetadataConfig();
+ * config.extractHeaders = true;
+ * config.extractLinks = true;
+ *
+ * const result = convertWithMetadata(html, null, config);
+ * console.log(result.markdown); // # Hello World\n\n[Link](https://example.com)
+ * console.log(result.metadata.headers); // [{ level: 1, text: "Hello World", ... }]
+ * console.log(result.metadata.links); // [{ href: "https://example.com", text: "Link", ... }]
+ * ```
+ * @param {string} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+export function convertWithMetadata(html, options, metadata_config) {
+    try {
+        const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+        const ptr0 = passStringToWasm0(html, wasm.__wbindgen_export, wasm.__wbindgen_export2);
+        const len0 = WASM_VECTOR_LEN;
+        let ptr1 = 0;
+        if (!isLikeNone(metadata_config)) {
+            _assertClass(metadata_config, WasmMetadataConfig);
+            ptr1 = metadata_config.__destroy_into_raw();
+        }
+        wasm.convertWithMetadata(retptr, ptr0, len0, addHeapObject(options), ptr1);
+        var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+        var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+        var r2 = getDataViewMemory0().getInt32(retptr + 4 * 2, true);
+        if (r2) {
+            throw takeObject(r1);
+        }
+        return takeObject(r0);
+    } finally {
+        wasm.__wbindgen_add_to_stack_pointer(16);
+    }
+}
+
 /**
  * @param {string} html
  * @param {WasmConversionOptionsHandle} handle
@@ -1048,6 +1252,11 @@ export function __wbg_new_1ba21ce319a06297() {
     return addHeapObject(ret);
 };
 
+export function __wbg_new_25f239778d6112b9() {
+    const ret = new Array();
+    return addHeapObject(ret);
+};
+
 export function __wbg_new_6421f6084cc5bc5a(arg0) {
     const ret = new Uint8Array(getObject(arg0));
     return addHeapObject(ret);
@@ -1086,6 +1295,15 @@ export function __wbg_set_3f1d0b984ed272ed(arg0, arg1, arg2) {
     getObject(arg0)[takeObject(arg1)] = takeObject(arg2);
 };
 
+export function __wbg_set_781438a03c0c3c81() { return handleError(function (arg0, arg1, arg2) {
+    const ret = Reflect.set(getObject(arg0), getObject(arg1), getObject(arg2));
+    return ret;
+}, arguments) };
+
+export function __wbg_set_7df433eea03a5c14(arg0, arg1, arg2) {
+    getObject(arg0)[arg1 >>> 0] = takeObject(arg2);
+};
+
 export function __wbg_set_efaaf145b9377369(arg0, arg1, arg2) {
     const ret = getObject(arg0).set(getObject(arg1), getObject(arg2));
     return addHeapObject(ret);
@@ -1126,6 +1344,12 @@ export function __wbindgen_cast_4625c577ab2ec9ee(arg0) {
     return addHeapObject(ret);
 };
 
+export function __wbindgen_cast_d6cd19b81560fd6e(arg0) {
+    // Cast intrinsic for `F64 -> Externref`.
+    const ret = arg0;
+    return addHeapObject(ret);
+};
+
 export function __wbindgen_object_clone_ref(arg0) {
     const ret = getObject(arg0);
     return addHeapObject(ret);

package/dist/html_to_markdown_wasm_bg.wasm.d.ts

@@ -6,11 +6,14 @@ export const __wbg_wasmhtmlextraction_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimage_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimageconfig_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimagewarning_free: (a: number, b: number) => void;
+export const __wbg_wasmmetadataconfig_free: (a: number, b: number) => void;
 export const convert: (a: number, b: number, c: number, d: number) => void;
 export const convertBytes: (a: number, b: number, c: number) => void;
 export const convertBytesWithInlineImages: (a: number, b: number, c: number, d: number) => void;
+export const convertBytesWithMetadata: (a: number, b: number, c: number, d: number) => void;
 export const convertBytesWithOptionsHandle: (a: number, b: number, c: number) => void;
 export const convertWithInlineImages: (a: number, b: number, c: number, d: number, e: number) => void;
+export const convertWithMetadata: (a: number, b: number, c: number, d: number, e: number) => void;
 export const convertWithOptionsHandle: (a: number, b: number, c: number, d: number) => void;
 export const createConversionOptionsHandle: (a: number, b: number) => void;
 export const wasmconversionoptionshandle_new: (a: number, b: number) => void;
@@ -30,6 +33,19 @@ export const wasminlineimageconfig_set_filenamePrefix: (a: number, b: number, c:
 export const wasminlineimageconfig_set_inferDimensions: (a: number, b: number) => void;
 export const wasminlineimagewarning_index: (a: number) => number;
 export const wasminlineimagewarning_message: (a: number, b: number) => void;
+export const wasmmetadataconfig_extract_document: (a: number) => number;
+export const wasmmetadataconfig_extract_headers: (a: number) => number;
+export const wasmmetadataconfig_extract_images: (a: number) => number;
+export const wasmmetadataconfig_extract_links: (a: number) => number;
+export const wasmmetadataconfig_extract_structured_data: (a: number) => number;
+export const wasmmetadataconfig_max_structured_data_size: (a: number) => number;
+export const wasmmetadataconfig_new: () => number;
+export const wasmmetadataconfig_set_extract_document: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_headers: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_images: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_links: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_structured_data: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_max_structured_data_size: (a: number, b: number) => void;
 export const init: () => void;
 export const __wbindgen_export: (a: number, b: number) => number;
 export const __wbindgen_export2: (a: number, b: number, c: number, d: number) => number;

