@ryangarber/markitdown-ts 0.1.2 → 0.1.4

package/dist/index.cjs

@@ -29,8 +29,8 @@ let linkedom = require("linkedom");
 let mammoth = require("mammoth");
 mammoth = __toESM(mammoth, 1);
 let unpdf = require("unpdf");
-let xlsx = require("xlsx");
-xlsx = __toESM(xlsx, 1);
+let exceljs = require("exceljs");
+exceljs = __toESM(exceljs, 1);
 let _xmldom_xmldom = require("@xmldom/xmldom");
 let fflate = require("fflate");
 //#region src/bytes.ts
@@ -160,12 +160,13 @@ var HtmlConverter = class {
 	}
 	async convertHtml(html) {
 		const document = parseHtmlDocument(html);
-		document.querySelectorAll("script, style").forEach((element) => {
+		const title = document.title;
+		document.querySelectorAll("script, style, title").forEach((element) => {
 			element.remove();
 		});
-		const markdown = new CustomTurnDown().convertSoup(document.body ?? document);
+		const markdown = new CustomTurnDown().convertSoup(document);
 		return {
-			title: document.title || null,
+			title: title || null,
 			markdown,
 			text_content: markdown
 		};
@@ -176,14 +177,17 @@ var HtmlConverter = class {
 var DocxConverter = class extends HtmlConverter {
 	async convert(source, options) {
 		if (options.file_extension?.toLowerCase() !== ".docx") return null;
-		const html = await mammoth.default.convertToHtml({ arrayBuffer: toArrayBuffer(source) }, {
+		let mammothInput = { arrayBuffer: toArrayBuffer(source) };
+		if (typeof process !== "undefined" && process.versions !== null && process.versions.node !== null) mammothInput = { buffer: Buffer.from(source) };
+		const mammothOptions = {
 			styleMap: options.styleMap,
 			includeEmbeddedStyleMap: options.includeEmbeddedStyleMap,
 			includeDefaultStyleMap: options.includeDefaultStyleMap,
 			ignoreEmptyParagraphs: options.ignoreEmptyParagraphs,
-			idPrefix: options.idPrefix,
-			transformDocument: options.transformDocument
-		});
+			idPrefix: options.idPrefix
+		};
+		if (options.transformDocument) mammothOptions.transformDocument = options.transformDocument;
+		const html = await mammoth.default.convertToHtml(mammothInput, mammothOptions);
 		return this.convertHtml(html.value);
 	}
 };
@@ -305,26 +309,45 @@ var WikipediaConverter = class {
 };
 //#endregion
 //#region src/converters/xlsx.ts
+const { Workbook } = exceljs.default;
 var XlsxConverter = class extends HtmlConverter {
 	async convert(source, options) {
 		if (options.file_extension?.toLowerCase() !== ".xlsx") return null;
-		const workbook = xlsx.read(source, { type: "array" });
+		const workbook = new Workbook();
+		await workbook.xlsx.load(toArrayBuffer(source));
 		const sections = [];
-		for (const sheetName of workbook.SheetNames) {
-			const sheet = workbook.Sheets[sheetName];
-			if (sheet?.["!ref"]) {
-				const result = await this.convertHtml(xlsx.utils.sheet_to_html(sheet));
-				sections.push(`## ${sheetName}\n${result?.markdown.trim() ?? ""}`);
-			}
+		for (const worksheet of workbook.worksheets) {
+			const html = worksheetToHtml(worksheet);
+			if (!html) continue;
+			const result = await this.convertHtml(html);
+			sections.push(`## ${worksheet.name}\n${result?.markdown.trim() ?? ""}`);
 		}
 		const markdown = sections.join("\n\n");
 		return {
-			title: workbook.Props?.Title || "Untitled",
+			title: workbook.title || "Untitled",
 			markdown,
 			text_content: markdown
 		};
 	}
 };
+function worksheetToHtml(worksheet) {
+	const columnCount = worksheet.columnCount;
+	if (!columnCount || !worksheet.actualRowCount) return "";
+	const rows = [];
+	worksheet.eachRow({ includeEmpty: true }, (row) => {
+		const cells = [];
+		for (let column = 1; column <= columnCount; column++) cells.push(`<td>${escapeHtml(cellText(row.getCell(column)))}</td>`);
+		rows.push(`<tr>${cells.join("")}</tr>`);
+	});
+	return `<html><body><table>${rows.join("")}</table></body></html>`;
+}
+function cellText(cell) {
+	if (cell.isMerged && cell.master !== cell) return "";
+	return cell.text;
+}
+function escapeHtml(value) {
+	return value.replace(/&/g, "&amp;").replace(/</g, "&lt;").replace(/>/g, "&gt;").replace(/"/g, "&quot;");
+}
 //#endregion
 //#region src/converters/xml-rss-atom.ts
 var RSSConverter = class {
@@ -527,7 +550,7 @@ var MarkItDown = class {
 		if (typeof source === "string" || source instanceof URL) return this.convertUrl(source.toString(), options);
 		if (isResponse(source)) return this.convertResponse(source, options);
 		const extensions = /* @__PURE__ */ new Set();
-		this.addExtension(extensions, options.file_extension);
+		this.addExtension(extensions, options.file_extension?.replace(/^\./, ""));
 		if (isBlob(source)) {
 			this.addExtension(extensions, extensionFromMime(source.type));
 			if ("name" in source && typeof source.name === "string") this.addExtension(extensions, extensionFromPathname(source.name));
@@ -580,6 +603,7 @@ var MarkItDown = class {
 				return result;
 			}
 		} catch (error) {
+			if (options.debug) console.error("Converter threw error:", error);
 			lastError = error;
 		}
 		const formats = [...extensions].join(", ");

package/dist/index.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"index.cjs","names":["TurndownService","turndownPluginGfm","parseHTML","getDocumentProxy","extractText","getMeta","XLSX","DOMParser","unzipSync"],"sources":["../src/bytes.ts","../src/custom-turndown.ts","../src/html-document.ts","../src/converters/bingserp.ts","../src/converters/html.ts","../src/converters/docx.ts","../src/converters/ipynb.ts","../src/converters/pdf.ts","../src/mime.ts","../src/converters/plain-text.ts","../src/converters/wikipedia.ts","../src/converters/xlsx.ts","../src/converters/xml-rss-atom.ts","../src/converters/youtube.ts","../src/converters/zip.ts","../src/markitdown.ts"],"sourcesContent":["import type { ByteSource } from \"./types\";\n\nexport async function toUint8Array(source: ByteSource): Promise<Uint8Array> {\n  if (source instanceof Uint8Array) {\n    return source;\n  }\n\n  if (isBlob(source)) {\n    return new Uint8Array(await source.arrayBuffer());\n  }\n\n  if (ArrayBuffer.isView(source)) {\n    return new Uint8Array(source.buffer, source.byteOffset, source.byteLength);\n  }\n\n  return new Uint8Array(source);\n}\n\nexport function toArrayBuffer(source: Uint8Array): ArrayBuffer {\n  return source.slice().buffer;\n}\n\nexport function decodeText(source: Uint8Array): string {\n  return new TextDecoder().decode(source);\n}\n\nexport function isBlob(value: unknown): value is Blob {\n  return typeof Blob !== \"undefined\" && value instanceof Blob;\n}\n\nexport function normalizeExtension(extension: string): string {\n  if (!extension) {\n    return \"\";\n  }\n\n  const normalized = extension.toLowerCase();\n  return normalized.startsWith(\".\") ? normalized : `.${normalized}`;\n}\n","import turndownPluginGfm from \"@joplin/turndown-plugin-gfm\";\nimport TurndownService from \"turndown\";\n\nexport class CustomTurnDown {\n  convertSoup(document: string | Node): string {\n    const service = new TurndownService({ headingStyle: \"atx\" });\n    service.use(turndownPluginGfm.gfm);\n\n    service.addRule(\"anchor tags\", {\n      filter: [\"a\"],\n      replacement(content, node) {\n        if (!content) return \"\";\n\n        const prefix = content.startsWith(\" \") ? \" \" : \"\";\n        const suffix = content.endsWith(\" \") ? \" \" : \"\";\n        const text = content.trim().replace(/\\n\\n.*/g, \"\");\n        if (!text) return \"\";\n\n        const element = node as unknown as Element;\n        const href = element.getAttribute(\"href\");\n        let title = element.getAttribute(\"title\") ?? \"\";\n        if (!href) return `${prefix}${text}${suffix}`;\n\n        try {\n          const url = new URL(href);\n          if (url.protocol !== \"https:\" && url.protocol !== \"http:\") {\n            return `${prefix}${text}${suffix}`;\n          }\n        } catch {\n          if (\n            !href.startsWith(\"/\") &&\n            !href.startsWith(\"#\") &&\n            !href.startsWith(\"./\") &&\n            !href.startsWith(\"../\")\n          ) {\n            return `${prefix}${text}${suffix}`;\n          }\n        }\n\n        if (text.replace(/\\\\_/g, \"_\") === href && !title) {\n          return `<${href}>`;\n        }\n        if (!title) title = href;\n        return `${prefix}[${text}](${href}${title ? ` \"${title}\"` : \"\"})${suffix}`;\n      }\n    });\n\n    service.addRule(\"img tags\", {\n      filter: [\"img\"],\n      replacement(_content, node) {\n        const element = node as unknown as Element;\n        const alt = element.getAttribute(\"alt\") ?? \"\";\n        let source = element.getAttribute(\"src\") ?? \"\";\n        const title = element.getAttribute(\"title\");\n        if (source.startsWith(\"data:\")) {\n          source = `${source.split(\",\", 1)[0]}...`;\n        }\n        return `![${alt}](${source}${title ? ` \"${title}\"` : \"\"})`;\n      }\n    });\n\n    return service.turndown(document as TurndownService.Node);\n  }\n}\n","import { parseHTML } from \"linkedom\";\n\nexport function parseHtmlDocument(html: string): Document {\n  return parseHTML(html).document as unknown as Document;\n}\n","import { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nconst BING_SEARCH_URL = /^https:\\/\\/www\\.bing\\.com\\/search\\?q=/i;\nconst BASE64_ALPHABET = \"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/\";\n\nexport class BingSerpConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    const url = options.url ?? \"\";\n    if ((extension !== \".html\" && extension !== \".htm\") || !BING_SEARCH_URL.test(url)) {\n      return null;\n    }\n\n    const document = parseHtmlDocument(decodeText(source));\n    const query = new URL(url).searchParams.get(\"q\") ?? \"\";\n\n    document.querySelectorAll(\".tptt\").forEach((element) => {\n      if (element.textContent) {\n        element.textContent += \" \";\n      }\n    });\n    document.querySelectorAll(\".algoSlug_icon\").forEach((element) => {\n      element.remove();\n    });\n\n    const turndown = new CustomTurnDown();\n    const results = [...document.querySelectorAll(\".b_algo\")].map((result) => {\n      result.querySelectorAll(\"a[href]\").forEach((anchor) => {\n        const href = anchor.getAttribute(\"href\");\n        if (!href) {\n          return;\n        }\n        try {\n          const encoded = new URL(href, url).searchParams.get(\"u\");\n          if (encoded) {\n            anchor.setAttribute(\"href\", decodeBingUrl(encoded));\n          }\n        } catch {\n          // Preserve malformed links as they appeared in the source document.\n        }\n      });\n\n      return turndown\n        .convertSoup(result)\n        .split(/\\n+/)\n        .map((line) => line.trim())\n        .filter(Boolean)\n        .join(\"\\n\");\n    });\n\n    const markdown = `## A Bing search for '${query}' found the following results:\\n\\n${results.join(\"\\n\\n\")}`;\n    return { title: document.title || null, markdown, text_content: markdown };\n  }\n}\n\nfunction decodeBingUrl(encoded: string): string {\n  const input = encoded.slice(2).replace(/-/g, \"+\").replace(/_/g, \"/\").replace(/=+$/, \"\");\n  const bytes: number[] = [];\n  let bits = 0;\n  let bitCount = 0;\n\n  for (const character of input) {\n    const value = BASE64_ALPHABET.indexOf(character);\n    if (value < 0) {\n      return encoded;\n    }\n    bits = (bits << 6) | value;\n    bitCount += 6;\n    if (bitCount >= 8) {\n      bitCount -= 8;\n      bytes.push((bits >> bitCount) & 0xff);\n    }\n  }\n\n  return new TextDecoder().decode(new Uint8Array(bytes));\n}\n","import { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class HtmlConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if (extension !== \".html\" && extension !