@kreuzberg/html-to-markdown-wasm 2.27.3 → 2.28.0

package/dist/README.md

@@ -64,6 +64,7 @@ High-performance HTML to Markdown conversion powered by Rust. Ships as native bi
 - **12 language bindings** with consistent output across all runtimes
 - **Metadata extraction** — title, headers, links, images, structured data (JSON-LD, Microdata, RDFa)
 - **Visitor pattern** — custom callbacks for content filtering, URL rewriting, domain-specific dialects
+- **Table extraction** — extract structured table data (cells, headers, rendered markdown) during conversion
 - **Secure by default** — built-in HTML sanitization via ammonia
 
 ## Quick Start

package/dist/html_to_markdown_wasm.d.ts

@@ -151,6 +151,32 @@ export function convertWithMetadata(html: string, options?: WasmConversionOption
 
 export function convertWithOptionsHandle(html: string, handle: WasmConversionOptionsHandle): string;
 
+/**
+ * Convert HTML to Markdown with structured table extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Optional metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `content` (string), `tables` (array), and `metadata` (object|null) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithTables } from 'html-to-markdown-wasm';
+ *
+ * const html = '<table><tr><th>Name</th></tr><tr><td>Alice</td></tr></table>';
+ * const result = convertWithTables(html, null, null);
+ * console.log(result.content);
+ * console.log(result.tables[0].cells);
+ * ```
+ */
+export function convertWithTables(html: string, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function createConversionOptionsHandle(options?: WasmConversionOptions | null): WasmConversionOptionsHandle;
 
 /**

package/dist/html_to_markdown_wasm_bg.js

@@ -738,6 +738,57 @@ export function convertWithOptionsHandle(html, handle) {
     }
 }
 
+/**
+ * Convert HTML to Markdown with structured table extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Optional metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `content` (string), `tables` (array), and `metadata` (object|null) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithTables } from 'html-to-markdown-wasm';
+ *
+ * const html = '<table><tr><th>Name</th></tr><tr><td>Alice</td></tr></table>';
+ * const result = convertWithTables(html, null, null);
+ * console.log(result.content);
+ * console.log(result.tables[0].cells);
+ * ```
+ * @param {string} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+export function convertWithTables(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.convertWithTables(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 {WasmConversionOptions | null | undefined} [options]
  * @returns {WasmConversionOptionsHandle}

package/dist/html_to_markdown_wasm_bg.wasm.d.ts

@@ -15,6 +15,7 @@ export const convertBytesWithOptionsHandle: (a: number, b: number, c: number) =>
 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 convertWithTables: (a: number, b: number, c: number, d: number, e: number) => void;
 export const createConversionOptionsHandle: (a: number, b: number) => void;
 export const wasmconversionoptionshandle_new: (a: number, b: number) => void;
 export const wasmhtmlextraction_inlineImages: (a: number, b: number) => void;

package/dist/package.json

@@ -5,7 +5,7 @@
     "Na'aman Hirschfeld <nhirschfeld@gmail.com>"
   ],
   "description": "HTML to Markdown conversion for WebAssembly targets",
-  "version": "2.27.3",
+  "version": "2.28.0",
   "license": "MIT",
   "repository": {
     "type": "git",

package/dist-node/README.md

@@ -64,6 +64,7 @@ High-performance HTML to Markdown conversion powered by Rust. Ships as native bi
 - **12 language bindings** with consistent output across all runtimes
 - **Metadata extraction** — title, headers, links, images, structured data (JSON-LD, Microdata, RDFa)
 - **Visitor pattern** — custom callbacks for content filtering, URL rewriting, domain-specific dialects
+- **Table extraction** — extract structured table data (cells, headers, rendered markdown) during conversion
 - **Secure by default** — built-in HTML sanitization via ammonia
 
 ## Quick Start

package/dist-node/html_to_markdown_wasm.d.ts

@@ -151,6 +151,32 @@ export function convertWithMetadata(html: string, options?: WasmConversionOption
 
 export function convertWithOptionsHandle(html: string, handle: WasmConversionOptionsHandle): string;
 
+/**
+ * Convert HTML to Markdown with structured table extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Optional metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `content` (string), `tables` (array), and `metadata` (object|null) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithTables } from 'html-to-markdown-wasm';
+ *
+ * const html = '<table><tr><th>Name</th></tr><tr><td>Alice</td></tr></table>';
+ * const result = convertWithTables(html, null, null);
+ * console.log(result.content);
+ * console.log(result.tables[0].cells);
+ * ```
+ */
+export function convertWithTables(html: string, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function createConversionOptionsHandle(options?: WasmConversionOptions | null): WasmConversionOptionsHandle;
 
 /**

package/dist-node/html_to_markdown_wasm.js

@@ -758,6 +758,58 @@ function convertWithOptionsHandle(html, handle) {
 }
 exports.convertWithOptionsHandle = convertWithOptionsHandle;
 
+/**
+ * Convert HTML to Markdown with structured table extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Optional metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `content` (string), `tables` (array), and `metadata` (object|null) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithTables } from 'html-to-markdown-wasm';
+ *
+ * const html = '<table><tr><th>Name</th></tr><tr><td>Alice</td></tr></table>';
+ * const result = convertWithTables(html, null, null);
+ * console.log(result.content);
+ * console.log(result.tables[0].cells);
+ * ```
+ * @param {string} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+function convertWithTables(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.convertWithTables(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.convertWithTables = convertWithTables;
+
 /**
  * @param {WasmConversionOptions | null | undefined} [options]
  * @returns {WasmConversionOptionsHandle}