package/dist/package.json

@@ -4,7 +4,7 @@
   "collaborators": [
     "Na'aman Hirschfeld <nhirschfeld@gmail.com>"
   ],
-  "version": "2.14.1",
+  "version": "2.14.3",
   "license": "MIT",
   "repository": {
     "type": "git",

package/dist-node/README.md

@@ -11,7 +11,7 @@ High-performance HTML → Markdown conversion powered by Rust. Shipping as a Rus
 [![Hex.pm](https://img.shields.io/hexpm/v/html_to_markdown.svg)](https://hex.pm/packages/html_to_markdown)
 [![NuGet](https://img.shields.io/nuget/v/Goldziher.HtmlToMarkdown.svg)](https://www.nuget.org/packages/Goldziher.HtmlToMarkdown/)
 [![Maven Central](https://img.shields.io/maven-central/v/io.github.goldziher/html-to-markdown.svg)](https://central.sonatype.com/artifact/io.github.goldziher/html-to-markdown)
-[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown)
+[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Goldziher/html-to-markdown/blob/main/LICENSE)
 [![Discord](https://img.shields.io/badge/Discord-Join%20our%20community-7289da)](https://discord.gg/pXxagNK2zN)
 
@@ -377,7 +377,7 @@ import (
 	"fmt"
 	"log"
 
-	"github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown"
+	"github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown"
 )
 
 func main() {

package/dist-node/html_to_markdown_wasm.d.ts

@@ -46,6 +46,23 @@ export class WasmInlineImageWarning {
   readonly message: string;
 }
 
+export class WasmMetadataConfig {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Create a new metadata configuration with defaults
+   *
+   * All extraction types enabled by default with 1MB structured data limit
+   */
+  constructor();
+  extract_links: boolean;
+  extract_images: boolean;
+  extract_headers: boolean;
+  extract_document: boolean;
+  extract_structured_data: boolean;
+  max_structured_data_size: number;
+}
+
 /**
  * Convert HTML to Markdown
  *
@@ -70,10 +87,56 @@ export function convertBytes(html: Uint8Array, options?: WasmConversionOptions |
 
 export function convertBytesWithInlineImages(html: Uint8Array, options?: WasmConversionOptions | null, image_config?: WasmInlineImageConfig | null): WasmHtmlExtraction;
 
+/**
+ * Convert HTML bytes to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML bytes to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ */
+export function convertBytesWithMetadata(html: Uint8Array, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function convertBytesWithOptionsHandle(html: Uint8Array, handle: WasmConversionOptionsHandle): string;
 
 export function convertWithInlineImages(html: string, options?: WasmConversionOptions | null, image_config?: WasmInlineImageConfig | null): WasmHtmlExtraction;
 
+/**
+ * Convert HTML to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithMetadata, WasmMetadataConfig } from 'html-to-markdown-wasm';
+ *
+ * const html = '<h1>Hello World</h1><a href="https://example.com">Link</a>';
+ * const config = new WasmMetadataConfig();
+ * config.extractHeaders = true;
+ * config.extractLinks = true;
+ *
+ * const result = convertWithMetadata(html, null, config);
+ * console.log(result.markdown); // # Hello World\n\n[Link](https://example.com)
+ * console.log(result.metadata.headers); // [{ level: 1, text: "Hello World", ... }]
+ * console.log(result.metadata.links); // [{ href: "https://example.com", text: "Link", ... }]
+ * ```
+ */
+export function convertWithMetadata(html: string, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function convertWithOptionsHandle(html: string, handle: WasmConversionOptionsHandle): string;
 
 export function createConversionOptionsHandle(options?: WasmConversionOptions | null): WasmConversionOptionsHandle;

package/dist-node/html_to_markdown_wasm.js

@@ -244,6 +244,10 @@ const WasmInlineImageWarningFinalization = (typeof FinalizationRegistry === 'und
     ? { register: () => {}, unregister: () => {} }
     : new FinalizationRegistry(ptr => wasm.__wbg_wasminlineimagewarning_free(ptr >>> 0, 1));
 