== \".htm\") {\n      return null;\n    }\n\n    return this.convertHtml(decodeText(source));\n  }\n\n  async convertHtml(html: string): Promise<ConverterResult> {\n    const document = parseHtmlDocument(html);\n    document.querySelectorAll(\"script, style\").forEach((element) => {\n      element.remove();\n    });\n\n    const markdown = new CustomTurnDown().convertSoup(document.body ?? document);\n    return {\n      title: document.title || null,\n      markdown,\n      text_content: markdown\n    };\n  }\n}\n","import mammoth from \"mammoth\";\nimport { toArrayBuffer } from \"../bytes\";\nimport type { ConverterOptions, ConverterResult } from \"../types\";\nimport { HtmlConverter } from \"./html\";\n\nexport class DocxConverter extends HtmlConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".docx\") {\n      return null;\n    }\n\n    const html = await mammoth.convertToHtml(\n      { arrayBuffer: toArrayBuffer(source) },\n      {\n        styleMap: options.styleMap,\n        includeEmbeddedStyleMap: options.includeEmbeddedStyleMap,\n        includeDefaultStyleMap: options.includeDefaultStyleMap,\n        ignoreEmptyParagraphs: options.ignoreEmptyParagraphs,\n        idPrefix: options.idPrefix,\n        transformDocument: options.transformDocument\n      }\n    );\n    return this.convertHtml(html.value);\n  }\n}\n","import { decodeText } from \"../bytes\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\ntype NotebookCell = {\n  cell_type?: string;\n  source?: string[];\n};\n\ntype Notebook = {\n  cells?: NotebookCell[];\n  metadata?: { title?: string };\n};\n\nexport class IpynbConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".ipynb\") {\n      return null;\n    }\n\n    const notebook = JSON.parse(decodeText(source)) as Notebook;\n    const output: string[] = [];\n    let title: string | null = notebook.metadata?.title ?? null;\n\n    for (const cell of notebook.cells ?? []) {\n      const lines = cell.source ?? [];\n      if (cell.cell_type === \"markdown\") {\n        output.push(lines.join(\"\"));\n        title ??=\n          lines\n            .find((line) => line.startsWith(\"# \"))\n            ?.slice(2)\n            .trim() ?? null;\n      } else if (cell.cell_type === \"code\") {\n        output.push(`\\`\\`\\`python\\n${lines.join(\"\")}\\n\\`\\`\\``);\n      } else if (cell.cell_type === \"raw\") {\n        output.push(`\\`\\`\\`\\n${lines.join(\"\")}\\n\\`\\`\\``);\n      }\n    }\n\n    const markdown = output.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n}\n","import { extractText, getDocumentProxy, getMeta } from \"unpdf\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class PdfConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".pdf\") {\n      return null;\n    }\n\n    const pdf = await getDocumentProxy(source.slice());\n    try {\n      const [{ text }, { info }] = await Promise.all([\n        extractText(pdf, { mergePages: true }),\n        getMeta(pdf)\n      ]);\n      const title = typeof info.Title === \"string\" && info.Title.trim() ? info.Title : null;\n\n      return { title, markdown: text, text_content: text };\n    } finally {\n      await pdf.loadingTask.destroy();\n    }\n  }\n}\n","const MIME_EXTENSIONS: Record<string, string> = {\n  \"application/atom+xml\": \".atom\",\n  \"application/json\": \".json\",\n  \"application/pdf\": \".pdf\",\n  \"application/zip\": \".zip\",\n  \"application/rss+xml\": \".rss\",\n  \"application/x-pdf\": \".pdf\",\n  \"application/vnd.openxmlformats-officedocument.spreadsheetml.sheet\": \".xlsx\",\n  \"application/vnd.openxmlformats-officedocument.wordprocessingml.document\": \".docx\",\n  \"application/xml\": \".xml\",\n  \"text/csv\": \".csv\",\n  \"text/html\": \".html\",\n  \"text/markdown\": \".md\",\n  \"text/plain\": \".txt\",\n  \"text/xml\": \".xml\"\n};\n\nconst TEXT_EXTENSIONS = new Set([\".csv\", \".json\", \".log\", \".markdown\", \".md\", \".text\", \".txt\"]);\n\nexport function extensionFromMime(contentType: string | null | undefined): string | null {\n  if (!contentType) {\n    return null;\n  }\n\n  const mime = contentType.split(\";\", 1)[0]?.trim().toLowerCase() ?? \"\";\n  return MIME_EXTENSIONS[mime] ?? (mime.startsWith(\"text/\") ? \".txt\" : null);\n}\n\nexport function isTextExtension(extension: string): boolean {\n  return TEXT_EXTENSIONS.has(extension.toLowerCase());\n}\n\nexport function extensionFromPathname(pathname: string): string | null {\n  const filename = pathname.split(\"/\").pop() ?? \"\";\n  const dot = filename.lastIndexOf(\".\");\n  return dot > 0 ? filename.slice(dot).toLowerCase() : null;\n}\n","import { decodeText } from \"../bytes\";\nimport { isTextExtension } from \"../mime\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class PlainTextConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    if (!isTextExtension(options.file_extension ?? \"\")) {\n      return null;\n    }\n\n    const content = decodeText(source);\n    return { title: null, markdown: content, text_content: content };\n  }\n}\n","import { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nconst WIKIPEDIA_URL = /^https?:\\/\\/[a-z]{2,3}\\.wikipedia\\.org\\//i;\n\nexport class WikipediaConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if ((extension !== \".html\" && extension !== \".htm\") || !WIKIPEDIA_URL.test(options.url ?? \"\")) {\n      return null;\n    }\n\n    const document = parseHtmlDocument(decodeText(source));\n    document.querySelectorAll(\"script, style\").forEach((element) => {\n      element.remove();\n    });\n\n    const body = document.querySelector(\"div#mw-content-text\");\n    const pageTitle = document.querySelector(\"span.mw-page-title-main\")?.textContent?.trim();\n    const title = pageTitle || document.title || null;\n    const markdown = body\n      ? `# ${title ?? \"Wikipedia\"}\\n\\n${new CustomTurnDown().convertSoup(body)}`\n      : new CustomTurnDown().convertSoup(document);\n\n    return { title, markdown, text_content: markdown };\n  }\n}\n","import * as XLSX from \"xlsx\";\nimport type { ConverterOptions, ConverterResult } from \"../types\";\nimport { HtmlConverter } from \"./html\";\n\nexport class XlsxConverter extends HtmlConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".xlsx\") {\n      return null;\n    }\n\n    const workbook = XLSX.read(source, { type: \"array\" });\n    const sections: string[] = [];\n\n    for (const sheetName of workbook.SheetNames) {\n      const sheet = workbook.Sheets[sheetName];\n      if (sheet?.[\"!ref\"]) {\n        const result = await this.convertHtml(XLSX.utils.sheet_to_html(sheet));\n        sections.push(`## ${sheetName}\\n${result?.markdown.trim() ?? \"\"}`);\n      }\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return {\n      title: workbook.Props?.Title || \"Untitled\",\n      markdown,\n      text_content: markdown\n    };\n  }\n}\n","import type { Document, Element } from \"@xmldom/xmldom\";\nimport { DOMParser } from \"@xmldom/xmldom\";\nimport { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class RSSConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if (![\".xml\", \".rss\", \".atom\"].includes(extension)) {\n      return null;\n    }\n\n    const document = new DOMParser().parseFromString(decodeText(source), \"text/xml\");\n    if (document.getElementsByTagName(\"rss\").length > 0) {\n      return this.parseRss(document);\n    }\n    if (document.getElementsByTagName(\"feed\")[0]?.getElementsByTagName(\"entry\").length) {\n      return this.parseAtom(document);\n    }\n    return null;\n  }\n\n  private parseAtom(document: Document): ConverterResult {\n    const root = document.getElementsByTagName(\"feed\")[0];\n    if (!root) return null;\n\n    const title = getText(root, \"title\");\n    const subtitle = getText(root, \"subtitle\");\n    const sections = [`# ${title ?? \"Feed\"}`];\n    if (subtitle) sections.push(subtitle);\n\n    for (const entry of [...root.getElementsByTagName(\"entry\")]) {\n      const entryTitle = getText(entry, \"title\");\n      const updated = getText(entry, \"updated\");\n      const summary = getText(entry, \"summary\");\n      const content = getText(entry, \"content\");\n      const lines: string[] = [];\n      if (entryTitle) lines.push(`## ${entryTitle}`);\n      if (updated) lines.push(`Updated on: ${updated}`);\n      if (summary) lines.push(parseContent(summary));\n      if (content) lines.push(parseContent(content));\n      sections.push(lines.join(\"\\n\"));\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n\n  private parseRss(document: Document): ConverterResult {\n    const channel = document.getElementsByTagName(\"rss\")[0]?.getElementsByTagName(\"channel\")[0];\n    if (!channel) return null;\n\n    const title = getText(channel, \"title\");\n    const description = getText(channel, \"description\");\n    const sections: string[] = title ? [`# ${title}`] : [];\n    if (description) sections.push(description);\n\n    for (const item of [...channel.getElementsByTagName(\"item\")]) {\n      const itemTitle = getText(item, \"title\");\n      const published = getText(item, \"pubDate\");\n      const itemDescription = getText(item, \"description\");\n      const content = getText(item, \"content:encoded\");\n      const lines: string[] = [];\n      if (itemTitle) lines.push(`## ${itemTitle}`);\n      if (published) lines.push(`Published on: ${published}`);\n      if (itemDescription) lines.push(parseContent(itemDescription));\n      if (content) lines.push(parseContent(content));\n      sections.push(lines.join(\"\\n\"));\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n}\n\nfunction parseContent(content: string): string {\n  return new CustomTurnDown().convertSoup(parseHtmlDocument(content).documentElement);\n}\n\nfunction getText(element: Element, tagName: string): string | null {\n  const node = element.getElementsByTagName(tagName)[0];\n  return node?.textContent?.trim() || null;\n}\n","import { decodeText } from \"../bytes\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class YouTubeConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if (\n      (extension !== \".html\" && extension !== \".htm\") ||\n      !options.url?.startsWith(\"https://www.youtube.com/watch?\")\n    ) {\n      return null;\n    }\n\n    const document = parseHtmlDocument(decodeText(source));\n    const metadata: Record<string, string> = { title: document.title };\n\n    document.querySelectorAll(\"meta\").forEach((meta) => {\n      const content = meta.getAttribute(\"content\");\n      const key =\n        meta.getAttribute(\"itemprop\") ?? meta.getAttribute(\"property\") ?? meta.getAttribute(\"name\");\n      if (key && content) {\n        metadata[key] = content;\n      }\n    });\n\n    try {\n      for (const script of document.querySelectorAll(\"script\")) {\n        const content = script.textContent ?? \"\";\n        if (!content.includes(\"ytInitialData\")) {\n          continue;\n        }\n        const start = content.indexOf(\"{\");\n        const end = content.lastIndexOf(\"}\");\n        if (start >= 0 && end > start) {\n          const description = findKey(\n            JSON.parse(content.slice(start, end + 1)),\n            \"attributedDescriptionBodyText\"\n          );\n          if (isRecord(description) && typeof description.content === \"string\") {\n            metadata.description = description.content;\n          }\n        }\n        break;\n      }\n    } catch {\n      // Meta description remains available when the page implementation changes.\n    }\n\n    const title = first(metadata, [\"title\", \"og:title\", \"name\"]) || document.title || null;\n    const lines = [\"# YouTube\"];\n    if (title) {\n      lines.push(`## ${title}`);\n    }\n\n    const stats: string[] = [];\n    const views = first(metadata, [\"interactionCount\"]);\n    const keywords = first(metadata, [\"keywords\"]);\n    const runtime = first(metadata, [\"duration\"]);\n    if (views) stats.push(`- **Views:** ${views}`);\n    if (keywords) stats.push(`- **Keywords:** ${keywords}`);\n    if (runtime) stats.push(`- **Runtime:** ${runtime}`);\n    if (stats.length > 0) {\n      lines.push(`### Video Metadata\\n${stats.join(\"\\n\")}`);\n    }\n\n    const description = first(metadata, [\"description\", \"og:description\"]);\n    if (description) {\n      lines.push(`### Description\\n${description}`);\n    }\n\n    const markdown = lines.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n}\n\nfunction first(metadata: Record<string, string>, keys: string[]): string | null {\n  for (const key of keys) {\n    if (metadata[key]) return metadata[key];\n  }\n  return null;\n}\n\nfunction findKey(value: unknown, key: string): unknown {\n  if (Array.isArray(value)) {\n    for (const item of value) {\n      const match = findKey(item, key);\n      if (match !