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

@@ -15,6 +15,7 @@ export const convertBytesWithOptionsHandle: (a: number, b: number, c: number) =>
 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 convertWithTables: (a: number, b: number, c: number, d: number, e: number) => void;
 export const createConversionOptionsHandle: (a: number, b: number) => void;
 export const wasmconversionoptionshandle_new: (a: number, b: number) => void;
 export const wasmhtmlextraction_inlineImages: (a: number, b: number) => void;

package/dist-node/package.json

@@ -4,7 +4,7 @@
     "Na'aman Hirschfeld <nhirschfeld@gmail.com>"
   ],
   "description": "HTML to Markdown conversion for WebAssembly targets",
-  "version": "2.27.3",
+  "version": "2.28.0",
   "license": "MIT",
   "repository": {
     "type": "git",

package/dist-web/README.md

@@ -64,6 +64,7 @@ High-performance HTML to Markdown conversion powered by Rust. Ships as native bi
 - **12 language bindings** with consistent output across all runtimes
 - **Metadata extraction** — title, headers, links, images, structured data (JSON-LD, Microdata, RDFa)
 - **Visitor pattern** — custom callbacks for content filtering, URL rewriting, domain-specific dialects
+- **Table extraction** — extract structured table data (cells, headers, rendered markdown) during conversion
 - **Secure by default** — built-in HTML sanitization via ammonia
 
 ## Quick Start

package/dist-web/html_to_markdown_wasm.d.ts

@@ -151,6 +151,32 @@ export function convertWithMetadata(html: string, options?: WasmConversionOption
 
 export function convertWithOptionsHandle(html: string, handle: WasmConversionOptionsHandle): string;
 
+/**
+ * Convert HTML to Markdown with structured table extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Optional metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `content` (string), `tables` (array), and `metadata` (object|null) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithTables } from 'html-to-markdown-wasm';
+ *
+ * const html = '<table><tr><th>Name</th></tr><tr><td>Alice</td></tr></table>';
+ * const result = convertWithTables(html, null, null);
+ * console.log(result.content);
+ * console.log(result.tables[0].cells);
+ * ```
+ */
+export function convertWithTables(html: string, options?: WasmConversionOptions | null, metadata_config?: WasmMetadataConfig | null): any;
+
 export function createConversionOptionsHandle(options?: WasmConversionOptions | null): WasmConversionOptionsHandle;
 
 /**
@@ -176,6 +202,7 @@ export interface InitOutput {
     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 convertWithTables: (a: number, b: number, c: number, d: number, e: number) => void;
     readonly createConversionOptionsHandle: (a: number, b: number) => void;
     readonly wasmconversionoptionshandle_new: (a: number, b: number) => void;
     readonly wasmhtmlextraction_inlineImages: (a: number, b: number) => void;

package/dist-web/html_to_markdown_wasm.js

@@ -744,6 +744,57 @@ export function convertWithOptionsHandle(html, handle) {
     }
 }
 
+/**
+ * Convert HTML to Markdown with structured table extraction
+ *
+ * # Arguments
+ *
+ * * `html` - The HTML string to convert
+ * * `options` - Optional conversion options (as a JavaScript object)
+ * * `metadata_config` - Optional metadata extraction configuration
+ *
+ * # Returns
+ *
+ * JavaScript object with `content` (string), `tables` (array), and `metadata` (object|null) fields
+ *
+ * # Example
+ *
+ * ```javascript
+ * import { convertWithTables } from 'html-to-markdown-wasm';
+ *
+ * const html = '<table><tr><th>Name</th></tr><tr><td>Alice</td></tr></table>';
+ * const result = convertWithTables(html, null, null);
+ * console.log(result.content);
+ * console.log(result.tables[0].cells);
+ * ```
+ * @param {string} html
+ * @param {WasmConversionOptions | null | undefined} [options]
+ * @param {WasmMetadataConfig | null} [metadata_config]
+ * @returns {Record<string, string>}
+ */
+export function convertWithTables(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.convertWithTables(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 {WasmConversionOptions | null | undefined} [options]
  * @returns {WasmConversionOptionsHandle}

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

@@ -15,6 +15,7 @@ export const convertBytesWithOptionsHandle: (a: number, b: number, c: number) =>
 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 convertWithTables: (a: number, b: number, c: number, d: number, e: number) => void;
 export const createConversionOptionsHandle: (a: number, b: number) => void;
 export const wasmconversionoptionshandle_new: (a: number, b: number) => void;
 export const wasmhtmlextraction_inlineImages: (a: number, b: number) => void;

package/dist-web/package.json

@@ -5,7 +5,7 @@
     "Na'aman Hirschfeld <nhirschfeld@gmail.com>"
   ],
   "description": "HTML to Markdown conversion for WebAssembly targets",
-  "version": "2.27.3",
+  "version": "2.28.0",
   "license": "MIT",
   "repository": {
     "type": "git",

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@kreuzberg/html-to-markdown-wasm",
-	"version": "2.27.3",
+	"version": "2.28.0",
 	"description": "High-performance HTML to Markdown converter - WebAssembly bindings",
 	"main": "dist/html_to_markdown_wasm.js",
 	"types": "dist/html_to_markdown_wasm.d.ts",