+const WasmMetadataConfigFinalization = (typeof FinalizationRegistry === 'undefined')
+    ? { register: () => {}, unregister: () => {} }
+    : new FinalizationRegistry(ptr => wasm.__wbg_wasmmetadataconfig_free(ptr >>> 0, 1));
+
 class WasmConversionOptionsHandle {
     static __wrap(ptr) {
         ptr = ptr >>> 0;
@@ -594,6 +598,113 @@ class WasmInlineImageWarning {
 if (Symbol.dispose) WasmInlineImageWarning.prototype[Symbol.dispose] = WasmInlineImageWarning.prototype.free;
 exports.WasmInlineImageWarning = WasmInlineImageWarning;
 
+/**
+ * Metadata extraction configuration
+ */
+class WasmMetadataConfig {
+    __destroy_into_raw() {
+        const ptr = this.__wbg_ptr;
+        this.__wbg_ptr = 0;
+        WasmMetadataConfigFinalization.unregister(this);
+        return ptr;
+    }
+    free() {
+        const ptr = this.__destroy_into_raw();
+        wasm.__wbg_wasmmetadataconfig_free(ptr, 0);
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_links() {
+        const ret = wasm.wasmmetadataconfig_extract_links(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_images() {
+        const ret = wasm.wasmmetadataconfig_extract_images(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_headers() {
+        const ret = wasm.wasmmetadataconfig_extract_headers(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_document() {
+        const ret = wasm.wasmmetadataconfig_extract_document(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_links(value) {
+        wasm.wasmmetadataconfig_set_extract_links(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_images(value) {
+        wasm.wasmmetadataconfig_set_extract_images(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_headers(value) {
+        wasm.wasmmetadataconfig_set_extract_headers(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_document(value) {
+        wasm.wasmmetadataconfig_set_extract_document(this.__wbg_ptr, value);
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_structured_data() {
+        const ret = wasm.wasmmetadataconfig_extract_structured_data(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {number}
+     */
+    get max_structured_data_size() {
+        const ret = wasm.wasmmetadataconfig_max_structured_data_size(this.__wbg_ptr);
+        return ret >>> 0;
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_structured_data(value) {
+        wasm.wasmmetadataconfig_set_extract_structured_data(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {number} value
+     */
+    set max_structured_data_size(value) {
+        wasm.wasmmetadataconfig_set_max_structured_data_size(this.__wbg_ptr, value);
+    }
+    /**
+     * Create a new metadata configuration with defaults
+     *
+     * All extraction types enabled by default with 1MB structured data limit
+     */
+    constructor() {
+        const ret = wasm.wasmmetadataconfig_new();
+        this.__wbg_ptr = ret >>> 0;
+        WasmMetadataConfigFinalization.register(this, this.__wbg_ptr, this);
+        return this;
+    }
+}
+if (Symbol.dispose) WasmMetadataConfig.prototype[Symbol.dispose] = WasmMetadataConfig.prototype.free;
+exports.WasmMetadataConfig = WasmMetadataConfig;
+
 /**
  * Convert HTML to Markdown
  *
@@ -702,6 +813,45 @@ function convertBytesWithInlineImages(html, options, image_config) {
 }
 exports.convertBytesWithInlineImages = convertBytesWithInlineImages;
 
+/**
+ * Convert HTML bytes to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML bytes to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ * @param {Uint8Array} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+function convertBytesWithMetadata(html, options, metadata_config) {
+    try {
+        const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+        let ptr0 = 0;
+        if (!isLikeNone(metadata_config)) {
+            _assertClass(metadata_config, WasmMetadataConfig);
+            ptr0 = metadata_config.__destroy_into_raw();
+        }
+        wasm.convertBytesWithMetadata(retptr, addHeapObject(html), addHeapObject(options), ptr0);
+        var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+        var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+        var r2 = getDataViewMemory0().getInt32(retptr + 4 * 2, true);
+        if (r2) {
+            throw takeObject(r1);
+        }
+        return takeObject(r0);
+    } finally {
+        wasm.__wbindgen_add_to_stack_pointer(16);
+    }
+}
+exports.convertBytesWithMetadata = convertBytesWithMetadata;
+
 /**
  * @param {Uint8Array} html
  * @param {WasmConversionOptionsHandle} handle
@@ -764,6 +914,63 @@ function convertWithInlineImages(html, options, image_config) {
 }
 exports.convertWithInlineImages = convertWithInlineImages;
 