== undefined) return match;\n    }\n  } else if (isRecord(value)) {\n    if (key in value) return value[key];\n    for (const item of Object.values(value)) {\n      const match = findKey(item, key);\n      if (match !== undefined) return match;\n    }\n  }\n  return undefined;\n}\n\nfunction isRecord(value: unknown): value is Record<string, unknown> {\n  return typeof value === \"object\" && value !== null;\n}\n","import { unzipSync } from \"fflate\";\nimport { extensionFromPathname } from \"../mime\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class ZipConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".zip\") {\n      return null;\n    }\n\n    const files = unzipSync(source, {\n      filter: (file) => !file.name.endsWith(\"/\")\n    });\n    const converters = options._parent_converters ?? [];\n    const zipName = options.url ?? \"archive.zip\";\n    const sections = [`Content from the zip file \\`${zipName}\\`:`];\n\n    for (const [name, bytes] of Object.entries(files)) {\n      const result = await convertEntry(bytes, name, converters, options);\n      if (result) {\n        sections.push(`## File: ${name}\\n\\n${result.markdown}`);\n      }\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return { title: null, markdown, text_content: markdown };\n  }\n}\n\nasync function convertEntry(\n  source: Uint8Array,\n  name: string,\n  converters: DocumentConverter[],\n  options: ConverterOptions\n): Promise<ConverterResult> {\n  const extension = extensionFromPathname(name);\n  if (!extension) {\n    return null;\n  }\n\n  const entryOptions: ConverterOptions = {\n    ...options,\n    file_extension: extension,\n    url: undefined,\n    _parent_converters: converters\n  };\n\n  for (const converter of converters) {\n    try {\n      const result = await converter.convert(source, entryOptions);\n      if (result) {\n        return result;\n      }\n    } catch {\n      // Match the original ZipConverter: skip entries that fail conversion.\n    }\n  }\n\n  return null;\n}\n","import { isBlob, toUint8Array } from \"./bytes\";\nimport { BingSerpConverter } from \"./converters/bingserp\";\nimport { DocxConverter } from \"./converters/docx\";\nimport { HtmlConverter } from \"./converters/html\";\nimport { IpynbConverter } from \"./converters/ipynb\";\nimport { PdfConverter } from \"./converters/pdf\";\nimport { PlainTextConverter } from \"./converters/plain-text\";\nimport { WikipediaConverter } from \"./converters/wikipedia\";\nimport { XlsxConverter } from \"./converters/xlsx\";\nimport { RSSConverter } from \"./converters/xml-rss-atom\";\nimport { YouTubeConverter } from \"./converters/youtube\";\nimport { ZipConverter } from \"./converters/zip\";\nimport { extensionFromMime, extensionFromPathname } from \"./mime\";\nimport type { ByteSource, ConverterOptions, ConverterResult, DocumentConverter } from \"./types\";\n\nexport type ConversionSource = ByteSource | Response | URL | string;\n\nexport class MarkItDown {\n  private readonly converters: DocumentConverter[] = [];\n\n  constructor() {\n    this.registerConverter(new PlainTextConverter());\n    this.registerConverter(new HtmlConverter());\n    this.registerConverter(new ZipConverter());\n    this.registerConverter(new RSSConverter());\n    this.registerConverter(new WikipediaConverter());\n    this.registerConverter(new YouTubeConverter());\n    this.registerConverter(new BingSerpConverter());\n    this.registerConverter(new DocxConverter());\n    this.registerConverter(new XlsxConverter());\n    this.registerConverter(new PdfConverter());\n    this.registerConverter(new IpynbConverter());\n  }\n\n  /** Convert a URL, response, blob, array buffer, data view, or typed array. */\n  async convert(\n    source: ConversionSource,\n    options: ConverterOptions = {}\n  ): Promise<ConverterResult> {\n    if (typeof source === \"string\" || source instanceof URL) {\n      return this.convertUrl(source.toString(), options);\n    }\n\n    if (isResponse(source)) {\n      return this.convertResponse(source, options);\n    }\n\n    const extensions = new Set<string>();\n    this.addExtension(extensions, options.file_extension);\n\n    if (isBlob(source)) {\n      this.addExtension(extensions, extensionFromMime(source.type));\n      if (\"name\" in source && typeof source.name === \"string\") {\n        this.addExtension(extensions, extensionFromPathname(source.name));\n      }\n    }\n\n    if (extensions.size === 0) {\n      throw new Error(\n        \"Could not determine the file type. Provide `file_extension` when converting bytes.\"\n      );\n    }\n\n    return this.convertBytes(await toUint8Array(source), extensions, options);\n  }\n\n  /**\n   * @deprecated Pass the byte source directly to `convert`.\n   */\n  async convertBuffer(\n    source: ByteSource,\n    options: ConverterOptions & { file_extension: string }\n  ): Promise<ConverterResult> {\n    return this.convert(source, options);\n  }\n\n  private async convertUrl(\n    source: string,\n    { fetch: fetchImplementation = globalThis.fetch, ...options }: ConverterOptions\n  ): Promise<ConverterResult> {\n    const url = new URL(source);\n    if (url.protocol !== \"http:\" && url.protocol !== \"https:\") {\n      throw new Error(`Unsupported URL protocol: ${url.protocol}`);\n    }\n    if (!fetchImplementation) {\n      throw new Error(\"No fetch implementation is available.\");\n    }\n\n    const response = await fetchImplementation(url);\n    if (!response.ok) {\n      throw new Error(`Failed to fetch URL: ${url}, status: ${response.status}`);\n    }\n\n    return this.convertResponse(response, { ...options, url: url.toString() });\n  }\n\n  private async convertResponse(\n    response: Response,\n    options: ConverterOptions\n  ): Promise<ConverterResult> {\n    const extensions = new Set<string>();\n    this.addExtension(extensions, options.file_extension);\n    this.addExtension(extensions, extensionFromMime(response.headers.get(\"content-type\")));\n\n    const disposition = response.headers.get(\"content-disposition\") ?? \"\";\n    const filename = disposition.match(/filename\\*?=(?:UTF-8''|[\"']?)([^;\"']+)/i)?.[1];\n    if (filename) {\n      this.addExtension(extensions, extensionFromPathname(decodeURIComponent(filename.trim())));\n    }\n\n    const responseUrl = options.url || response.url;\n    if (responseUrl) {\n      this.addExtension(extensions, extensionFromPathname(new URL(responseUrl).pathname));\n    }\n\n    if (extensions.size === 0) {\n      throw new Error(\n        \"Could not determine the file type. Provide `file_extension` or a Content-Type header.\"\n      );\n    }\n\n    return this.convertBytes(new Uint8Array(await response.arrayBuffer()), extensions, {\n      ...options,\n      url: responseUrl\n    });\n  }\n\n  private async convertBytes(\n    source: Uint8Array,\n    extensions: Set<string>,\n    options: ConverterOptions\n  ): Promise<ConverterResult> {\n    let lastError: unknown;\n\n    for (const extension of extensions) {\n      for (const converter of this.converters) {\n        try {\n          const result = await converter.convert(source, {\n            ...options,\n            file_extension: extension,\n            _parent_converters: this.converters\n          });\n          if (result) {\n            result.markdown = result.markdown\n              .replace(/\\r\\n|\\r|\\n/g, \"\\n\")\n              .trim()\n              .replace(/\\n{3,}/g, \"\\n\\n\");\n            result.text_content = result.markdown;\n            return result;\n          }\n        } catch (error) {\n          lastError = error;\n        }\n      }\n    }\n\n    const formats = [...extensions].join(\", \");\n    if (lastError) {\n      throw new Error(`Could not convert ${formats} to Markdown: ${String(lastError)}`);\n    }\n    throw new Error(`Unsupported file type: ${formats}.`);\n  }\n\n  private addExtension(extensions: Set<string>, extension: string | null | undefined): void {\n    if (!extension) {\n      return;\n    }\n    extensions.add(\n      extension.startsWith(\".\") ? extension.toLowerCase() : `.${extension.toLowerCase()}`\n    );\n  }\n\n  private registerConverter(converter: DocumentConverter): void {\n    this.converters.unshift(converter);\n  }\n}\n\nfunction isResponse(value: unknown): value is Response {\n  return typeof Response !== \"undefined\" && value instanceof Response;\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAEA,eAAsB,aAAa,QAAyC;CAC1E,IAAI,kBAAkB,YACpB,OAAO;CAGT,IAAI,OAAO,MAAM,GACf,OAAO,IAAI,WAAW,MAAM,OAAO,YAAY,CAAC;CAGlD,IAAI,YAAY,OAAO,MAAM,GAC3B,OAAO,IAAI,WAAW,OAAO,QAAQ,OAAO,YAAY,OAAO,UAAU;CAG3E,OAAO,IAAI,WAAW,MAAM;AAC9B;AAEA,SAAgB,cAAc,QAAiC;CAC7D,OAAO,OAAO,MAAM,CAAC,CAAC;AACxB;AAEA,SAAgB,WAAW,QAA4B;CACrD,OAAO,IAAI,YAAY,CAAC,CAAC,OAAO,MAAM;AACxC;AAEA,SAAgB,OAAO,OAA+B;CACpD,OAAO,OAAO,SAAS,eAAe,iBAAiB;AACzD;;;ACzBA,IAAa,iBAAb,MAA4B;CAC1B,YAAY,UAAiC;EAC3C,MAAM,UAAU,IAAIA,SAAAA,QAAgB,EAAE,cAAc,MAAM,CAAC;EAC3D,QAAQ,IAAIC,4BAAAA,QAAkB,GAAG;EAEjC,QAAQ,QAAQ,eAAe;GAC7B,QAAQ,CAAC,GAAG;GACZ,YAAY,SAAS,MAAM;IACzB,IAAI,CAAC,SAAS,OAAO;IAErB,MAAM,SAAS,QAAQ,WAAW,GAAG,IAAI,MAAM;IAC/C,MAAM,SAAS,QAAQ,SAAS,GAAG,IAAI,MAAM;IAC7C,MAAM,OAAO,QAAQ,KAAK,CAAC,CAAC,QAAQ,WAAW,EAAE;IACjD,IAAI,CAAC,MAAM,OAAO;IAElB,MAAM,UAAU;IAChB,MAAM,OAAO,QAAQ,aAAa,MAAM;IACxC,IAAI,QAAQ,QAAQ,aAAa,OAAO,KAAK;IAC7C,IAAI,CAAC,MAAM,OAAO,GAAG,SAAS,OAAO;IAErC,IAAI;KACF,MAAM,MAAM,IAAI,IAAI,IAAI;KACxB,IAAI,IAAI,aAAa,YAAY,IAAI,aAAa,SAChD,OAAO,GAAG,SAAS,OAAO;IAE9B,QAAQ;KACN,IACE,CAAC,KAAK,WAAW,GAAG,KACpB,CAAC,KAAK,WAAW,GAAG,KACpB,CAAC,KAAK,WAAW,IAAI,KACrB,CAAC,KAAK,WAAW,KAAK,GAEtB,OAAO,GAAG,SAAS,OAAO;IAE9B;IAEA,IAAI,KAAK,QAAQ,QAAQ,GAAG,MAAM,QAAQ,CAAC,OACzC,OAAO,IAAI,KAAK;IAElB,IAAI,CAAC,OAAO,QAAQ;IACpB,OAAO,GAAG,OAAO,GAAG,KAAK,IAAI,OAAO,QAAQ,KAAK,MAAM,KAAK,GAAG,GAAG;GACpE;EACF,CAAC;EAED,QAAQ,QAAQ,YAAY;GAC1B,QAAQ,CAAC,KAAK;GACd,YAAY,UAAU,MAAM;IAC1B,MAAM,UAAU;IAChB,MAAM,MAAM,QAAQ,aAAa,KAAK,KAAK;IAC3C,IAAI,SAAS,QAAQ,aAAa,KAAK,KAAK;IAC5C,MAAM,QAAQ,QAAQ,aAAa,OAAO;IAC1C,IAAI,OAAO,WAAW,OAAO,GAC3B,SAAS,GAAG,OAAO,MAAM,KAAK,CAAC,CAAC,CAAC,GAAG;IAEtC,OAAO,KAAK,IAAI,IAAI,SAAS,QAAQ,KAAK,MAAM,KAAK,GAAG;GAC1D;EACF,CAAC;EAED,OAAO,QAAQ,SAAS,QAAgC;CAC1D;AACF;;;AC7DA,SAAgB,kBAAkB,MAAwB;CACxD,QAAA,GAAOC,SAAAA,UAAAA,CAAU,IAAI,CAAC,CAAC;AACzB;;;ACCA,MAAM,kBAAkB;AACxB,MAAM,kBAAkB;AAExB,IAAa,oBAAb,MAA4D;CAC1D,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,MAAM,MAAM,QAAQ,OAAO;EAC3B,IAAK,cAAc,WAAW,cAAc,UAAW,CAAC,gBAAgB,KAAK,GAAG,GAC9E,OAAO;EAGT,MAAM,WAAW,kBAAkB,WAAW,MAAM,CAAC;EACrD,MAAM,QAAQ,IAAI,IAAI,GAAG,CAAC,CAAC,aAAa,IAAI,GAAG,KAAK;EAEpD,SAAS,iBAAiB,OAAO,CAAC,CAAC,SAAS,YAAY;GACtD,IAAI,QAAQ,aACV,QAAQ,eAAe;EAE3B,CAAC;EACD,SAAS,iBAAiB,gBAAgB,CAAC,CAAC,SAAS,YAAY;GAC/D,QAAQ,OAAO;EACjB,CAAC;EAED,MAAM,WAAW,IAAI,eAAe;EAyBpC,MAAM,WAAW,yBAAyB,MAAM,oCAxBhC,CAAC,GAAG,SAAS,iBAAiB,SAAS,CAAC,CAAC,CAAC,KAAK,WAAW;GACxE,OAAO,iBAAiB,SAAS,CAAC,CAAC,SAAS,WAAW;IACrD,MAAM,OAAO,OAAO,aAAa,MAAM;IACvC,IAAI,CAAC,MACH;IAEF,IAAI;KACF,MAAM,UAAU,IAAI,IAAI,MAAM,GAAG,CAAC,CAAC,aAAa,IAAI,GAAG;KACvD,IAAI,SACF,OAAO,aAAa,QAAQ,cAAc,OAAO,CAAC;IAEtD,QAAQ,CAER;GACF,CAAC;GAED,OAAO,SACJ,YAAY,MAAM,CAAC,CACnB,MAAM,KAAK,CAAC,CACZ,KAAK,SAAS,KAAK,KAAK,CAAC,CAAC,CAC1B,OAAO,OAAO,CAAC,CACf,KAAK,IAAI;EACd,CAE0F,CAAC,CAAC,KAAK,MAAM;EACvG,OAAO;GAAE,OAAO,SAAS,SAAS;GAAM;GAAU,cAAc;EAAS;CAC3E;AACF;AAEA,SAAS,cAAc,SAAyB;CAC9C,MAAM,QAAQ,QAAQ,MAAM,CAAC,CAAC,CAAC,QAAQ,MAAM,GAAG,CAAC,CAAC,QAAQ,MAAM,GAAG,CAAC,CAAC,QAAQ,OAAO,EAAE;CACtF,MAAM,QAAkB,CAAC;CACzB,IAAI,OAAO;CACX,IAAI,WAAW;CAEf,KAAK,MAAM,aAAa,OAAO;EAC7B,MAAM,QAAQ,gBAAgB,QAAQ,SAAS;EAC/C,IAAI,QAAQ,GACV,OAAO;EAET,OAAQ,QAAQ,IAAK;EACrB,YAAY;EACZ,IAAI,YAAY,GAAG;GACjB,YAAY;GACZ,MAAM,KAAM,QAAQ,WAAY,GAAI;EACtC;CACF;CAEA,OAAO,IAAI,YAAY,CAAC,CAAC,OAAO,IAAI,WAAW,KAAK,CAAC;AACvD;;;ACzEA,IAAa,gBAAb,MAAwD;CACtD,MAAM,QAAQ,QAAoB,SAAqD;EACrF,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IAAI,cAAc,WAAW,cAAc,QACzC,OAAO;EAGT,OAAO,KAAK,YAAY,WAAW,MAAM,CAAC;CAC5C;CAEA,MAAM,YAAY,MAAwC;EACxD,MAAM,WAAW,kBAAkB,IAAI;EACvC,SAAS,iBAAiB,eAAe,CAAC,CAAC,SAAS,YAAY;GAC9D,QAAQ,OAAO;EACjB,CAAC;EAED,MAAM,WAAW,IAAI,eAAe,CAAC,CAAC,YAAY,SAAS,QAAQ,QAAQ;EAC3E,OAAO;GACL,OAAO,SAAS,SAAS;GACzB;GACA,cAAc;EAChB;CACF;AACF;;;ACvBA,IAAa,gBAAb,cAAmC,cAAc;CAC/C,MAAM,QAAQ,QAAoB,SAAqD;EACrF,IAAI,QAAQ,gBAAgB,YAAY,MAAM,SAC5C,OAAO;EAGT,MAAM,OAAO,MAAM,QAAA,QAAQ,cACzB,EAAE,aAAa,cAAc,MAAM,EAAE,GACrC;GACE,UAAU,QAAQ;GAClB,yBAAyB,QAAQ;GACjC,wBAAwB,QAAQ;GAChC,uBAAuB,QAAQ;GAC/B,UAAU,QAAQ;GAClB,mBAAmB,QAAQ;EAC7B,CACF;EACA,OAAO,KAAK,YAAY,KAAK,KAAK;CACpC;AACF;;;ACXA,IAAa,iBAAb,MAAyD;CACvD