+/**
+ * Convert HTML to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithMetadata, WasmMetadataConfig } from 'html-to-markdown-wasm';
+ *
+ * const html = '<h1>Hello World</h1><a href="https://example.com">Link</a>';
+ * const config = new WasmMetadataConfig();
+ * config.extractHeaders = true;
+ * config.extractLinks = true;
+ *
+ * const result = convertWithMetadata(html, null, config);
+ * console.log(result.markdown); // # Hello World\n\n[Link](https://example.com)
+ * console.log(result.metadata.headers); // [{ level: 1, text: "Hello World", ... }]
+ * console.log(result.metadata.links); // [{ href: "https://example.com", text: "Link", ... }]
+ * ```
+ * @param {string} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+function convertWithMetadata(html, options, metadata_config) {
+    try {
+        const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+        const ptr0 = passStringToWasm0(html, wasm.__wbindgen_export, wasm.__wbindgen_export2);
+        const len0 = WASM_VECTOR_LEN;
+        let ptr1 = 0;
+        if (!isLikeNone(metadata_config)) {
+            _assertClass(metadata_config, WasmMetadataConfig);
+            ptr1 = metadata_config.__destroy_into_raw();
+        }
+        wasm.convertWithMetadata(retptr, ptr0, len0, addHeapObject(options), ptr1);
+        var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+        var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+        var r2 = getDataViewMemory0().getInt32(retptr + 4 * 2, true);
+        if (r2) {
+            throw takeObject(r1);
+        }
+        return takeObject(r0);
+    } finally {
+        wasm.__wbindgen_add_to_stack_pointer(16);
+    }
+}
+exports.convertWithMetadata = convertWithMetadata;
+
 /**
  * @param {string} html
  * @param {WasmConversionOptionsHandle} handle
@@ -1052,6 +1259,11 @@ exports.__wbg_new_1ba21ce319a06297 = function() {
     return addHeapObject(ret);
 };
 
+exports.__wbg_new_25f239778d6112b9 = function() {
+    const ret = new Array();
+    return addHeapObject(ret);
+};
+
 exports.__wbg_new_6421f6084cc5bc5a = function(arg0) {
     const ret = new Uint8Array(getObject(arg0));
     return addHeapObject(ret);
@@ -1090,6 +1302,15 @@ exports.__wbg_set_3f1d0b984ed272ed = function(arg0, arg1, arg2) {
     getObject(arg0)[takeObject(arg1)] = takeObject(arg2);
 };
 
+exports.__wbg_set_781438a03c0c3c81 = function() { return handleError(function (arg0, arg1, arg2) {
+    const ret = Reflect.set(getObject(arg0), getObject(arg1), getObject(arg2));
+    return ret;
+}, arguments) };
+
+exports.__wbg_set_7df433eea03a5c14 = function(arg0, arg1, arg2) {
+    getObject(arg0)[arg1 >>> 0] = takeObject(arg2);
+};
+
 exports.__wbg_set_efaaf145b9377369 = function(arg0, arg1, arg2) {
     const ret = getObject(arg0).set(getObject(arg1), getObject(arg2));
     return addHeapObject(ret);
@@ -1130,6 +1351,12 @@ exports.__wbindgen_cast_4625c577ab2ec9ee = function(arg0) {
     return addHeapObject(ret);
 };
 
+exports.__wbindgen_cast_d6cd19b81560fd6e = function(arg0) {
+    // Cast intrinsic for `F64 -> Externref`.
+    const ret = arg0;
+    return addHeapObject(ret);
+};
+
 exports.__wbindgen_object_clone_ref = function(arg0) {
     const ret = getObject(arg0);
     return addHeapObject(ret);

package/dist-node/html_to_markdown_wasm_bg.wasm.d.ts

@@ -6,11 +6,14 @@ export const __wbg_wasmhtmlextraction_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimage_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimageconfig_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimagewarning_free: (a: number, b: number) => void;
+export const __wbg_wasmmetadataconfig_free: (a: number, b: number) => void;
 export const convert: (a: number, b: number, c: number, d: number) => void;
 export const convertBytes: (a: number, b: number, c: number) => void;
 export const convertBytesWithInlineImages: (a: number, b: number, c: number, d: number) => void;
+export const convertBytesWithMetadata: (a: number, b: number, c: number, d: number) => void;
 export const convertBytesWithOptionsHandle: (a: number, b: number, c: number) => void;
 export const convertWithInlineImages: (a: number, b: number, c: number, d: number, e: number) => void;
+export const convertWithMetadata: (a: number, b: number, c: number, d: number, e: number) => void;
 export const convertWithOptionsHandle: (a: number, b: number, c: number, d: number) => void;
 export const createConversionOptionsHandle: (a: number, b: number) => void;
 export const wasmconversionoptionshandle_new: (a: number, b: number) => void;
@@ -30,6 +33,19 @@ export const wasminlineimageconfig_set_filenamePrefix: (a: number, b: number, c:
 export const wasminlineimageconfig_set_inferDimensions: (a: number, b: number) => void;
 export const wasminlineimagewarning_index: (a: number) => number;
 export const wasminlineimagewarning_message: (a: number, b: number) => void;
+export const wasmmetadataconfig_extract_document: (a: number) => number;
+export const wasmmetadataconfig_extract_headers: (a: number) => number;
+export const wasmmetadataconfig_extract_images: (a: number) => number;
+export const wasmmetadataconfig_extract_links: (a: number) => number;
+export const wasmmetadataconfig_extract_structured_data: (a: number) => number;
+export const wasmmetadataconfig_max_structured_data_size: (a: number) => number;
+export const wasmmetadataconfig_new: () => number;
+export const wasmmetadataconfig_set_extract_document: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_headers: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_images: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_links: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_structured_data: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_max_structured_data_size: (a: number, b: number) => void;
 export const init: () => void;
 export const __wbindgen_export: (a: number, b: number) => number;
 export const __wbindgen_export2: (a: number, b: number, c: number, d: number) => number;

package/dist-node/package.json

@@ -3,7 +3,7 @@
   "collaborators": [
     "Na'aman Hirschfeld <nhirschfeld@gmail.com>"
   ],
-  "version": "2.14.1",
+  "version": "2.14.3",
   "license": "MIT",
   "repository": {
     "type": "git",

package/dist-web/README.md

@@ -11,7 +11,7 @@ High-performance HTML → Markdown conversion powered by Rust. Shipping as a Rus
 [![Hex.pm](https://img.shields.io/hexpm/v/html_to_markdown.svg)](https://hex.pm/packages/html_to_markdown)
 [![NuGet](https://img.shields.io/nuget/v/Goldziher.HtmlToMarkdown.svg)](https://www.nuget.org/packages/Goldziher.HtmlToMarkdown/)
 [![Maven Central](https://img.shields.io/maven-central/v/io.github.goldziher/html-to-markdown.svg)](https://central.sonatype.com/artifact/io.github.goldziher/html-to-markdown)
-[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown)
+[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Goldziher/html-to-markdown/blob/main/LICENSE)
 [![Discord](https://img.shields.io/badge/Discord-Join%20our%20community-7289da)](https://discord.gg/pXxagNK2zN)
 
@@ -377,7 +377,7 @@ import (
 	"fmt"
 	"log"
 
-	"github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown"
+	"github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown"
 )
 
 func main() {

package/dist-web/html_to_markdown_wasm.d.ts

@@ -46,6 +46,23 @@ export class WasmInlineImageWarning {
   readonly message: string;
 }
 
+export class WasmMetadataConfig {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Create a new metadata configuration with defaults
+   *
+   * All extraction types enabled by default with 1MB structured data limit
+   */
+  constructor();
+  extract_links: boolean;
+  extract_images: boolean;
+  extract_headers: boolean;
+  extract_document: boolean;
+  extract_structured_data: boolean;
+  max_structured_data_size: number;
+}
+
 /**
  * Convert HTML to Markdown
  *
@@ -70,10 +87,56 @@ export function convertBytes(html: Uint8Array, options?: WasmConversionOptions |
 
 export function convertBytesWithInlineImages(html: Uint8Array, options?: WasmConversionOptions | null, image_config?: WasmInlineImageConfig | null): WasmHtmlExtraction;
 
+/**
+ * Convert HTML bytes to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML bytes to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ */
+export function convertBytesWithMetadata(html: Uint8Array, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function convertBytesWithOptionsHandle(html: Uint8Array, handle: WasmConversionOptionsHandle): string;
 
 export function convertWithInlineImages(html: string, options?: WasmConversionOptions | null, image_config?: WasmInlineImageConfig | null): WasmHtmlExtraction;
 
+/**
+ * Convert HTML to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithMetadata, WasmMetadataConfig } from 'html-to-markdown-wasm';
+ *
+ * const html = '<h1>Hello World</h1><a href="https://example.com">Link</a>';
+ * const config = new WasmMetadataConfig();
+ * config.extractHeaders = true;
+ * config.extractLinks = true;
+ *
+ * const result = convertWithMetadata(html, null, config);
+ * console.log(result.markdown); // # Hello World\n\n[Link](https://example.com)
+ * console.log(result.metadata.headers); // [{ level: 1, text: "Hello World", ... }]
+ * console.log(result.metadata.links); // [{ href: "https://example.com", text: "Link", ... }]
+ * ```
+ */
+export function convertWithMetadata(html: string, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function convertWithOptionsHandle(html: string, handle: WasmConversionOptionsHandle): string;
 