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,IAAI,QAAQ,gBAAgB,YAAY,MAAM,UAC5C,OAAO;EAGT,MAAM,WAAW,KAAK,MAAM,WAAW,MAAM,CAAC;EAC9C,MAAM,SAAmB,CAAC;EAC1B,IAAI,QAAuB,SAAS,UAAU,SAAS;EAEvD,KAAK,MAAM,QAAQ,SAAS,SAAS,CAAC,GAAG;GACvC,MAAM,QAAQ,KAAK,UAAU,CAAC;GAC9B,IAAI,KAAK,cAAc,YAAY;IACjC,OAAO,KAAK,MAAM,KAAK,EAAE,CAAC;IAC1B,UAAA,QACE,MACG,MAAM,SAAS,KAAK,WAAW,IAAI,CAAC,CAAC,EACpC,MAAM,CAAC,CAAC,CACT,KAAK,KAAK;GACjB,OAAO,IAAI,KAAK,cAAc,QAC5B,OAAO,KAAK,iBAAiB,MAAM,KAAK,EAAE,EAAE,SAAS;QAChD,IAAI,KAAK,cAAc,OAC5B,OAAO,KAAK,WAAW,MAAM,KAAK,EAAE,EAAE,SAAS;EAEnD;EAEA,MAAM,WAAW,OAAO,KAAK,MAAM;EACnC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;;;ACvCA,IAAa,eAAb,MAAuD;CACrD,MAAM,QAAQ,QAAoB,SAAqD;EACrF,IAAI,QAAQ,gBAAgB,YAAY,MAAM,QAC5C,OAAO;EAGT,MAAM,MAAM,OAAA,GAAMC,MAAAA,iBAAAA,CAAiB,OAAO,MAAM,CAAC;EACjD,IAAI;GACF,MAAM,CAAC,EAAE,QAAQ,EAAE,UAAU,MAAM,QAAQ,IAAI,EAAA,GAC7CC,MAAAA,YAAAA,CAAY,KAAK,EAAE,YAAY,KAAK,CAAC,IAAA,GACrCC,MAAAA,QAAAA,CAAQ,GAAG,CACb,CAAC;GAGD,OAAO;IAAE,OAFK,OAAO,KAAK,UAAU,YAAY,KAAK,MAAM,KAAK,IAAI,KAAK,QAAQ;IAEjE,UAAU;IAAM,cAAc;GAAK;EACrD,UAAU;GACR,MAAM,IAAI,YAAY,QAAQ;EAChC;CACF;AACF;;;ACtBA,MAAM,kBAA0C;CAC9C,wBAAwB;CACxB,oBAAoB;CACpB,mBAAmB;CACnB,mBAAmB;CACnB,uBAAuB;CACvB,qBAAqB;CACrB,qEAAqE;CACrE,2EAA2E;CAC3E,mBAAmB;CACnB,YAAY;CACZ,aAAa;CACb,iBAAiB;CACjB,cAAc;CACd,YAAY;AACd;AAEA,MAAM,kCAAkB,IAAI,IAAI;CAAC;CAAQ;CAAS;CAAQ;CAAa;CAAO;CAAS;AAAM,CAAC;AAE9F,SAAgB,kBAAkB,aAAuD;CACvF,IAAI,CAAC,aACH,OAAO;CAGT,MAAM,OAAO,YAAY,MAAM,KAAK,CAAC,CAAC,CAAC,EAAE,EAAE,KAAK,CAAC,CAAC,YAAY,KAAK;CACnE,OAAO,gBAAgB,UAAU,KAAK,WAAW,OAAO,IAAI,SAAS;AACvE;AAEA,SAAgB,gBAAgB,WAA4B;CAC1D,OAAO,gBAAgB,IAAI,UAAU,YAAY,CAAC;AACpD;AAEA,SAAgB,sBAAsB,UAAiC;CACrE,MAAM,WAAW,SAAS,MAAM,GAAG,CAAC,CAAC,IAAI,KAAK;CAC9C,MAAM,MAAM,SAAS,YAAY,GAAG;CACpC,OAAO,MAAM,IAAI,SAAS,MAAM,GAAG,CAAC,CAAC,YAAY,IAAI;AACvD;;;AChCA,IAAa,qBAAb,MAA6D;CAC3D,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,IAAI,CAAC,gBAAgB,QAAQ,kBAAkB,EAAE,GAC/C,OAAO;EAGT,MAAM,UAAU,WAAW,MAAM;EACjC,OAAO;GAAE,OAAO;GAAM,UAAU;GAAS,cAAc;EAAQ;CACjE;AACF;;;ACRA,MAAM,gBAAgB;AAEtB,IAAa,qBAAb,MAA6D;CAC3D,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IAAK,cAAc,WAAW,cAAc,UAAW,CAAC,cAAc,KAAK,QAAQ,OAAO,EAAE,GAC1F,OAAO;EAGT,MAAM,WAAW,kBAAkB,WAAW,MAAM,CAAC;EACrD,SAAS,iBAAiB,eAAe,CAAC,CAAC,SAAS,YAAY;GAC9D,QAAQ,OAAO;EACjB,CAAC;EAED,MAAM,OAAO,SAAS,cAAc,qBAAqB;EAEzD,MAAM,QADY,SAAS,cAAc,yBAAyB,CAAC,EAAE,aAAa,KAAK,KAC5D,SAAS,SAAS;EAC7C,MAAM,WAAW,OACb,KAAK,SAAS,YAAY,MAAM,IAAI,eAAe,CAAC,CAAC,YAAY,IAAI,MACrE,IAAI,eAAe,CAAC,CAAC,YAAY,QAAQ;EAE7C,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;;;ACxBA,IAAa,gBAAb,cAAmC,cAAc;CAC/C,MAAM,QAAQ,QAAoB,SAAqD;EACrF,IAAI,QAAQ,gBAAgB,YAAY,MAAM,SAC5C,OAAO;EAGT,MAAM,WAAWC,KAAK,KAAK,QAAQ,EAAE,MAAM,QAAQ,CAAC;EACpD,MAAM,WAAqB,CAAC;EAE5B,KAAK,MAAM,aAAa,SAAS,YAAY;GAC3C,MAAM,QAAQ,SAAS,OAAO;GAC9B,IAAI,QAAQ,SAAS;IACnB,MAAM,SAAS,MAAM,KAAK,YAAYA,KAAK,MAAM,cAAc,KAAK,CAAC;IACrE,SAAS,KAAK,MAAM,UAAU,IAAI,QAAQ,SAAS,KAAK,KAAK,IAAI;GACnE;EACF;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GACL,OAAO,SAAS,OAAO,SAAS;GAChC;GACA,cAAc;EAChB;CACF;AACF;;;ACrBA,IAAa,eAAb,MAAuD;CACrD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IAAI,CAAC;GAAC;GAAQ;GAAQ;EAAO,CAAC,CAAC,SAAS,SAAS,GAC/C,OAAO;EAGT,MAAM,WAAW,IAAIC,eAAAA,UAAU,CAAC,CAAC,gBAAgB,WAAW,MAAM,GAAG,UAAU;EAC/E,IAAI,SAAS,qBAAqB,KAAK,CAAC,CAAC,SAAS,GAChD,OAAO,KAAK,SAAS,QAAQ;EAE/B,IAAI,SAAS,qBAAqB,MAAM,CAAC,CAAC,EAAE,EAAE,qBAAqB,OAAO,CAAC,CAAC,QAC1E,OAAO,KAAK,UAAU,QAAQ;EAEhC,OAAO;CACT;CAEA,UAAkB,UAAqC;EACrD,MAAM,OAAO,SAAS,qBAAqB,MAAM,CAAC,CAAC;EACnD,IAAI,CAAC,MAAM,OAAO;EAElB,MAAM,QAAQ,QAAQ,MAAM,OAAO;EACnC,MAAM,WAAW,QAAQ,MAAM,UAAU;EACzC,MAAM,WAAW,CAAC,KAAK,SAAS,QAAQ;EACxC,IAAI,UAAU,SAAS,KAAK,QAAQ;EAEpC,KAAK,MAAM,SAAS,CAAC,GAAG,KAAK,qBAAqB,OAAO,CAAC,GAAG;GAC3D,MAAM,aAAa,QAAQ,OAAO,OAAO;GACzC,MAAM,UAAU,QAAQ,OAAO,SAAS;GACxC,MAAM,UAAU,QAAQ,OAAO,SAAS;GACxC,MAAM,UAAU,QAAQ,OAAO,SAAS;GACxC,MAAM,QAAkB,CAAC;GACzB,IAAI,YAAY,MAAM,KAAK,MAAM,YAAY;GAC7C,IAAI,SAAS,MAAM,KAAK,eAAe,SAAS;GAChD,IAAI,SAAS,MAAM,KAAK,aAAa,OAAO,CAAC;GAC7C,IAAI,SAAS,MAAM,KAAK,aAAa,OAAO,CAAC;GAC7C,SAAS,KAAK,MAAM,KAAK,IAAI,CAAC;EAChC;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;CAEA,SAAiB,UAAqC;EACpD,MAAM,UAAU,SAAS,qBAAqB,KAAK,CAAC,CAAC,EAAE,EAAE,qBAAqB,SAAS,CAAC,CAAC;EACzF,IAAI,CAAC,SAAS,OAAO;EAErB,MAAM,QAAQ,QAAQ,SAAS,OAAO;EACtC,MAAM,cAAc,QAAQ,SAAS,aAAa;EAClD,MAAM,WAAqB,QAAQ,CAAC,KAAK,OAAO,IAAI,CAAC;EACrD,IAAI,aAAa,SAAS,KAAK,WAAW;EAE1C,KAAK,MAAM,QAAQ,CAAC,GAAG,QAAQ,qBAAqB,MAAM,CAAC,GAAG;GAC5D,MAAM,YAAY,QAAQ,MAAM,OAAO;GACvC,MAAM,YAAY,QAAQ,MAAM,SAAS;GACzC,MAAM,kBAAkB,QAAQ,MAAM,aAAa;GACnD,MAAM,UAAU,QAAQ,MAAM,iBAAiB;GAC/C,MAAM,QAAkB,CAAC;GACzB,IAAI,WAAW,MAAM,KAAK,MAAM,WAAW;GAC3C,IAAI,WAAW,MAAM,KAAK,iBAAiB,WAAW;GACtD,IAAI,iBAAiB,MAAM,KAAK,aAAa,eAAe,CAAC;GAC7D,IAAI,SAAS,MAAM,KAAK,aAAa,OAAO,CAAC;GAC7C,SAAS,KAAK,MAAM,KAAK,IAAI,CAAC;EAChC;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;AAEA,SAAS,aAAa,SAAyB;CAC7C,OAAO,IAAI,eAAe,CAAC,CAAC,YAAY,kBAAkB,OAAO,CAAC,CAAC,eAAe;AACpF;AAEA,SAAS,QAAQ,SAAkB,SAAgC;CAEjE,OADa,QAAQ,qBAAqB,OAAO,CAAC,CAAC,EACxC,EAAE,aAAa,KAAK,KAAK;AACtC;;;AChFA,IAAa,mBAAb,MAA2D;CACzD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IACG,cAAc,WAAW,cAAc,UACxC,CAAC,QAAQ,KAAK,WAAW,gCAAgC,GAEzD,OAAO;EAGT,MAAM,WAAW,kBAAkB,WAAW,MAAM,CAAC;EACrD,MAAM,WAAmC,EAAE,OAAO,SAAS,MAAM;EAEjE,SAAS,iBAAiB,MAAM,CAAC,CAAC,SAAS,SAAS;GAClD,MAAM,UAAU,KAAK,aAAa,SAAS;GAC3C,MAAM,MACJ,KAAK,aAAa,UAAU,KAAK,KAAK,aAAa,UAAU,KAAK,KAAK,aAAa,MAAM;GAC5F,IAAI,OAAO,SACT,SAAS,OAAO;EAEpB,CAAC;EAED,IAAI;GACF,KAAK,MAAM,UAAU,SAAS,iBAAiB,QAAQ,GAAG;IACxD,MAAM,UAAU,OAAO,eAAe;IACtC,IAAI,CAAC,QAAQ,SAAS,eAAe,GACnC;IAEF,MAAM,QAAQ,QAAQ,QAAQ,GAAG;IACjC,MAAM,MAAM,QAAQ,YAAY,GAAG;IACnC,IAAI,SAAS,KAAK,MAAM,OAAO;KAC7B,MAAM,cAAc,QAClB,KAAK,MAAM,QAAQ,MAAM,OAAO,MAAM,CAAC,CAAC,GACxC,+BACF;KACA,IAAI,SAAS,WAAW,KAAK,OAAO,YAAY,YAAY,UAC1D,SAAS,cAAc,YAAY;IAEvC;IACA;GACF;EACF,QAAQ,CAER;EAEA,MAAM,QAAQ,MAAM,UAAU;GAAC;GAAS;GAAY;EAAM,CAAC,KAAK,SAAS,SAAS;EAClF,MAAM,QAAQ,CAAC,WAAW;EAC1B,IAAI,OACF,MAAM,KAAK,MAAM,OAAO;EAG1B,MAAM,QAAkB,CAAC;EACzB,MAAM,QAAQ,MAAM,UAAU,CAAC,kBAAkB,CAAC;EAClD,MAAM,WAAW,MAAM,UAAU,CAAC,UAAU,CAAC;EAC7C,MAAM,UAAU,MAAM,UAAU,CAAC,UAAU,CAAC;EAC5C,IAAI,OAAO,MAAM,KAAK,gBAAgB,OAAO;EAC7C,IAAI,UAAU,MAAM,KAAK,mBAAmB,UAAU;EACtD,IAAI,SAAS,MAAM,KAAK,kBAAkB,SAAS;EACnD,IAAI,MAAM,SAAS,GACjB,MAAM,KAAK,uBAAuB,MAAM,KAAK,IAAI,GAAG;EAGtD,MAAM,cAAc,MAAM,UAAU,CAAC,eAAe,gBAAgB,CAAC;EACrE,IAAI,aACF,MAAM,KAAK,oBAAoB,aAAa;EAG9C,MAAM,WAAW,MAAM,KAAK,MAAM;EAClC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;AAEA,SAAS,MAAM,UAAkC,MAA+B;CAC9E,KAAK,MAAM,OAAO,MAChB,IAAI,SAAS,MAAM,OAAO,SAAS;CAErC,OAAO;AACT;AAEA,SAAS,QAAQ,OAAgB,KAAsB;CACrD,IAAI,MAAM,QAAQ,KAAK,GACrB,KAAK,MAAM,QAAQ,OAAO;EACxB,MAAM,QAAQ,QAAQ,MAAM,GAAG;EAC/B,IAAI,UAAU,KAAA,GAAW,OAAO;CAClC;MACK,IAAI,SAAS,KAAK,GAAG;EAC1B,IAAI,OAAO,OAAO,OAAO,MAAM;EAC/B,KAAK,MAAM,QAAQ,OAAO,OAAO,KAAK,GAAG;GACvC,MAAM,QAAQ,QAAQ,MAAM,GAAG;GAC/B,IAAI,UAAU,KAAA,GAAW,OAAO;EAClC;CACF;AAEF;AAEA,SAAS,SAAS,OAAkD;CAClE,OAAO,OAAO,UAAU,YAAY,UAAU;AAChD;;;ACjGA,IAAa,eAAb,MAAuD;CACrD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,IAAI,QAAQ,gBAAgB,YAAY,MAAM,QAC5C,OAAO;EAGT,MAAM,SAAA,GAAQC,OAAAA,UAAAA,CAAU,QAAQ,EAC9B,SAAS,SAAS,CAAC,KAAK,KAAK,SAAS,GAAG,EAC3C,CAAC;EACD,MAAM,aAAa,QAAQ,sBAAsB,CAAC;EAElD,MAAM,WAAW,CAAC,+BADF,QAAQ,OAAO,cAC0B,IAAI;EAE7D,KAAK,MAAM,CAAC,MAAM,UAAU,OAAO,QAAQ,KAAK,GAAG;GACjD,MAAM,SAAS,MAAM,aAAa,OAAO,MAAM,YAAY,OAAO;GAClE,IAAI,QACF,SAAS,KAAK,YAAY,KAAK,MAAM,OAAO,UAAU;EAE1D;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GAAE,OAAO;GAAM;GAAU,cAAc;EAAS;CACzD;AACF;AAEA,eAAe,aACb,QACA,MACA,YACA,SAC0B;CAC1B,MAAM,YAAY,sBAAsB,IAAI;CAC5C,IAAI,CAAC,WACH,OAAO;CAGT,MAAM,eAAiC;EACrC,GAAG;EACH,gBAAgB;EAChB,KAAK,KAAA;EACL,oBAAoB;CACtB;CAEA,KAAK,MAAM,aAAa,YACtB,IAAI;EACF,MAAM,SAAS,MAAM,UAAU,QAAQ,QAAQ,YAAY;EAC3D,IAAI,QACF,OAAO;CAEX,QAAQ,CAER;CAGF,OAAO;AACT;;;AC1CA,IAAa,aAAb,MAAwB;CAGtB,cAAc;EAFd,KAAiB,aAAkC,CAAC;EAGlD,KAAK,kBAAkB,IAAI,mBAAmB,CAAC;EAC/C,KAAK,kBAAkB,IAAI,cAAc,CAAC;EAC1C,KAAK,kBAAkB,IAAI,aAAa,CAAC;EACzC,KAAK,kBAAkB,IAAI,aAAa,CAAC;EACzC,KAAK,kBAAkB,IAAI,mBAAmB,CAAC;EAC/C,KAAK,kBAAkB,IAAI,iBAAiB,CAAC;EAC7C,KAAK,kBAAkB,IAAI,kBAAkB,CAAC;EAC9C,KAAK,kBAAkB,IAAI,cAAc,CAAC;EAC1C,KAAK,kBAAkB,IAAI,cAAc,CAAC;EAC1C,KAAK,kBAAkB,IAAI,aAAa,CAAC;EACzC,KAAK,kBAAkB,IAAI,eAAe,CAAC;CAC7C;;CAGA,MAAM,QACJ,QACA,UAA4B,CAAC,GACH;EAC1B,IAAI,OAAO,WAAW,YAAY,kBAAkB,KAClD,OAAO,KAAK,WAAW,OAAO,SAAS,GAAG,OAAO;EAGnD,IAAI,WAAW,MAAM,GACnB,OAAO,KAAK,gBAAgB,QAAQ,OAAO;EAG7C,MAAM,6BAAa,IAAI,IAAY;EACnC,KAAK,aAAa,YAAY,QAAQ,cAAc;EAEpD,IAAI,OAAO,MAAM,GAAG;GAClB,KAAK,aAAa,YAAY,kBAAkB,OAAO,IAAI,CAAC;GAC5D,IAAI,UAAU,UAAU,OAAO,OAAO,SAAS,UAC7C,KAAK,aAAa,YAAY,sBAAsB,OAAO,IAAI,CAAC;EAEpE;EAEA,IAAI,WAAW,SAAS,GACtB,MAAM,IAAI,MACR,oFACF;EAGF,OAAO,KAAK,aAAa,MAAM,aAAa,MAAM,GAAG,YAAY,OAAO;CAC1E;;;;CAKA,MAAM,cACJ,QACA,SAC0B;EAC1B,OAAO,KAAK,QAAQ,QAAQ,OAAO;CACrC;CAEA,MAAc,WACZ,QACA,EAAE,OAAO,sBAAsB,WAAW,OAAO,GAAG,WAC1B;EAC1B,MAAM,MAAM,IAAI,IAAI,MAAM;EAC1B,IAAI,IAAI,aAAa,WAAW,IAAI,aAAa,UAC/C,MAAM,IAAI,MAAM,6BAA6B,IAAI,UAAU;EAE7D,IAAI,CAAC,qBACH,MAAM,IAAI,MAAM,uCAAuC;EAGzD,MAAM,WAAW,MAAM,oBAAoB,GAAG;EAC9C,IAAI,CAAC,SAAS,IACZ,MAAM,IAAI,MAAM,wBAAwB,IAAI,YAAY,SAAS,QAAQ;EAG3E,OAAO,KAAK,gBAAgB,UAAU;GAAE,GAAG;GAAS,KAAK,IAAI,SAAS;EAAE,CAAC;CAC3E;CAEA,MAAc,gBACZ,UACA,SAC0B;EAC1B,MAAM,6BAAa,IAAI,IAAY;EACnC,KAAK,aAAa,YAAY,QAAQ,cAAc;EACpD,KAAK,aAAa,YAAY,kBAAkB,SAAS,QAAQ,IAAI,cAAc,CAAC,CAAC;EAGrF,MAAM,YADc,SAAS,QAAQ,IAAI,qBAAqB,KAAK,GAAA,CACtC,MAAM,yCAAyC,CAAC,GAAG;EAChF,IAAI,UACF,KAAK,aAAa,YAAY,sBAAsB,mBAAmB,SAAS,KAAK,CAAC,CAAC,CAAC;EAG1F,MAAM,cAAc,QAAQ,OAAO,SAAS;EAC5C,IAAI,aACF,KAAK,aAAa,YAAY,sBAAsB,IAAI,IAAI,WAAW,CAAC,CAAC,QAAQ,CAAC;EAGpF,IAAI,WAAW,SAAS,GACtB,MAAM,IAAI,MACR,uFACF;EAGF,OAAO,KAAK,aAAa,IAAI,WAAW,MAAM,SAAS,YAAY,CAAC,GAAG,YAAY;GACjF,GAAG;GACH,KAAK;EACP,CAAC;CACH;CAEA,MAAc,aACZ,QACA,YACA,SAC0B;EAC1B,IAAI;EAEJ,KAAK,MAAM,aAAa,YACtB,KAAK,MAAM,aAAa,KAAK,YAC3B,IAAI;GACF,MAAM,SAAS,MAAM,UAAU,QAAQ,QAAQ;IAC7C,GAAG;IACH,gBAAgB;IAChB,oBAAoB,KAAK;GAC3B,CAAC;GACD,IAAI,QAAQ;IACV,OAAO,WAAW,OAAO,SACtB,QAAQ,eAAe,IAAI,CAAC,CAC5B,KAAK,CAAC,CACN,QAAQ,WAAW,MAAM;IAC5B,OAAO,eAAe,OAAO;IAC7B,OAAO;GACT;EACF,SAAS,OAAO;GACd,YAAY;EACd;EAIJ,MAAM,UAAU,CAAC,GAAG,UAAU,CAAC,CAAC,KAAK,IAAI;EACzC,IAAI,WACF,MAAM,IAAI,MAAM,qBAAqB,QAAQ,gBAAgB,OAAO,SAAS,GAAG;EAElF,MAAM,IAAI,MAAM,0BAA0B,QAAQ,EAAE;CACtD;CAEA,aAAqB,YAAyB,WAA4C;EACxF,IAAI,CAAC,WACH;EAEF,WAAW,IACT,UAAU,WAAW,GAAG,IAAI,UAAU,YAAY,IAAI,IAAI,UAAU,YAAY,GAClF;CACF;CAEA,kBAA0B,WAAoC;EAC5D,KAAK,WAAW,QAAQ,SAAS;CACnC;AACF;AAEA,SAAS,WAAW,OAAmC;CACrD,OAAO,OAAO,aAAa,eAAe,iBAAiB;AAC7D"}
+{"version":3,"file":"index.cjs","names":["TurndownService","turndownPluginGfm","parseHTML","getDocumentProxy","extractText","getMeta","ExcelJS","DOMParser","unzipSync"],"sources":["../src/bytes.ts","../src/custom-turndown.ts","../src/html-document.ts","../src/converters/bingserp.ts","../src/converters/html.ts","../src/converters/docx.ts","../src/converters/ipynb.ts","../src/converters/pdf.ts","../src/mime.ts","../src/converters/plain-text.ts","../src/converters/wikipedia.ts","../src/converters/xlsx.ts","../src/converters/xml-rss-atom.ts","../src/converters/youtube.ts","../src/converters/zip.ts","../src/markitdown.ts"],"sourcesContent":["import type { ByteSource } from \"./types\";\n\nexport async function toUint8Array(source: ByteSource): Promise<Uint8Array> {\n  if (source instanceof Uint8Array) {\n    return source;\n  }\n\n  if (isBlob(source)) {\n    return new Uint8Array(await source.arrayBuffer());\n  }\n\n  if (ArrayBuffer.isView(source)) {\n    return new Uint8Array(source.buffer, source.byteOffset, source.byteLength);\n  }\n\n  return new Uint8Array(source);\n}\n\nexport function toArrayBuffer(source: Uint8Array): ArrayBuffer {\n  return source.slice().buffer;\n}\n\nexport function decodeText(source: Uint8Array): string {\n  return new TextDecoder().decode(source);\n}\n\nexport function isBlob(value: unknown): value is Blob {\n  return typeof Blob !== \"undefined\" && value instanceof Blob;\n}\n\nexport function normalizeExtension(extension: string): string {\n  if (!extension) {\n    return \"\";\n  }\n\n  const normalized = extension.toLowerCase();\n  return normalized.startsWith(\".\") ? normalized : `.${normalized}`;\n}\n","import turndownPluginGfm from \"@joplin/turndown-plugin-gfm\";\nimport TurndownService from \"turndown\";\n\nexport class CustomTurnDown {\n  convertSoup(document: string | Node): string {\n    const service = new TurndownService({ headingStyle: \"atx\" });\n    service.use(turndownPluginGfm.gfm);\n\n    service.addRule(\"anchor tags\", {\n      filter: [\"a\"],\n      replacement(content, node) {\n        if (!content) return \"\";\n\n        const prefix = content.startsWith(\" \") ? \" \" : \"\";\n        const suffix = content.endsWith(\" \") ? \" \" : \"\";\n        const text = content.trim().replace(/\\n\\n.