 export function createConversionOptionsHandle(options?: WasmConversionOptions | null): WasmConversionOptionsHandle;
@@ -92,11 +155,14 @@ export interface InitOutput {
   readonly __wbg_wasminlineimage_free: (a: number, b: number) => void;
   readonly __wbg_wasminlineimageconfig_free: (a: number, b: number) => void;
   readonly __wbg_wasminlineimagewarning_free: (a: number, b: number) => void;
+  readonly __wbg_wasmmetadataconfig_free: (a: number, b: number) => void;
   readonly convert: (a: number, b: number, c: number, d: number) => void;
   readonly convertBytes: (a: number, b: number, c: number) => void;
   readonly convertBytesWithInlineImages: (a: number, b: number, c: number, d: number) => void;
+  readonly convertBytesWithMetadata: (a: number, b: number, c: number, d: number) => void;
   readonly convertBytesWithOptionsHandle: (a: number, b: number, c: number) => void;
   readonly convertWithInlineImages: (a: number, b: number, c: number, d: number, e: number) => void;
+  readonly convertWithMetadata: (a: number, b: number, c: number, d: number, e: number) => void;
   readonly convertWithOptionsHandle: (a: number, b: number, c: number, d: number) => void;
   readonly createConversionOptionsHandle: (a: number, b: number) => void;
   readonly wasmconversionoptionshandle_new: (a: number, b: number) => void;
@@ -116,6 +182,19 @@ export interface InitOutput {
   readonly wasminlineimageconfig_set_inferDimensions: (a: number, b: number) => void;
   readonly wasminlineimagewarning_index: (a: number) => number;
   readonly wasminlineimagewarning_message: (a: number, b: number) => void;
+  readonly wasmmetadataconfig_extract_document: (a: number) => number;
+  readonly wasmmetadataconfig_extract_headers: (a: number) => number;
+  readonly wasmmetadataconfig_extract_images: (a: number) => number;
+  readonly wasmmetadataconfig_extract_links: (a: number) => number;
+  readonly wasmmetadataconfig_extract_structured_data: (a: number) => number;
+  readonly wasmmetadataconfig_max_structured_data_size: (a: number) => number;
+  readonly wasmmetadataconfig_new: () => number;
+  readonly wasmmetadataconfig_set_extract_document: (a: number, b: number) => void;
+  readonly wasmmetadataconfig_set_extract_headers: (a: number, b: number) => void;
+  readonly wasmmetadataconfig_set_extract_images: (a: number, b: number) => void;
+  readonly wasmmetadataconfig_set_extract_links: (a: number, b: number) => void;
+  readonly wasmmetadataconfig_set_extract_structured_data: (a: number, b: number) => void;
+  readonly wasmmetadataconfig_set_max_structured_data_size: (a: number, b: number) => void;
   readonly init: () => void;
   readonly __wbindgen_export: (a: number, b: number) => number;
   readonly __wbindgen_export2: (a: number, b: number, c: number, d: number) => number;

package/dist-web/html_to_markdown_wasm.js

@@ -250,6 +250,10 @@ const WasmInlineImageWarningFinalization = (typeof FinalizationRegistry === 'und
     ? { register: () => {}, unregister: () => {} }
     : new FinalizationRegistry(ptr => wasm.__wbg_wasminlineimagewarning_free(ptr >>> 0, 1));
 
+const WasmMetadataConfigFinalization = (typeof FinalizationRegistry === 'undefined')
+    ? { register: () => {}, unregister: () => {} }
+    : new FinalizationRegistry(ptr => wasm.__wbg_wasmmetadataconfig_free(ptr >>> 0, 1));
+
 export class WasmConversionOptionsHandle {
     static __wrap(ptr) {
         ptr = ptr >>> 0;
@@ -595,6 +599,112 @@ export class WasmInlineImageWarning {
 }
 if (Symbol.dispose) WasmInlineImageWarning.prototype[Symbol.dispose] = WasmInlineImageWarning.prototype.free;
 