*/g, \"\");\n        if (!text) return \"\";\n\n        const element = node as unknown as Element;\n        const href = element.getAttribute(\"href\");\n        let title = element.getAttribute(\"title\") ?? \"\";\n        if (!href) return `${prefix}${text}${suffix}`;\n\n        try {\n          const url = new URL(href);\n          if (url.protocol !== \"https:\" && url.protocol !== \"http:\") {\n            return `${prefix}${text}${suffix}`;\n          }\n        } catch {\n          if (\n            !href.startsWith(\"/\") &&\n            !href.startsWith(\"#\") &&\n            !href.startsWith(\"./\") &&\n            !href.startsWith(\"../\")\n          ) {\n            return `${prefix}${text}${suffix}`;\n          }\n        }\n\n        if (text.replace(/\\\\_/g, \"_\") === href && !title) {\n          return `<${href}>`;\n        }\n        if (!title) title = href;\n        return `${prefix}[${text}](${href}${title ? ` \"${title}\"` : \"\"})${suffix}`;\n      }\n    });\n\n    service.addRule(\"img tags\", {\n      filter: [\"img\"],\n      replacement(_content, node) {\n        const element = node as unknown as Element;\n        const alt = element.getAttribute(\"alt\") ?? \"\";\n        let source = element.getAttribute(\"src\") ?? \"\";\n        const title = element.getAttribute(\"title\");\n        if (source.startsWith(\"data:\")) {\n          source = `${source.split(\",\", 1)[0]}...`;\n        }\n        return `![${alt}](${source}${title ? ` \"${title}\"` : \"\"})`;\n      }\n    });\n\n    return service.turndown(document as TurndownService.Node);\n  }\n}\n","import { parseHTML } from \"linkedom\";\n\nexport function parseHtmlDocument(html: string): Document {\n  return parseHTML(html).document as unknown as Document;\n}\n","import { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nconst BING_SEARCH_URL = /^https:\\/\\/www\\.bing\\.com\\/search\\?q=/i;\nconst BASE64_ALPHABET = \"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/\";\n\nexport class BingSerpConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    const url = options.url ?? \"\";\n    if ((extension !== \".html\" && extension !== \".htm\") || !BING_SEARCH_URL.test(url)) {\n      return null;\n    }\n\n    const document = parseHtmlDocument(decodeText(source));\n    const query = new URL(url).searchParams.get(\"q\") ?? \"\";\n\n    document.querySelectorAll(\".tptt\").forEach((element) => {\n      if (element.textContent) {\n        element.textContent += \" \";\n      }\n    });\n    document.querySelectorAll(\".algoSlug_icon\").forEach((element) => {\n      element.remove();\n    });\n\n    const turndown = new CustomTurnDown();\n    const results = [...document.querySelectorAll(\".b_algo\")].map((result) => {\n      result.querySelectorAll(\"a[href]\").forEach((anchor) => {\n        const href = anchor.getAttribute(\"href\");\n        if (!href) {\n          return;\n        }\n        try {\n          const encoded = new URL(href, url).searchParams.get(\"u\");\n          if (encoded) {\n            anchor.setAttribute(\"href\", decodeBingUrl(encoded));\n          }\n        } catch {\n          // Preserve malformed links as they appeared in the source document.\n        }\n      });\n\n      return turndown\n        .convertSoup(result)\n        .split(/\\n+/)\n        .map((line) => line.trim())\n        .filter(Boolean)\n        .join(\"\\n\");\n    });\n\n    const markdown = `## A Bing search for '${query}' found the following results:\\n\\n${results.join(\"\\n\\n\")}`;\n    return { title: document.title || null, markdown, text_content: markdown };\n  }\n}\n\nfunction decodeBingUrl(encoded: string): string {\n  const input = encoded.slice(2).replace(/-/g, \"+\").replace(/_/g, \"/\").replace(/=+$/, \"\");\n  const bytes: number[] = [];\n  let bits = 0;\n  let bitCount = 0;\n\n  for (const character of input) {\n    const value = BASE64_ALPHABET.indexOf(character);\n    if (value < 0) {\n      return encoded;\n    }\n    bits = (bits << 6) | value;\n    bitCount += 6;\n    if (bitCount >= 8) {\n      bitCount -= 8;\n      bytes.push((bits >> bitCount) & 0xff);\n    }\n  }\n\n  return new TextDecoder().decode(new Uint8Array(bytes));\n}\n","import { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class HtmlConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if (extension !== \".html\" && extension !== \".htm\") {\n      return null;\n    }\n\n    return this.convertHtml(decodeText(source));\n  }\n\n  async convertHtml(html: string): Promise<ConverterResult> {\n    const document = parseHtmlDocument(html);\n\n    const title = document.title;\n    document.querySelectorAll(\"script, style, title\").forEach((element) => {\n      element.remove();\n    });\n\n    const markdown = new CustomTurnDown().convertSoup(document);\n    return {\n      title: title || null,\n      markdown,\n      text_content: markdown\n    };\n  }\n}\n","import mammoth from \"mammoth\";\nimport { toArrayBuffer } from \"../bytes\";\nimport type { ConverterOptions, ConverterResult } from \"../types\";\nimport { HtmlConverter } from \"./html\";\n\nexport class DocxConverter extends HtmlConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".docx\") {\n      return null;\n    }\n\n    let mammothInput: Parameters<typeof mammoth.convertToHtml>[0] = {\n      arrayBuffer: toArrayBuffer(source)\n    };\n\n    if (\n      typeof process !== \"undefined\" &&\n      process.versions !== null &&\n      process.versions.node !== null\n    ) {\n      mammothInput = { buffer: Buffer.from(source) };\n    }\n\n    const mammothOptions: Parameters<typeof mammoth.convertToHtml>[1] = {\n      styleMap: options.styleMap,\n      includeEmbeddedStyleMap: options.includeEmbeddedStyleMap,\n      includeDefaultStyleMap: options.includeDefaultStyleMap,\n      ignoreEmptyParagraphs: options.ignoreEmptyParagraphs,\n      idPrefix: options.idPrefix\n    };\n\n    if (options.transformDocument) {\n      mammothOptions.transformDocument = options.transformDocument;\n    }\n\n    const html = await mammoth.convertToHtml(mammothInput, mammothOptions);\n\n    return this.convertHtml(html.value);\n  }\n}\n","import { decodeText } from \"../bytes\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\ntype NotebookCell = {\n  cell_type?: string;\n  source?: string[];\n};\n\ntype Notebook = {\n  cells?: NotebookCell[];\n  metadata?: { title?: string };\n};\n\nexport class IpynbConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".ipynb\") {\n      return null;\n    }\n\n    const notebook = JSON.parse(decodeText(source)) as Notebook;\n    const output: string[] = [];\n    let title: string | null = notebook.metadata?.title ?? null;\n\n    for (const cell of notebook.cells ?? []) {\n      const lines = cell.source ?? [];\n      if (cell.cell_type === \"markdown\") {\n        output.push(lines.join(\"\"));\n        title ??=\n          lines\n            .find((line) => line.startsWith(\"# \"))\n            ?.slice(2)\n            .trim() ?? null;\n      } else if (cell.cell_type === \"code\") {\n        output.push(`\\`\\`\\`python\\n${lines.join(\"\")}\\n\\`\\`\\``);\n      } else if (cell.cell_type === \"raw\") {\n        output.push(`\\`\\`\\`\\n${lines.join(\"\")}\\n\\`\\`\\``);\n      }\n    }\n\n    const markdown = output.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n}\n","import { extractText, getDocumentProxy, getMeta } from \"unpdf\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class PdfConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".pdf\") {\n      return null;\n    }\n\n    const pdf = await getDocumentProxy(source.slice());\n    try {\n      const [{ text }, { info }] = await Promise.all([\n        extractText(pdf, { mergePages: true }),\n        getMeta(pdf)\n      ]);\n      const title = typeof info.Title === \"string\" && info.Title.trim() ? info.Title : null;\n\n      return { title, markdown: text, text_content: text };\n    } finally {\n      await pdf.loadingTask.destroy();\n    }\n  }\n}\n","const MIME_EXTENSIONS: Record<string, string> = {\n  \"application/atom+xml\": \".atom\",\n  \"application/json\": \".json\",\n  \"application/pdf\": \".pdf\",\n  \"application/zip\": \".zip\",\n  \"application/rss+xml\": \".rss\",\n  \"application/x-pdf\": \".pdf\",\n  \"application/vnd.openxmlformats-officedocument.spreadsheetml.sheet\": \".xlsx\",\n  \"application/vnd.openxmlformats-officedocument.wordprocessingml.document\": \".docx\",\n  \"application/xml\": \".xml\",\n  \"text/csv\": \".csv\",\n  \"text/html\": \".html\",\n  \"text/markdown\": \".md\",\n  \"text/plain\": \".txt\",\n  \"text/xml\": \".xml\"\n};\n\nconst TEXT_EXTENSIONS = new Set([\".csv\", \".json\", \".log\", \".markdown\", \".md\", \".text\", \".txt\"]);\n\nexport function extensionFromMime(contentType: string | null | undefined): string | null {\n  if (!contentType) {\n    return null;\n  }\n\n  const mime = contentType.split(\";\", 1)[0]?.trim().toLowerCase() ?? \"\";\n  return MIME_EXTENSIONS[mime] ?? (mime.startsWith(\"text/\") ? \".txt\" : null);\n}\n\nexport function isTextExtension(extension: string): boolean {\n  return TEXT_EXTENSIONS.has(extension.toLowerCase());\n}\n\nexport function extensionFromPathname(pathname: string): string | null {\n  const filename = pathname.split(\"/\").pop() ?? \"\";\n  const dot = filename.lastIndexOf(\".\");\n  return dot > 0 ? filename.slice(dot).toLowerCase() : null;\n}\n","import { decodeText } from \"../bytes\";\nimport { isTextExtension } from \"../mime\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class PlainTextConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    if (!isTextExtension(options.file_extension ?? \"\")) {\n      return null;\n    }\n\n    const content = decodeText(source);\n    return { title: null, markdown: content, text_content: content };\n  }\n}\n","import { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nconst WIKIPEDIA_URL = /^https?:\\/\\/[a-z]{2,3}\\.wikipedia\\.org\\//i;\n\nexport class WikipediaConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if ((extension !== \".html\" && extension !== \".htm\") || !WIKIPEDIA_URL.test(options.url ?? \"\")) {\n      return null;\n    }\n\n    const document = parseHtmlDocument(decodeText(source));\n    document.querySelectorAll(\"script, style\").forEach((element) => {\n      element.remove();\n    });\n\n    const body = document.querySelector(\"div#mw-content-text\");\n    const pageTitle = document.querySelector(\"span.mw-page-title-main\")?.textContent?.trim();\n    const title = pageTitle || document.title || null;\n    const markdown = body\n      ? `# ${title ?? \"Wikipedia\"}\\n\\n${new CustomTurnDown().convertSoup(body)}`\n      : new CustomTurnDown().convertSoup(document);\n\n    return { title, markdown, text_content: markdown };\n  }\n}\n","import ExcelJS, { type Cell, type Worksheet } from \"exceljs\";\nimport { toArrayBuffer } from \"../bytes\";\nimport type { ConverterOptions, ConverterResult } from \"../types\";\nimport { HtmlConverter } from \"./html\";\n\nconst { Workbook } = ExcelJS;\n\nexport class XlsxConverter extends HtmlConverter {\n  async convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".xlsx\") {\n      return null;\n    }\n\n    const workbook = new Workbook();\n    await workbook.xlsx.load(toArrayBuffer(source));\n    const sections: string[] = [];\n\n    for (const worksheet of workbook.worksheets) {\n      const html = worksheetToHtml(worksheet);\n      if (!html) {\n        continue;\n      }\n      const result = await this.convertHtml(html);\n      sections.push(`## ${worksheet.name}\\n${result?.markdown.trim() ?? \"\"}`);\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return {\n      title: workbook.title || \"Untitled\",\n      markdown,\n      text_content: markdown\n    };\n  }\n}\n\nfunction worksheetToHtml(worksheet: Worksheet): string {\n  const columnCount = worksheet.columnCount;\n  if (!columnCount || !worksheet.actualRowCount) {\n    return \"\";\n  }\n\n  const rows: string[] = [];\n  worksheet.eachRow({ includeEmpty: true }, (row) => {\n    const cells: string[] = [];\n    for (let column = 1; column <= columnCount; column++) {\n      cells.push(`<td>${escapeHtml(cellText(row.getCell(column)))}</td>`);\n    }\n    rows.push(`<tr>${cells.join(\"\")}</tr>`);\n  });\n\n  return `<html><body><table>${rows.join(\"\")}</table></body></html>`;\n}\n\nfunction cellText(cell: Cell): string {\n  if (cell.isMerged && cell.master !