+/**
+ * Metadata extraction configuration
+ */
+export class WasmMetadataConfig {
+    __destroy_into_raw() {
+        const ptr = this.__wbg_ptr;
+        this.__wbg_ptr = 0;
+        WasmMetadataConfigFinalization.unregister(this);
+        return ptr;
+    }
+    free() {
+        const ptr = this.__destroy_into_raw();
+        wasm.__wbg_wasmmetadataconfig_free(ptr, 0);
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_links() {
+        const ret = wasm.wasmmetadataconfig_extract_links(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_images() {
+        const ret = wasm.wasmmetadataconfig_extract_images(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_headers() {
+        const ret = wasm.wasmmetadataconfig_extract_headers(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_document() {
+        const ret = wasm.wasmmetadataconfig_extract_document(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_links(value) {
+        wasm.wasmmetadataconfig_set_extract_links(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_images(value) {
+        wasm.wasmmetadataconfig_set_extract_images(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_headers(value) {
+        wasm.wasmmetadataconfig_set_extract_headers(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_document(value) {
+        wasm.wasmmetadataconfig_set_extract_document(this.__wbg_ptr, value);
+    }
+    /**
+     * @returns {boolean}
+     */
+    get extract_structured_data() {
+        const ret = wasm.wasmmetadataconfig_extract_structured_data(this.__wbg_ptr);
+        return ret !== 0;
+    }
+    /**
+     * @returns {number}
+     */
+    get max_structured_data_size() {
+        const ret = wasm.wasmmetadataconfig_max_structured_data_size(this.__wbg_ptr);
+        return ret >>> 0;
+    }
+    /**
+     * @param {boolean} value
+     */
+    set extract_structured_data(value) {
+        wasm.wasmmetadataconfig_set_extract_structured_data(this.__wbg_ptr, value);
+    }
+    /**
+     * @param {number} value
+     */
+    set max_structured_data_size(value) {
+        wasm.wasmmetadataconfig_set_max_structured_data_size(this.__wbg_ptr, value);
+    }
+    /**
+     * Create a new metadata configuration with defaults
+     *
+     * All extraction types enabled by default with 1MB structured data limit
+     */
+    constructor() {
+        const ret = wasm.wasmmetadataconfig_new();
+        this.__wbg_ptr = ret >>> 0;
+        WasmMetadataConfigFinalization.register(this, this.__wbg_ptr, this);
+        return this;
+    }
+}
+if (Symbol.dispose) WasmMetadataConfig.prototype[Symbol.dispose] = WasmMetadataConfig.prototype.free;
+
 /**
  * Convert HTML to Markdown
  *
@@ -700,6 +810,44 @@ export function convertBytesWithInlineImages(html, options, image_config) {
     }
 }
 
+/**
+ * Convert HTML bytes to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML bytes to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ * @param {Uint8Array} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+export function convertBytesWithMetadata(html, options, metadata_config) {
+    try {
+        const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+        let ptr0 = 0;
+        if (!isLikeNone(metadata_config)) {
+            _assertClass(metadata_config, WasmMetadataConfig);
+            ptr0 = metadata_config.__destroy_into_raw();
+        }
+        wasm.convertBytesWithMetadata(retptr, addHeapObject(html), addHeapObject(options), ptr0);
+        var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+        var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+        var r2 = getDataViewMemory0().getInt32(retptr + 4 * 2, true);
+        if (r2) {
+            throw takeObject(r1);
+        }
+        return takeObject(r0);
+    } finally {
+        wasm.__wbindgen_add_to_stack_pointer(16);
+    }
+}
+
 /**
  * @param {Uint8Array} html
  * @param {WasmConversionOptionsHandle} handle
@@ -760,6 +908,62 @@ export function convertWithInlineImages(html, options, image_config) {
     }
 }
 
+/**
+ * Convert HTML to Markdown with metadata extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `markdown` (string) and `metadata` (object) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithMetadata, WasmMetadataConfig } from 'html-to-markdown-wasm';
+ *
+ * const html = '<h1>Hello World</h1><a href="https://example.com">Link</a>';
+ * const config = new WasmMetadataConfig();
+ * config.extractHeaders = true;
+ * config.extractLinks = true;
+ *
+ * const result = convertWithMetadata(html, null, config);
+ * console.log(result.markdown); // # Hello World\n\n[Link](https://example.com)
+ * console.log(result.metadata.headers); // [{ level: 1, text: "Hello World", ... }]
+ * console.log(result.metadata.links); // [{ href: "https://example.com", text: "Link", ... }]
+ * ```
+ * @param {string} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+export function convertWithMetadata(html, options, metadata_config) {
+    try {
+        const retptr = wasm.__wbindgen_add_to_stack_pointer(-16);
+        const ptr0 = passStringToWasm0(html, wasm.__wbindgen_export, wasm.__wbindgen_export2);
+        const len0 = WASM_VECTOR_LEN;
+        let ptr1 = 0;
+        if (!isLikeNone(metadata_config)) {
+            _assertClass(metadata_config, WasmMetadataConfig);
+            ptr1 = metadata_config.__destroy_into_raw();
+        }
+        wasm.convertWithMetadata(retptr, ptr0, len0, addHeapObject(options), ptr1);
+        var r0 = getDataViewMemory0().getInt32(retptr + 4 * 0, true);
+        var r1 = getDataViewMemory0().getInt32(retptr + 4 * 1, true);
+        var r2 = getDataViewMemory0().getInt32(retptr + 4 * 2, true);
+        if (r2) {
+            throw takeObject(r1);
+        }
+        return takeObject(r0);
+    } finally {
+        wasm.__wbindgen_add_to_stack_pointer(16);
+    }
+}
+
 /**
  * @param {string} html
  * @param {WasmConversionOptionsHandle} handle
@@ -1043,6 +1247,10 @@ function __wbg_get_imports() {
         const ret = new Object();
         return addHeapObject(ret);
     };
+    imports.wbg.__wbg_new_25f239778d6112b9 = function() {
+        const ret = new Array();
+        return addHeapObject(ret);
+    };
     imports.wbg.__wbg_new_6421f6084cc5bc5a = function(arg0) {
         const ret = new Uint8Array(getObject(arg0));
         return addHeapObject(ret);
@@ -1073,6 +1281,13 @@ function __wbg_get_imports() {
     imports.wbg.__wbg_set_3f1d0b984ed272ed = function(arg0, arg1, arg2) {
         getObject(arg0)[takeObject(arg1)] = takeObject(arg2);
     };
+    imports.wbg.__wbg_set_781438a03c0c3c81 = function() { return handleError(function (arg0, arg1, arg2) {
+        const ret = Reflect.set(getObject(arg0), getObject(arg1), getObject(arg2));
+        return ret;
+    }, arguments) };
+    imports.wbg.__wbg_set_7df433eea03a5c14 = function(arg0, arg1, arg2) {
+        getObject(arg0)[arg1 >>> 0] = takeObject(arg2);
+    };
     imports.wbg.__wbg_set_efaaf145b9377369 = function(arg0, arg1, arg2) {
         const ret = getObject(arg0).set(getObject(arg1), getObject(arg2));
         return addHeapObject(ret);
@@ -1106,6 +1321,11 @@ function __wbg_get_imports() {
         const ret = BigInt.asUintN(64, arg0);
         return addHeapObject(ret);
     };
+    imports.wbg.__wbindgen_cast_d6cd19b81560fd6e = function(arg0) {
+        // Cast intrinsic for `F64 -> Externref`.
+        const ret = arg0;
+        return addHeapObject(ret);
+    };
     imports.wbg.__wbindgen_object_clone_ref = function(arg0) {
         const ret = getObject(arg0);
         return addHeapObject(ret);