== cell) {\n    return \"\";\n  }\n  return cell.text;\n}\n\nfunction escapeHtml(value: string): string {\n  return value\n    .replace(/&/g, \"&amp;\")\n    .replace(/</g, \"&lt;\")\n    .replace(/>/g, \"&gt;\")\n    .replace(/\"/g, \"&quot;\");\n}\n","import type { Document, Element } from \"@xmldom/xmldom\";\nimport { DOMParser } from \"@xmldom/xmldom\";\nimport { decodeText } from \"../bytes\";\nimport { CustomTurnDown } from \"../custom-turndown\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class RSSConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if (![\".xml\", \".rss\", \".atom\"].includes(extension)) {\n      return null;\n    }\n\n    const document = new DOMParser().parseFromString(decodeText(source), \"text/xml\");\n    if (document.getElementsByTagName(\"rss\").length > 0) {\n      return this.parseRss(document);\n    }\n    if (document.getElementsByTagName(\"feed\")[0]?.getElementsByTagName(\"entry\").length) {\n      return this.parseAtom(document);\n    }\n    return null;\n  }\n\n  private parseAtom(document: Document): ConverterResult {\n    const root = document.getElementsByTagName(\"feed\")[0];\n    if (!root) return null;\n\n    const title = getText(root, \"title\");\n    const subtitle = getText(root, \"subtitle\");\n    const sections = [`# ${title ?? \"Feed\"}`];\n    if (subtitle) sections.push(subtitle);\n\n    for (const entry of [...root.getElementsByTagName(\"entry\")]) {\n      const entryTitle = getText(entry, \"title\");\n      const updated = getText(entry, \"updated\");\n      const summary = getText(entry, \"summary\");\n      const content = getText(entry, \"content\");\n      const lines: string[] = [];\n      if (entryTitle) lines.push(`## ${entryTitle}`);\n      if (updated) lines.push(`Updated on: ${updated}`);\n      if (summary) lines.push(parseContent(summary));\n      if (content) lines.push(parseContent(content));\n      sections.push(lines.join(\"\\n\"));\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n\n  private parseRss(document: Document): ConverterResult {\n    const channel = document.getElementsByTagName(\"rss\")[0]?.getElementsByTagName(\"channel\")[0];\n    if (!channel) return null;\n\n    const title = getText(channel, \"title\");\n    const description = getText(channel, \"description\");\n    const sections: string[] = title ? [`# ${title}`] : [];\n    if (description) sections.push(description);\n\n    for (const item of [...channel.getElementsByTagName(\"item\")]) {\n      const itemTitle = getText(item, \"title\");\n      const published = getText(item, \"pubDate\");\n      const itemDescription = getText(item, \"description\");\n      const content = getText(item, \"content:encoded\");\n      const lines: string[] = [];\n      if (itemTitle) lines.push(`## ${itemTitle}`);\n      if (published) lines.push(`Published on: ${published}`);\n      if (itemDescription) lines.push(parseContent(itemDescription));\n      if (content) lines.push(parseContent(content));\n      sections.push(lines.join(\"\\n\"));\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n}\n\nfunction parseContent(content: string): string {\n  return new CustomTurnDown().convertSoup(parseHtmlDocument(content).documentElement);\n}\n\nfunction getText(element: Element, tagName: string): string | null {\n  const node = element.getElementsByTagName(tagName)[0];\n  return node?.textContent?.trim() || null;\n}\n","import { decodeText } from \"../bytes\";\nimport { parseHtmlDocument } from \"../html-document\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class YouTubeConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    const extension = options.file_extension?.toLowerCase() ?? \"\";\n    if (\n      (extension !== \".html\" && extension !== \".htm\") ||\n      !options.url?.startsWith(\"https://www.youtube.com/watch?\")\n    ) {\n      return null;\n    }\n\n    const document = parseHtmlDocument(decodeText(source));\n    const metadata: Record<string, string> = { title: document.title };\n\n    document.querySelectorAll(\"meta\").forEach((meta) => {\n      const content = meta.getAttribute(\"content\");\n      const key =\n        meta.getAttribute(\"itemprop\") ?? meta.getAttribute(\"property\") ?? meta.getAttribute(\"name\");\n      if (key && content) {\n        metadata[key] = content;\n      }\n    });\n\n    try {\n      for (const script of document.querySelectorAll(\"script\")) {\n        const content = script.textContent ?? \"\";\n        if (!content.includes(\"ytInitialData\")) {\n          continue;\n        }\n        const start = content.indexOf(\"{\");\n        const end = content.lastIndexOf(\"}\");\n        if (start >= 0 && end > start) {\n          const description = findKey(\n            JSON.parse(content.slice(start, end + 1)),\n            \"attributedDescriptionBodyText\"\n          );\n          if (isRecord(description) && typeof description.content === \"string\") {\n            metadata.description = description.content;\n          }\n        }\n        break;\n      }\n    } catch {\n      // Meta description remains available when the page implementation changes.\n    }\n\n    const title = first(metadata, [\"title\", \"og:title\", \"name\"]) || document.title || null;\n    const lines = [\"# YouTube\"];\n    if (title) {\n      lines.push(`## ${title}`);\n    }\n\n    const stats: string[] = [];\n    const views = first(metadata, [\"interactionCount\"]);\n    const keywords = first(metadata, [\"keywords\"]);\n    const runtime = first(metadata, [\"duration\"]);\n    if (views) stats.push(`- **Views:** ${views}`);\n    if (keywords) stats.push(`- **Keywords:** ${keywords}`);\n    if (runtime) stats.push(`- **Runtime:** ${runtime}`);\n    if (stats.length > 0) {\n      lines.push(`### Video Metadata\\n${stats.join(\"\\n\")}`);\n    }\n\n    const description = first(metadata, [\"description\", \"og:description\"]);\n    if (description) {\n      lines.push(`### Description\\n${description}`);\n    }\n\n    const markdown = lines.join(\"\\n\\n\");\n    return { title, markdown, text_content: markdown };\n  }\n}\n\nfunction first(metadata: Record<string, string>, keys: string[]): string | null {\n  for (const key of keys) {\n    if (metadata[key]) return metadata[key];\n  }\n  return null;\n}\n\nfunction findKey(value: unknown, key: string): unknown {\n  if (Array.isArray(value)) {\n    for (const item of value) {\n      const match = findKey(item, key);\n      if (match !== undefined) return match;\n    }\n  } else if (isRecord(value)) {\n    if (key in value) return value[key];\n    for (const item of Object.values(value)) {\n      const match = findKey(item, key);\n      if (match !== undefined) return match;\n    }\n  }\n  return undefined;\n}\n\nfunction isRecord(value: unknown): value is Record<string, unknown> {\n  return typeof value === \"object\" && value !== null;\n}\n","import { unzipSync } from \"fflate\";\nimport { extensionFromPathname } from \"../mime\";\nimport type { ConverterOptions, ConverterResult, DocumentConverter } from \"../types\";\n\nexport class ZipConverter implements DocumentConverter {\n  async convert(source: Uint8Array, options: ConverterOptions = {}): Promise<ConverterResult> {\n    if (options.file_extension?.toLowerCase() !== \".zip\") {\n      return null;\n    }\n\n    const files = unzipSync(source, {\n      filter: (file) => !file.name.endsWith(\"/\")\n    });\n    const converters = options._parent_converters ?? [];\n    const zipName = options.url ?? \"archive.zip\";\n    const sections = [`Content from the zip file \\`${zipName}\\`:`];\n\n    for (const [name, bytes] of Object.entries(files)) {\n      const result = await convertEntry(bytes, name, converters, options);\n      if (result) {\n        sections.push(`## File: ${name}\\n\\n${result.markdown}`);\n      }\n    }\n\n    const markdown = sections.join(\"\\n\\n\");\n    return { title: null, markdown, text_content: markdown };\n  }\n}\n\nasync function convertEntry(\n  source: Uint8Array,\n  name: string,\n  converters: DocumentConverter[],\n  options: ConverterOptions\n): Promise<ConverterResult> {\n  const extension = extensionFromPathname(name);\n  if (!extension) {\n    return null;\n  }\n\n  const entryOptions: ConverterOptions = {\n    ...options,\n    file_extension: extension,\n    url: undefined,\n    _parent_converters: converters\n  };\n\n  for (const converter of converters) {\n    try {\n      const result = await converter.convert(source, entryOptions);\n      if (result) {\n        return result;\n      }\n    } catch {\n      // Match the original ZipConverter: skip entries that fail conversion.\n    }\n  }\n\n  return null;\n}\n","import { isBlob, toUint8Array } from \"./bytes\";\nimport { BingSerpConverter } from \"./converters/bingserp\";\nimport { DocxConverter } from \"./converters/docx\";\nimport { HtmlConverter } from \"./converters/html\";\nimport { IpynbConverter } from \"./converters/ipynb\";\nimport { PdfConverter } from \"./converters/pdf\";\nimport { PlainTextConverter } from \"./converters/plain-text\";\nimport { WikipediaConverter } from \"./converters/wikipedia\";\nimport { XlsxConverter } from \"./converters/xlsx\";\nimport { RSSConverter } from \"./converters/xml-rss-atom\";\nimport { YouTubeConverter } from \"./converters/youtube\";\nimport { ZipConverter } from \"./converters/zip\";\nimport { extensionFromMime, extensionFromPathname } from \"./mime\";\nimport type { ByteSource, ConverterOptions, ConverterResult, DocumentConverter } from \"./types\";\n\nexport type ConversionSource = ByteSource | Response | URL | string;\n\nexport class MarkItDown {\n  private readonly converters: DocumentConverter[] = [];\n\n  constructor() {\n    this.registerConverter(new PlainTextConverter());\n    this.registerConverter(new HtmlConverter());\n    this.registerConverter(new ZipConverter());\n    this.registerConverter(new RSSConverter());\n    this.registerConverter(new WikipediaConverter());\n    this.registerConverter(new YouTubeConverter());\n    this.registerConverter(new BingSerpConverter());\n    this.registerConverter(new DocxConverter());\n    this.registerConverter(new XlsxConverter());\n    this.registerConverter(new PdfConverter());\n    this.registerConverter(new IpynbConverter());\n  }\n\n  /** Convert a URL, response, blob, array buffer, data view, or typed array. */\n  async convert(\n    source: ConversionSource,\n    options: ConverterOptions = {}\n  ): Promise<ConverterResult> {\n    if (typeof source === \"string\" || source instanceof URL) {\n      return this.convertUrl(source.toString(), options);\n    }\n\n    if (isResponse(source)) {\n      return this.convertResponse(source, options);\n    }\n\n    const extensions = new Set<string>();\n    this.addExtension(extensions, options.file_extension?.replace(/^\\./, \"\"));\n\n    if (isBlob(source)) {\n      this.addExtension(extensions, extensionFromMime(source.type));\n      if (\"name\" in source && typeof source.name === \"string\") {\n        this.addExtension(extensions, extensionFromPathname(source.name));\n      }\n    }\n\n    if (extensions.size === 0) {\n      throw new Error(\n        \"Could not determine the file type. Provide `file_extension` when converting bytes.\"\n      );\n    }\n\n    return this.convertBytes(await toUint8Array(source), extensions, options);\n  }\n\n  /**\n   * @deprecated Pass the byte source directly to `convert`.\n   */\n  async convertBuffer(\n    source: ByteSource,\n    options: ConverterOptions & { file_extension: string }\n  ): Promise<ConverterResult> {\n    return this.convert(source, options);\n  }\n\n  private async convertUrl(\n    source: string,\n    { fetch: fetchImplementation = globalThis.fetch, ...options }: ConverterOptions\n  ): Promise<ConverterResult> {\n    const url = new URL(source);\n    if (url.protocol !== \"http:\" && url.protocol !