package/dist-web/html_to_markdown_wasm_bg.wasm.d.ts

@@ -6,11 +6,14 @@ export const __wbg_wasmhtmlextraction_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimage_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimageconfig_free: (a: number, b: number) => void;
 export const __wbg_wasminlineimagewarning_free: (a: number, b: number) => void;
+export const __wbg_wasmmetadataconfig_free: (a: number, b: number) => void;
 export const convert: (a: number, b: number, c: number, d: number) => void;
 export const convertBytes: (a: number, b: number, c: number) => void;
 export const convertBytesWithInlineImages: (a: number, b: number, c: number, d: number) => void;
+export const convertBytesWithMetadata: (a: number, b: number, c: number, d: number) => void;
 export const convertBytesWithOptionsHandle: (a: number, b: number, c: number) => void;
 export const convertWithInlineImages: (a: number, b: number, c: number, d: number, e: number) => void;
+export const convertWithMetadata: (a: number, b: number, c: number, d: number, e: number) => void;
 export const convertWithOptionsHandle: (a: number, b: number, c: number, d: number) => void;
 export const createConversionOptionsHandle: (a: number, b: number) => void;
 export const wasmconversionoptionshandle_new: (a: number, b: number) => void;
@@ -30,6 +33,19 @@ export const wasminlineimageconfig_set_filenamePrefix: (a: number, b: number, c:
 export const wasminlineimageconfig_set_inferDimensions: (a: number, b: number) => void;
 export const wasminlineimagewarning_index: (a: number) => number;
 export const wasminlineimagewarning_message: (a: number, b: number) => void;
+export const wasmmetadataconfig_extract_document: (a: number) => number;
+export const wasmmetadataconfig_extract_headers: (a: number) => number;
+export const wasmmetadataconfig_extract_images: (a: number) => number;
+export const wasmmetadataconfig_extract_links: (a: number) => number;
+export const wasmmetadataconfig_extract_structured_data: (a: number) => number;
+export const wasmmetadataconfig_max_structured_data_size: (a: number) => number;
+export const wasmmetadataconfig_new: () => number;
+export const wasmmetadataconfig_set_extract_document: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_headers: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_images: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_links: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_extract_structured_data: (a: number, b: number) => void;
+export const wasmmetadataconfig_set_max_structured_data_size: (a: number, b: number) => void;
 export const init: () => void;
 export const __wbindgen_export: (a: number, b: number) => number;
 export const __wbindgen_export2: (a: number, b: number, c: number, d: number) => number;

package/dist-web/package.json

@@ -4,7 +4,7 @@
   "collaborators": [
     "Na'aman Hirschfeld <nhirschfeld@gmail.com>"
   ],
-  "version": "2.14.1",
+  "version": "2.14.3",
   "license": "MIT",
   "repository": {
     "type": "git",

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "html-to-markdown-wasm",
-	"version": "2.14.1",
+	"version": "2.14.3",
 	"description": "High-performance HTML to Markdown converter - WebAssembly bindings",
 	"main": "dist/html_to_markdown_wasm.js",
 	"types": "dist/html_to_markdown_wasm.d.ts",