== \"https:\") {\n      throw new Error(`Unsupported URL protocol: ${url.protocol}`);\n    }\n    if (!fetchImplementation) {\n      throw new Error(\"No fetch implementation is available.\");\n    }\n\n    const response = await fetchImplementation(url);\n    if (!response.ok) {\n      throw new Error(`Failed to fetch URL: ${url}, status: ${response.status}`);\n    }\n\n    return this.convertResponse(response, { ...options, url: url.toString() });\n  }\n\n  private async convertResponse(\n    response: Response,\n    options: ConverterOptions\n  ): Promise<ConverterResult> {\n    const extensions = new Set<string>();\n    this.addExtension(extensions, options.file_extension);\n    this.addExtension(extensions, extensionFromMime(response.headers.get(\"content-type\")));\n\n    const disposition = response.headers.get(\"content-disposition\") ?? \"\";\n    const filename = disposition.match(/filename\\*?=(?:UTF-8''|[\"']?)([^;\"']+)/i)?.[1];\n    if (filename) {\n      this.addExtension(extensions, extensionFromPathname(decodeURIComponent(filename.trim())));\n    }\n\n    const responseUrl = options.url || response.url;\n    if (responseUrl) {\n      this.addExtension(extensions, extensionFromPathname(new URL(responseUrl).pathname));\n    }\n\n    if (extensions.size === 0) {\n      throw new Error(\n        \"Could not determine the file type. Provide `file_extension` or a Content-Type header.\"\n      );\n    }\n\n    return this.convertBytes(new Uint8Array(await response.arrayBuffer()), extensions, {\n      ...options,\n      url: responseUrl\n    });\n  }\n\n  private async convertBytes(\n    source: Uint8Array,\n    extensions: Set<string>,\n    options: ConverterOptions\n  ): Promise<ConverterResult> {\n    let lastError: unknown;\n\n    for (const extension of extensions) {\n      for (const converter of this.converters) {\n        try {\n          const result = await converter.convert(source, {\n            ...options,\n            file_extension: extension,\n            _parent_converters: this.converters\n          });\n          if (result) {\n            result.markdown = result.markdown\n              .replace(/\\r\\n|\\r|\\n/g, \"\\n\")\n              .trim()\n              .replace(/\\n{3,}/g, \"\\n\\n\");\n            result.text_content = result.markdown;\n            return result;\n          }\n        } catch (error) {\n          if (options.debug) {\n            console.error(\"Converter threw error:\", error);\n          }\n          lastError = error;\n        }\n      }\n    }\n\n    const formats = [...extensions].join(\", \");\n    if (lastError) {\n      throw new Error(`Could not convert ${formats} to Markdown: ${String(lastError)}`);\n    }\n    throw new Error(`Unsupported file type: ${formats}.`);\n  }\n\n  private addExtension(extensions: Set<string>, extension: string | null | undefined): void {\n    if (!extension) {\n      return;\n    }\n    extensions.add(\n      extension.startsWith(\".\") ? extension.toLowerCase() : `.${extension.toLowerCase()}`\n    );\n  }\n\n  private registerConverter(converter: DocumentConverter): void {\n    this.converters.unshift(converter);\n  }\n}\n\nfunction isResponse(value: unknown): value is Response {\n  return typeof Response !== \"undefined\" && value instanceof Response;\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAEA,eAAsB,aAAa,QAAyC;CAC1E,IAAI,kBAAkB,YACpB,OAAO;CAGT,IAAI,OAAO,MAAM,GACf,OAAO,IAAI,WAAW,MAAM,OAAO,YAAY,CAAC;CAGlD,IAAI,YAAY,OAAO,MAAM,GAC3B,OAAO,IAAI,WAAW,OAAO,QAAQ,OAAO,YAAY,OAAO,UAAU;CAG3E,OAAO,IAAI,WAAW,MAAM;AAC9B;AAEA,SAAgB,cAAc,QAAiC;CAC7D,OAAO,OAAO,MAAM,CAAC,CAAC;AACxB;AAEA,SAAgB,WAAW,QAA4B;CACrD,OAAO,IAAI,YAAY,CAAC,CAAC,OAAO,MAAM;AACxC;AAEA,SAAgB,OAAO,OAA+B;CACpD,OAAO,OAAO,SAAS,eAAe,iBAAiB;AACzD;;;ACzBA,IAAa,iBAAb,MAA4B;CAC1B,YAAY,UAAiC;EAC3C,MAAM,UAAU,IAAIA,SAAAA,QAAgB,EAAE,cAAc,MAAM,CAAC;EAC3D,QAAQ,IAAIC,4BAAAA,QAAkB,GAAG;EAEjC,QAAQ,QAAQ,eAAe;GAC7B,QAAQ,CAAC,GAAG;GACZ,YAAY,SAAS,MAAM;IACzB,IAAI,CAAC,SAAS,OAAO;IAErB,MAAM,SAAS,QAAQ,WAAW,GAAG,IAAI,MAAM;IAC/C,MAAM,SAAS,QAAQ,SAAS,GAAG,IAAI,MAAM;IAC7C,MAAM,OAAO,QAAQ,KAAK,CAAC,CAAC,QAAQ,WAAW,EAAE;IACjD,IAAI,CAAC,MAAM,OAAO;IAElB,MAAM,UAAU;IAChB,MAAM,OAAO,QAAQ,aAAa,MAAM;IACxC,IAAI,QAAQ,QAAQ,aAAa,OAAO,KAAK;IAC7C,IAAI,CAAC,MAAM,OAAO,GAAG,SAAS,OAAO;IAErC,IAAI;KACF,MAAM,MAAM,IAAI,IAAI,IAAI;KACxB,IAAI,IAAI,aAAa,YAAY,IAAI,aAAa,SAChD,OAAO,GAAG,SAAS,OAAO;IAE9B,QAAQ;KACN,IACE,CAAC,KAAK,WAAW,GAAG,KACpB,CAAC,KAAK,WAAW,GAAG,KACpB,CAAC,KAAK,WAAW,IAAI,KACrB,CAAC,KAAK,WAAW,KAAK,GAEtB,OAAO,GAAG,SAAS,OAAO;IAE9B;IAEA,IAAI,KAAK,QAAQ,QAAQ,GAAG,MAAM,QAAQ,CAAC,OACzC,OAAO,IAAI,KAAK;IAElB,IAAI,CAAC,OAAO,QAAQ;IACpB,OAAO,GAAG,OAAO,GAAG,KAAK,IAAI,OAAO,QAAQ,KAAK,MAAM,KAAK,GAAG,GAAG;GACpE;EACF,CAAC;EAED,QAAQ,QAAQ,YAAY;GAC1B,QAAQ,CAAC,KAAK;GACd,YAAY,UAAU,MAAM;IAC1B,MAAM,UAAU;IAChB,MAAM,MAAM,QAAQ,aAAa,KAAK,KAAK;IAC3C,IAAI,SAAS,QAAQ,aAAa,KAAK,KAAK;IAC5C,MAAM,QAAQ,QAAQ,aAAa,OAAO;IAC1C,IAAI,OAAO,WAAW,OAAO,GAC3B,SAAS,GAAG,OAAO,MAAM,KAAK,CAAC,CAAC,CAAC,GAAG;IAEtC,OAAO,KAAK,IAAI,IAAI,SAAS,QAAQ,KAAK,MAAM,KAAK,GAAG;GAC1D;EACF,CAAC;EAED,OAAO,QAAQ,SAAS,QAAgC;CAC1D;AACF;;;AC7DA,SAAgB,kBAAkB,MAAwB;CACxD,QAAA,GAAOC,SAAAA,UAAAA,CAAU,IAAI,CAAC,CAAC;AACzB;;;ACCA,MAAM,kBAAkB;AACxB,MAAM,kBAAkB;AAExB,IAAa,oBAAb,MAA4D;CAC1D,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,MAAM,MAAM,QAAQ,OAAO;EAC3B,IAAK,cAAc,WAAW,cAAc,UAAW,CAAC,gBAAgB,KAAK,GAAG,GAC9E,OAAO;EAGT,MAAM,WAAW,kBAAkB,WAAW,MAAM,CAAC;EACrD,MAAM,QAAQ,IAAI,IAAI,GAAG,CAAC,CAAC,aAAa,IAAI,GAAG,KAAK;EAEpD,SAAS,iBAAiB,OAAO,CAAC,CAAC,SAAS,YAAY;GACtD,IAAI,QAAQ,aACV,QAAQ,eAAe;EAE3B,CAAC;EACD,SAAS,iBAAiB,gBAAgB,CAAC,CAAC,SAAS,YAAY;GAC/D,QAAQ,OAAO;EACjB,CAAC;EAED,MAAM,WAAW,IAAI,eAAe;EAyBpC,MAAM,WAAW,yBAAyB,MAAM,oCAxBhC,CAAC,GAAG,SAAS,iBAAiB,SAAS,CAAC,CAAC,CAAC,KAAK,WAAW;GACxE,OAAO,iBAAiB,SAAS,CAAC,CAAC,SAAS,WAAW;IACrD,MAAM,OAAO,OAAO,aAAa,MAAM;IACvC,IAAI,CAAC,MACH;IAEF,IAAI;KACF,MAAM,UAAU,IAAI,IAAI,MAAM,GAAG,CAAC,CAAC,aAAa,IAAI,GAAG;KACvD,IAAI,SACF,OAAO,aAAa,QAAQ,cAAc,OAAO,CAAC;IAEtD,QAAQ,CAER;GACF,CAAC;GAED,OAAO,SACJ,YAAY,MAAM,CAAC,CACnB,MAAM,KAAK,CAAC,CACZ,KAAK,SAAS,KAAK,KAAK,CAAC,CAAC,CAC1B,OAAO,OAAO,CAAC,CACf,KAAK,IAAI;EACd,CAE0F,CAAC,CAAC,KAAK,MAAM;EACvG,OAAO;GAAE,OAAO,SAAS,SAAS;GAAM;GAAU,cAAc;EAAS;CAC3E;AACF;AAEA,SAAS,cAAc,SAAyB;CAC9C,MAAM,QAAQ,QAAQ,MAAM,CAAC,CAAC,CAAC,QAAQ,MAAM,GAAG,CAAC,CAAC,QAAQ,MAAM,GAAG,CAAC,CAAC,QAAQ,OAAO,EAAE;CACtF,MAAM,QAAkB,CAAC;CACzB,IAAI,OAAO;CACX,IAAI,WAAW;CAEf,KAAK,MAAM,aAAa,OAAO;EAC7B,MAAM,QAAQ,gBAAgB,QAAQ,SAAS;EAC/C,IAAI,QAAQ,GACV,OAAO;EAET,OAAQ,QAAQ,IAAK;EACrB,YAAY;EACZ,IAAI,YAAY,GAAG;GACjB,YAAY;GACZ,MAAM,KAAM,QAAQ,WAAY,GAAI;EACtC;CACF;CAEA,OAAO,IAAI,YAAY,CAAC,CAAC,OAAO,IAAI,WAAW,KAAK,CAAC;AACvD;;;ACzEA,IAAa,gBAAb,MAAwD;CACtD,MAAM,QAAQ,QAAoB,SAAqD;EACrF,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IAAI,cAAc,WAAW,cAAc,QACzC,OAAO;EAGT,OAAO,KAAK,YAAY,WAAW,MAAM,CAAC;CAC5C;CAEA,MAAM,YAAY,MAAwC;EACxD,MAAM,WAAW,kBAAkB,IAAI;EAEvC,MAAM,QAAQ,SAAS;EACvB,SAAS,iBAAiB,sBAAsB,CAAC,CAAC,SAAS,YAAY;GACrE,QAAQ,OAAO;EACjB,CAAC;EAED,MAAM,WAAW,IAAI,eAAe,CAAC,CAAC,YAAY,QAAQ;EAC1D,OAAO;GACL,OAAO,SAAS;GAChB;GACA,cAAc;EAChB;CACF;AACF;;;ACzBA,IAAa,gBAAb,cAAmC,cAAc;CAC/C,MAAM,QAAQ,QAAoB,SAAqD;EACrF,IAAI,QAAQ,gBAAgB,YAAY,MAAM,SAC5C,OAAO;EAGT,IAAI,eAA4D,EAC9D,aAAa,cAAc,MAAM,EACnC;EAEA,IACE,OAAO,YAAY,eACnB,QAAQ,aAAa,QACrB,QAAQ,SAAS,SAAS,MAE1B,eAAe,EAAE,QAAQ,OAAO,KAAK,MAAM,EAAE;EAG/C,MAAM,iBAA8D;GAClE,UAAU,QAAQ;GAClB,yBAAyB,QAAQ;GACjC,wBAAwB,QAAQ;GAChC,uBAAuB,QAAQ;GAC/B,UAAU,QAAQ;EACpB;EAEA,IAAI,QAAQ,mBACV,eAAe,oBAAoB,QAAQ;EAG7C,MAAM,OAAO,MAAM,QAAA,QAAQ,cAAc,cAAc,cAAc;EAErE,OAAO,KAAK,YAAY,KAAK,KAAK;CACpC;AACF;;;AC1BA,IAAa,iBAAb,MAAyD;CACvD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,IAAI,QAAQ,gBAAgB,YAAY,MAAM,UAC5C,OAAO;EAGT,MAAM,WAAW,KAAK,MAAM,WAAW,MAAM,CAAC;EAC9C,MAAM,SAAmB,CAAC;EAC1B,IAAI,QAAuB,SAAS,UAAU,SAAS;EAEvD,KAAK,MAAM,QAAQ,SAAS,SAAS,CAAC,GAAG;GACvC,MAAM,QAAQ,KAAK,UAAU,CAAC;GAC9B,IAAI,KAAK,cAAc,YAAY;IACjC,OAAO,KAAK,MAAM,KAAK,EAAE,CAAC;IAC1B,UAAA,QACE,MACG,MAAM,SAAS,KAAK,WAAW,IAAI,CAAC,CAAC,EACpC,MAAM,CAAC,CAAC,CACT,KAAK,KAAK;GACjB,OAAO,IAAI,KAAK,cAAc,QAC5B,OAAO,KAAK,iBAAiB,MAAM,KAAK,EAAE,EAAE,SAAS;QAChD,IAAI,KAAK,cAAc,OAC5B,OAAO,KAAK,WAAW,MAAM,KAAK,EAAE,EAAE,SAAS;EAEnD;EAEA,MAAM,WAAW,OAAO,KAAK,MAAM;EACnC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;;;ACvCA,IAAa,eAAb,MAAuD;CACrD,MAAM,QAAQ,QAAoB,SAAqD;EACrF,IAAI,QAAQ,gBAAgB,YAAY,MAAM,QAC5C,OAAO;EAGT,MAAM,MAAM,OAAA,GAAMC,MAAAA,iBAAAA,CAAiB,OAAO,MAAM,CAAC;EACjD,IAAI;GACF,MAAM,CAAC,EAAE,QAAQ,EAAE,UAAU,MAAM,QAAQ,IAAI,EAAA,GAC7CC,MAAAA,YAAAA,CAAY,KAAK,EAAE,YAAY,KAAK,CAAC,IAAA,GACrCC,MAAAA,QAAAA,CAAQ,GAAG,CACb,CAAC;GAGD,OAAO;IAAE,OAFK,OAAO,KAAK,UAAU,YAAY,KAAK,MAAM,KAAK,IAAI,KAAK,QAAQ;IAEjE,UAAU;IAAM,cAAc;GAAK;EACrD,UAAU;GACR,MAAM,IAAI,YAAY,QAAQ;EAChC;CACF;AACF;;;ACtBA,MAAM,kBAA0C;CAC9C,wBAAwB;CACxB,oBAAoB;CACpB,mBAAmB;CACnB,mBAAmB;CACnB,uBAAuB;CACvB,qBAAqB;CACrB,qEAAqE;CACrE,2EAA2E;CAC3E,mBAAmB;CACnB,YAAY;CACZ,aAAa;CACb,iBAAiB;CACjB,cAAc;CACd,YAAY;AACd;AAEA,MAAM,kCAAkB,IAAI,IAAI;CAAC;CAAQ;CAAS;CAAQ;CAAa;CAAO;CAAS;AAAM,CAAC;AAE9F,SAAgB,kBAAkB,aAAuD;CACvF,IAAI,CAAC,aACH,OAAO;CAGT,MAAM,OAAO,YAAY,MAAM,KAAK,CAAC,CAAC,CAAC,EAAE,EAAE,KAAK,CAAC,CAAC,YAAY,KAAK;CACnE,OAAO,gBAAgB,UAAU,KAAK,WAAW,OAAO,IAAI,SAAS;AACvE;AAEA,SAAgB,gBAAgB,WAA4B;CAC1D,OAAO,gBAAgB,IAAI,UAAU,YAAY,CAAC;AACpD;AAEA,SAAgB,sBAAsB,UAAiC;CACrE,MAAM,WAAW,SAAS,MAAM,GAAG,CAAC,CAAC,IAAI,KAAK;CAC9C,MAAM,MAAM,SAAS,YAAY,GAAG;CACpC,OAAO,MAAM,IAAI,SAAS,MAAM,GAAG,CAAC,CAAC,YAAY,IAAI;AACvD;;;AChCA,IAAa,qBAAb,MAA6D;CAC3D,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,IAAI,CAAC,gBAAgB,QAAQ,kBAAkB,EAAE,GAC/C,OAAO;EAGT,MAAM,UAAU,WAAW,MAAM;EACjC,OAAO;GAAE,OAAO;GAAM,UAAU;GAAS,cAAc;EAAQ;CACjE;AACF;;;ACRA,MAAM,gBAAgB;AAEtB,IAAa,qBAAb,MAA6D;CAC3D,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IAAK,cAAc,WAAW,cAAc,UAAW,CAAC,cAAc,KAAK,QAAQ,OAAO,EAAE,GAC1F,OAAO;EAGT,MAAM,WAAW,kBAAkB,WAAW,MAAM,CAAC;EACrD,SAAS,iBAAiB,eAAe,CAAC,CAAC,SAAS,YAAY;GAC9D,QAAQ,OAAO;EACjB,CAAC;EAED,MAAM,OAAO,SAAS,cAAc,qBAAqB;EAEzD,MAAM,QADY,SAAS,cAAc,yBAAyB,CAAC,EAAE,aAAa,KAAK,KAC5D,SAAS,SAAS;EAC7C,MAAM,WAAW,OACb,KAAK,SAAS,YAAY,MAAM,IAAI,eAAe,CAAC,CAAC,YAAY,IAAI,MACrE,IAAI,eAAe,CAAC,CAAC,YAAY,QAAQ;EAE7C,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;;;ACvBA,MAAM,EAAE,aAAaC,QAAAA;AAErB,IAAa,gBAAb,cAAmC,cAAc;CAC/C,MAAM,QAAQ,QAAoB,SAAqD;EACrF,IAAI,QAAQ,gBAAgB,YAAY,MAAM,SAC5C,OAAO;EAGT,MAAM,WAAW,IAAI,SAAS;EAC9B,MAAM,SAAS,KAAK,KAAK,cAAc,MAAM,CAAC;EAC9C,MAAM,WAAqB,CAAC;EAE5B,KAAK,MAAM,aAAa,SAAS,YAAY;GAC3C,MAAM,OAAO,gBAAgB,SAAS;GACtC,IAAI,CAAC,MACH;GAEF,MAAM,SAAS,MAAM,KAAK,YAAY,IAAI;GAC1C,SAAS,KAAK,MAAM,UAAU,KAAK,IAAI,QAAQ,SAAS,KAAK,KAAK,IAAI;EACxE;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GACL,OAAO,SAAS,SAAS;GACzB;GACA,cAAc;EAChB;CACF;AACF;AAEA,SAAS,gBAAgB,WAA8B;CACrD,MAAM,cAAc,UAAU;CAC9B,IAAI,CAAC,eAAe,CAAC,UAAU,gBAC7B,OAAO;CAGT,MAAM,OAAiB,CAAC;CACxB,UAAU,QAAQ,EAAE,cAAc,KAAK,IAAI,QAAQ;EACjD,MAAM,QAAkB,CAAC;EACzB,KAAK,IAAI,SAAS,GAAG,UAAU,aAAa,UAC1C,MAAM,KAAK,OAAO,WAAW,SAAS,IAAI,QAAQ,MAAM,CAAC,CAAC,EAAE,MAAM;EAEpE,KAAK,KAAK,OAAO,MAAM,KAAK,EAAE,EAAE,MAAM;CACxC,CAAC;CAED,OAAO,sBAAsB,KAAK,KAAK,EAAE,EAAE;AAC7C;AAEA,SAAS,SAAS,MAAoB;CACpC,IAAI,KAAK,YAAY,KAAK,WAAW,MACnC,OAAO;CAET,OAAO,KAAK;AACd;AAEA,SAAS,WAAW,OAAuB;CACzC,OAAO,MACJ,QAAQ,MAAM,OAAO,CAAC,CACtB,QAAQ,MAAM,MAAM,CAAC,CACrB,QAAQ,MAAM,MAAM,CAAC,CACrB,QAAQ,MAAM,QAAQ;AAC3B;;;AC3DA,IAAa,eAAb,MAAuD;CACrD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IAAI,CAAC;GAAC;GAAQ;GAAQ;EAAO,CAAC,CAAC,SAAS,SAAS,GAC/C,OAAO;EAGT,MAAM,WAAW,IAAIC,eAAAA,UAAU,CAAC,CAAC,gBAAgB,WAAW,MAAM,GAAG,UAAU;EAC/E,IAAI,SAAS,qBAAqB,KAAK,CAAC,CAAC,SAAS,GAChD,OAAO,KAAK,SAAS,QAAQ;EAE/B,IAAI,SAAS,qBAAqB,MAAM,CAAC,CAAC,EAAE,EAAE,qBAAqB,OAAO,CAAC,CAAC,QAC1E,OAAO,KAAK,UAAU,QAAQ;EAEhC,OAAO;CACT;CAEA,UAAkB,UAAqC;EACrD,MAAM,OAAO,SAAS,qBAAqB,MAAM,CAAC,CAAC;EACnD,IAAI,CAAC,MAAM,OAAO;EAElB,MAAM,QAAQ,QAAQ,MAAM,OAAO;EACnC,MAAM,WAAW,QAAQ,MAAM,UAAU;EACzC,MAAM,WAAW,CAAC,KAAK,SAAS,QAAQ;EACxC,IAAI,UAAU,SAAS,KAAK,QAAQ;EAEpC,KAAK,MAAM,SAAS,CAAC,GAAG,KAAK,qBAAqB,OAAO,CAAC,GAAG;GAC3D,MAAM,aAAa,QAAQ,OAAO,OAAO;GACzC,MAAM,UAAU,QAAQ,OAAO,SAAS;GACxC,MAAM,UAAU,QAAQ,OAAO,SAAS;GACxC,MAAM,UAAU,QAAQ,OAAO,SAAS;GACxC,MAAM,QAAkB,CAAC;GACzB,IAAI,YAAY,MAAM,KAAK,MAAM,YAAY;GAC7C,IAAI,SAAS,MAAM,KAAK,eAAe,SAAS;GAChD,IAAI,SAAS,MAAM,KAAK,aAAa,OAAO,CAAC;GAC7C,IAAI,SAAS,MAAM,KAAK,aAAa,OAAO,CAAC;GAC7C,SAAS,KAAK,MAAM,KAAK,IAAI,CAAC;EAChC;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;CAEA,SAAiB,UAAqC;EACpD,MAAM,UAAU,SAAS,qBAAqB,KAAK,CAAC,CAAC,EAAE,EAAE,qBAAqB,SAAS,CAAC,CAAC;EACzF,IAAI,CAAC,SAAS,OAAO;EAErB,MAAM,QAAQ,QAAQ,SAAS,OAAO;EACtC,MAAM,cAAc,QAAQ,SAAS,aAAa;EAClD,MAAM,WAAqB,QAAQ,CAAC,KAAK,OAAO,IAAI,CAAC;EACrD,IAAI,aAAa,SAAS,KAAK,WAAW;EAE1C,KAAK,MAAM,QAAQ,CAAC,GAAG,QAAQ,qBAAqB,MAAM,CAAC,GAAG;GAC5D,MAAM,YAAY,QAAQ,MAAM,OAAO;GACvC,MAAM,YAAY,QAAQ,MAAM,SAAS;GACzC,MAAM,kBAAkB,QAAQ,MAAM,aAAa;GACnD,MAAM,UAAU,QAAQ,MAAM,iBAAiB;GAC/C,MAAM,QAAkB,CAAC;GACzB,IAAI,WAAW,MAAM,KAAK,MAAM,WAAW;GAC3C,IAAI,WAAW,MAAM,KAAK,iBAAiB,WAAW;GACtD,IAAI,iBAAiB,MAAM,KAAK,aAAa,eAAe,CAAC;GAC7D,IAAI,SAAS,MAAM,KAAK,aAAa,OAAO,CAAC;GAC7C,SAAS,KAAK,MAAM,KAAK,IAAI,CAAC;EAChC;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;AAEA,SAAS,aAAa,SAAyB;CAC7C,OAAO,IAAI,eAAe,CAAC,CAAC,YAAY,kBAAkB,OAAO,CAAC,CAAC,eAAe;AACpF;AAEA,SAAS,QAAQ,SAAkB,SAAgC;CAEjE,OADa,QAAQ,qBAAqB,OAAO,CAAC,CAAC,EACxC,EAAE,aAAa,KAAK,KAAK;AACtC;;;AChFA,IAAa,mBAAb,MAA2D;CACzD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,MAAM,YAAY,QAAQ,gBAAgB,YAAY,KAAK;EAC3D,IACG,cAAc,WAAW,cAAc,UACxC,CAAC,QAAQ,KAAK,WAAW,gCAAgC,GAEzD,OAAO;EAGT,MAAM,WAAW,kBAAkB,WAAW,MAAM,CAAC;EACrD,MAAM,WAAmC,EAAE,OAAO,SAAS,MAAM;EAEjE,SAAS,iBAAiB,MAAM,CAAC,CAAC,SAAS,SAAS;GAClD,MAAM,UAAU,KAAK,aAAa,SAAS;GAC3C,MAAM,MACJ,KAAK,aAAa,UAAU,KAAK,KAAK,aAAa,UAAU,KAAK,KAAK,aAAa,MAAM;GAC5F,IAAI,OAAO,SACT,SAAS,OAAO;EAEpB,CAAC;EAED,IAAI;GACF,KAAK,MAAM,UAAU,SAAS,iBAAiB,QAAQ,GAAG;IACxD,MAAM,UAAU,OAAO,eAAe;IACtC,IAAI,CAAC,QAAQ,SAAS,eAAe,GACnC;IAEF,MAAM,QAAQ,QAAQ,QAAQ,GAAG;IACjC,MAAM,MAAM,QAAQ,YAAY,GAAG;IACnC,IAAI,SAAS,KAAK,MAAM,OAAO;KAC7B,MAAM,cAAc,QAClB,KAAK,MAAM,QAAQ,MAAM,OAAO,MAAM,CAAC,CAAC,GACxC,+BACF;KACA,IAAI,SAAS,WAAW,KAAK,OAAO,YAAY,YAAY,UAC1D,SAAS,cAAc,YAAY;IAEvC;IACA;GACF;EACF,QAAQ,CAER;EAEA,MAAM,QAAQ,MAAM,UAAU;GAAC;GAAS;GAAY;EAAM,CAAC,KAAK,SAAS,SAAS;EAClF,MAAM,QAAQ,CAAC,WAAW;EAC1B,IAAI,OACF,MAAM,KAAK,MAAM,OAAO;EAG1B,MAAM,QAAkB,CAAC;EACzB,MAAM,QAAQ,MAAM,UAAU,CAAC,kBAAkB,CAAC;EAClD,MAAM,WAAW,MAAM,UAAU,CAAC,UAAU,CAAC;EAC7C,MAAM,UAAU,MAAM,UAAU,CAAC,UAAU,CAAC;EAC5C,IAAI,OAAO,MAAM,KAAK,gBAAgB,OAAO;EAC7C,IAAI,UAAU,MAAM,KAAK,mBAAmB,UAAU;EACtD,IAAI,SAAS,MAAM,KAAK,kBAAkB,SAAS;EACnD,IAAI,MAAM,SAAS,GACjB,MAAM,KAAK,uBAAuB,MAAM,KAAK,IAAI,GAAG;EAGtD,MAAM,cAAc,MAAM,UAAU,CAAC,eAAe,gBAAgB,CAAC;EACrE,IAAI,aACF,MAAM,KAAK,oBAAoB,aAAa;EAG9C,MAAM,WAAW,MAAM,KAAK,MAAM;EAClC,OAAO;GAAE;GAAO;GAAU,cAAc;EAAS;CACnD;AACF;AAEA,SAAS,MAAM,UAAkC,MAA+B;CAC9E,KAAK,MAAM,OAAO,MAChB,IAAI,SAAS,MAAM,OAAO,SAAS;CAErC,OAAO;AACT;AAEA,SAAS,QAAQ,OAAgB,KAAsB;CACrD,IAAI,MAAM,QAAQ,KAAK,GACrB,KAAK,MAAM,QAAQ,OAAO;EACxB,MAAM,QAAQ,QAAQ,MAAM,GAAG;EAC/B,IAAI,UAAU,KAAA,GAAW,OAAO;CAClC;MACK,IAAI,SAAS,KAAK,GAAG;EAC1B,IAAI,OAAO,OAAO,OAAO,MAAM;EAC/B,KAAK,MAAM,QAAQ,OAAO,OAAO,KAAK,GAAG;GACvC,MAAM,QAAQ,QAAQ,MAAM,GAAG;GAC/B,IAAI,UAAU,KAAA,GAAW,OAAO;EAClC;CACF;AAEF;AAEA,SAAS,SAAS,OAAkD;CAClE,OAAO,OAAO,UAAU,YAAY,UAAU;AAChD;;;ACjGA,IAAa,eAAb,MAAuD;CACrD,MAAM,QAAQ,QAAoB,UAA4B,CAAC,GAA6B;EAC1F,IAAI,QAAQ,gBAAgB,YAAY,MAAM,QAC5C,OAAO;EAGT,MAAM,SAAA,GAAQC,OAAAA,UAAAA,CAAU,QAAQ,EAC9B,SAAS,SAAS,CAAC,KAAK,KAAK,SAAS,GAAG,EAC3C,CAAC;EACD,MAAM,aAAa,QAAQ,sBAAsB,CAAC;EAElD,MAAM,WAAW,CAAC,+BADF,QAAQ,OAAO,cAC0B,IAAI;EAE7D,KAAK,MAAM,CAAC,MAAM,UAAU,OAAO,QAAQ,KAAK,GAAG;GACjD,MAAM,SAAS,MAAM,aAAa,OAAO,MAAM,YAAY,OAAO;GAClE,IAAI,QACF,SAAS,KAAK,YAAY,KAAK,MAAM,OAAO,UAAU;EAE1D;EAEA,MAAM,WAAW,SAAS,KAAK,MAAM;EACrC,OAAO;GAAE,OAAO;GAAM;GAAU,cAAc;EAAS;CACzD;AACF;AAEA,eAAe,aACb,QACA,MACA,YACA,SAC0B;CAC1B,MAAM,YAAY,sBAAsB,IAAI;CAC5C,IAAI,CAAC,WACH,OAAO;CAGT,MAAM,eAAiC;EACrC,GAAG;EACH,gBAAgB;EAChB,KAAK,KAAA;EACL,oBAAoB;CACtB;CAEA,KAAK,MAAM,aAAa,YACtB,IAAI;EACF,MAAM,SAAS,MAAM,UAAU,QAAQ,QAAQ,YAAY;EAC3D,IAAI,QACF,OAAO;CAEX,QAAQ,CAER;CAGF,OAAO;AACT;;;AC1CA,IAAa,aAAb,MAAwB;CAGtB,cAAc;EAFd,KAAiB,aAAkC,CAAC;EAGlD,KAAK,kBAAkB,IAAI,mBAAmB,CAAC;EAC/C,KAAK,kBAAkB,IAAI,cAAc,CAAC;EAC1C,KAAK,kBAAkB,IAAI,aAAa,CAAC;EACzC,KAAK,kBAAkB,IAAI,aAAa,CAAC;EACzC,KAAK,kBAAkB,IAAI,mBAAmB,CAAC;EAC/C,KAAK,kBAAkB,IAAI,iBAAiB,CAAC;EAC7C,KAAK,kBAAkB,IAAI,kBAAkB,CAAC;EAC9C,KAAK,kBAAkB,IAAI,cAAc,CAAC;EAC1C,KAAK,kBAAkB,IAAI,cAAc,CAAC;EAC1C,KAAK,kBAAkB,IAAI,aAAa,CAAC;EACzC,KAAK,kBAAkB,IAAI,eAAe,CAAC;CAC7C;;CAGA,MAAM,QACJ,QACA,UAA4B,CAAC,GACH;EAC1B,IAAI,OAAO,WAAW,YAAY,kBAAkB,KAClD,OAAO,KAAK,WAAW,OAAO,SAAS,GAAG,OAAO;EAGnD,IAAI,WAAW,MAAM,GACnB,OAAO,KAAK,gBAAgB,QAAQ,OAAO;EAG7C,MAAM,6BAAa,IAAI,IAAY;EACnC,KAAK,aAAa,YAAY,QAAQ,gBAAgB,QAAQ,OAAO,EAAE,CAAC;EAExE,IAAI,OAAO,MAAM,GAAG;GAClB,KAAK,aAAa,YAAY,kBAAkB,OAAO,IAAI,CAAC;GAC5D,IAAI,UAAU,UAAU,OAAO,OAAO,SAAS,UAC7C,KAAK,aAAa,YAAY,sBAAsB,OAAO,IAAI,CAAC;EAEpE;EAEA,IAAI,WAAW,SAAS,GACtB,MAAM,IAAI,MACR,oFACF;EAGF,OAAO,KAAK,aAAa,MAAM,aAAa,MAAM,GAAG,YAAY,OAAO;CAC1E;;;;CAKA,MAAM,cACJ,QACA,SAC0B;EAC1B,OAAO,KAAK,QAAQ,QAAQ,OAAO;CACrC;CAEA,MAAc,WACZ,QACA,EAAE,OAAO,sBAAsB,WAAW,OAAO,GAAG,WAC1B;EAC1B,MAAM,MAAM,IAAI,IAAI,MAAM;EAC1B,IAAI,IAAI,aAAa,WAAW,IAAI,aAAa,UAC/C,MAAM,IAAI,MAAM,6BAA6B,IAAI,UAAU;EAE7D,IAAI,CAAC,qBACH,MAAM,IAAI,MAAM,uCAAuC;EAGzD,MAAM,WAAW,MAAM,oBAAoB,GAAG;EAC9C,IAAI,CAAC,SAAS,IACZ,MAAM,IAAI,MAAM,wBAAwB,IAAI,YAAY,SAAS,QAAQ;EAG3E,OAAO,KAAK,gBAAgB,UAAU;GAAE,GAAG;GAAS,KAAK,IAAI,SAAS;EAAE,CAAC;CAC3E;CAEA,MAAc,gBACZ,UACA,SAC0B;EAC1B,MAAM,6BAAa,IAAI,IAAY;EACnC,KAAK,aAAa,YAAY,QAAQ,cAAc;EACpD,KAAK,aAAa,YAAY,kBAAkB,SAAS,QAAQ,IAAI,cAAc,CAAC,CAAC;EAGrF,MAAM,YADc,SAAS,QAAQ,IAAI,qBAAqB,KAAK,GAAA,CACtC,MAAM,yCAAyC,CAAC,GAAG;EAChF,IAAI,UACF,KAAK,aAAa,YAAY,sBAAsB,mBAAmB,SAAS,KAAK,CAAC,CAAC,CAAC;EAG1F,MAAM,cAAc,QAAQ,OAAO,SAAS;EAC5C,IAAI,aACF,KAAK,aAAa,YAAY,sBAAsB,IAAI,IAAI,WAAW,CAAC,CAAC,QAAQ,CAAC;EAGpF,IAAI,WAAW,SAAS,GACtB,MAAM,IAAI,MACR,uFACF;EAGF,OAAO,KAAK,aAAa,IAAI,WAAW,MAAM,SAAS,YAAY,CAAC,GAAG,YAAY;GACjF,GAAG;GACH,KAAK;EACP,CAAC;CACH;CAEA,MAAc,aACZ,QACA,YACA,SAC0B;EAC1B,IAAI;EAEJ,KAAK,MAAM,aAAa,YACtB,KAAK,MAAM,aAAa,KAAK,YAC3B,IAAI;GACF,MAAM,SAAS,MAAM,UAAU,QAAQ,QAAQ;IAC7C,GAAG;IACH,gBAAgB;IAChB,oBAAoB,KAAK;GAC3B,CAAC;GACD,IAAI,QAAQ;IACV,OAAO,WAAW,OAAO,SACtB,QAAQ,eAAe,IAAI,CAAC,CAC5B,KAAK,CAAC,CACN,QAAQ,WAAW,MAAM;IAC5B,OAAO,eAAe,OAAO;IAC7B,OAAO;GACT;EACF,SAAS,OAAO;GACd,IAAI,QAAQ,OACV,QAAQ,MAAM,0BAA0B,KAAK;GAE/C,YAAY;EACd;EAIJ,MAAM,UAAU,CAAC,GAAG,UAAU,CAAC,CAAC,KAAK,IAAI;EACzC,IAAI,WACF,MAAM,IAAI,MAAM,qBAAqB,QAAQ,gBAAgB,OAAO,SAAS,GAAG;EAElF,MAAM,IAAI,MAAM,0BAA0B,QAAQ,EAAE;CACtD;CAEA,aAAqB,YAAyB,WAA4C;EACxF,IAAI,CAAC,WACH;EAEF,WAAW,IACT,UAAU,WAAW,GAAG,IAAI,UAAU,YAAY,IAAI,IAAI,UAAU,YAAY,GAClF;CACF;CAEA,kBAA0B,WAAoC;EAC5D,KAAK,WAAW,QAAQ,SAAS;CACnC;AACF;AAEA,SAAS,WAAW,OAAmC;CACrD,OAAO,OAAO,aAAa,eAAe,iBAAiB;AAC7D"}

package/dist/index.d.cts

@@ -27,6 +27,7 @@ type ConverterOptions = {
   idPrefix?: string;
   transformDocument?: (element: unknown) => unknown;
   _parent_converters?: DocumentConverter[];
+  debug?: boolean;
 };
 interface DocumentConverter {
   convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult>;

package/dist/index.d.cts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.cts","names":[],"sources":["../src/types.ts","../src/markitdown.ts"],"mappings":";KAAY;EACV;EACA;;EAEA;;;;;;;;KASU,aAAa,aAAa,kBAAkB,kBAAkB;KAE9D;;EAEV;;EAEA;;EAEA,eAAe;;EAEf;EACA;EACA;EACA;EACA;EACA,qBAAqB;EACrB,qBAAqB;;UAGN;EACf,QAAQ,QAAQ,YAAY,SAAS,mBAAmB,QAAQ;;;;KClBtD,mBAAmB,aAAa,WAAW;cAE1C;mBACM;EAEjB;;EAeM,QACJ,QAAQ,kBACR,UAAS,mBACR,QAAQ;;;;EA+BL,cACJ,QAAQ,YACR,SAAS;IAAqB;MAC7B,QAAQ;UAIG;UAoBA;UA+BA;UAoCN;UASA"}
+{"version":3,"file":"index.d.cts","names":[],"sources":["../src/types.ts","../src/markitdown.ts"],"mappings":";KAAY;EACV;EACA;;EAEA;;;;;;;;KASU,aAAa,aAAa,kBAAkB,kBAAkB;KAE9D;;EAEV;;EAEA;;EAEA,eAAe;;EAEf;EACA;EACA;EACA;EACA;EACA,qBAAqB;EACrB,qBAAqB;EACrB;;UAGe;EACf,QAAQ,QAAQ,YAAY,SAAS,mBAAmB,QAAQ;;;;KCnBtD,mBAAmB,aAAa,WAAW;cAE1C;mBACM;EAEjB;;EAeM,QACJ,QAAQ,kBACR,UAAS,mBACR,QAAQ;;;;EA+BL,cACJ,QAAQ,YACR,SAAS;IAAqB;MAC7B,QAAQ;UAIG;UAoBA;UA+BA;UAuCN;UASA"}

package/dist/index.d.ts

@@ -27,6 +27,7 @@ type ConverterOptions = {
   idPrefix?: string;
   transformDocument?: (element: unknown) => unknown;
   _parent_converters?: DocumentConverter[];
+  debug?: boolean;
 };
 interface DocumentConverter {
   convert(source: Uint8Array, options: ConverterOptions): Promise<ConverterResult>;

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","names":[],"sources":["../src/types.ts","../src/markitdown.ts"],"mappings":";KAAY;EACV;EACA;;EAEA;;;;;;;;KASU,aAAa,aAAa,kBAAkB,kBAAkB;KAE9D;;EAEV;;EAEA;;EAEA,eAAe;;EAEf;EACA;EACA;EACA;EACA;EACA,qBAAqB;EACrB,qBAAqB;;UAGN;EACf,QAAQ,QAAQ,YAAY,SAAS,mBAAmB,QAAQ;;;;KClBtD,mBAAmB,aAAa,WAAW;cAE1C;mBACM;EAEjB;;EAeM,QACJ,QAAQ,kBACR,UAAS,mBACR,QAAQ;;;;EA+BL,cACJ,QAAQ,YACR,SAAS;IAAqB;MAC7B,QAAQ;UAIG;UAoBA;UA+BA;UAoCN;UASA"}
+{"version":3,"file":"index.d.ts","names":[],"sources":["../src/types.ts","../src/markitdown.ts"],"mappings":";KAAY;EACV;EACA;;EAEA;;;;;;;;KASU,aAAa,aAAa,kBAAkB,kBAAkB;KAE9D;;EAEV;;EAEA;;EAEA,eAAe;;EAEf;EACA;EACA;EACA;EACA;EACA,qBAAqB;EACrB,qBAAqB;EACrB;;UAGe;EACf,QAAQ,QAAQ,YAAY,SAAS,mBAAmB,QAAQ;;;;KCnBtD,mBAAmB,aAAa,WAAW;cAE1C;mBACM;EAEjB;;EAeM,QACJ,QAAQ,kBACR,UAAS,mBACR,QAAQ;;;;EA+BL,cACJ,QAAQ,YACR,SAAS;IAAqB;MAC7B,QAAQ;UAIG;UAoBA;UA+BA;UAuCN;UASA"}

package/dist/index.js

@@ -3,7 +3,7 @@ import TurndownService from "turndown";
 import { parseHTML } from "linkedom";
 import mammoth from "mammoth";
 import { extractText, getDocumentProxy, getMeta } from "unpdf";
-import * as XLSX from "xlsx";
+import ExcelJS from "exceljs";
 import { DOMParser } from "@xmldom/xmldom";
 import { unzipSync } from "fflate";
 //#region src/bytes.ts
@@ -133,12 +133,13 @@ var HtmlConverter = class {
 	}
 	async convertHtml(html) {
 		const document = parseHtmlDocument(html);
-		document.querySelectorAll("script, style").forEach((element) => {
+		const title = document.title;
+		document.querySelectorAll("script, style, title").forEach((element) => {
 			element.remove();
 		});
-		const markdown = new CustomTurnDown().convertSoup(document.body ?? document);
+		const markdown = new CustomTurnDown().convertSoup(document);
 		return {
-			title: document.title || null,
+			title: title || null,
 			markdown,
 			text_content: markdown
 		};
@@ -149,14 +150,17 @@ var HtmlConverter = class {
 var DocxConverter = class extends HtmlConverter {
 	async convert(source, options) {
 		if (options.file_extension?.toLowerCase() !== ".docx") return null;
-		const html = await mammoth.convertToHtml({ arrayBuffer: toArrayBuffer(source) }, {
+		let mammothInput = { arrayBuffer: toArrayBuffer(source) };
+		if (typeof process !== "undefined" && process.versions !== null && process.versions.node !== null) mammothInput = { buffer: Buffer.from(source) };
+		const mammothOptions = {
 			styleMap: options.styleMap,
 			includeEmbeddedStyleMap: options.includeEmbeddedStyleMap,
 			includeDefaultStyleMap: options.includeDefaultStyleMap,
 			ignoreEmptyParagraphs: options.ignoreEmptyParagraphs,
-			idPrefix: options.idPrefix,
-			transformDocument: options.transformDocument
-		});
+			idPrefix: options.idPrefix
+		};
+		if (options.transformDocument) mammothOptions.transformDocument = options.transformDocument;
+		const html = await mammoth.convertToHtml(mammothInput, mammothOptions);
 		return this.convertHtml(html.value);
 	}
 };
@@ -278,26 +282,45 @@ var WikipediaConverter = class {
 };
 //#endregion
 //#region src/converters/xlsx.ts
+const { Workbook } = ExcelJS;
 var XlsxConverter = class extends HtmlConverter {
 	async convert(source, options) {
 		if (options.file_extension?.toLowerCase() !== ".xlsx") return null;
-		const workbook = XLSX.read(source, { type: "array" });
+		const workbook = new Workbook();
+		await workbook.xlsx.load(toArrayBuffer(source));
 		const sections = [];
-		for (const sheetName of workbook.SheetNames) {
-			const sheet = workbook.Sheets[sheetName];
-			if (sheet?.["!ref"]) {
-				const result = await this.convertHtml(XLSX.utils.sheet_to_html(sheet));
-				sections.push(`## ${sheetName}\n${result?.markdown.trim() ?? ""}`);
-			}
+		for (const worksheet of workbook.worksheets) {
+			const html = worksheetToHtml(worksheet);
+			if (!html) continue;
+			const result = await this.convertHtml(html);
+			sections.push(`## ${worksheet.name}\n${result?.markdown.trim() ?? ""}`);
 		}
 		const markdown = sections.join("\n\n");
 		return {
-			title: workbook.Props?.Title || "Untitled",
+			title: workbook.title || "Untitled",
 			markdown,
 			text_content: markdown
 		};
 	}
 };
+function worksheetToHtml(worksheet) {
+	const columnCount = worksheet.columnCount;
+	if (!columnCount || !worksheet.actualRowCount) return "";
+	const rows = [];
+	worksheet.eachRow({ includeEmpty: true }, (row) => {
+		const cells = [];
+		for (let column = 1; column <= columnCount; column++) cells.push(`<td>${escapeHtml(cellText(row.getCell(column)))}</td>`);
+		rows.push(`<tr>${cells.join("")}</tr>`);
+	});
+	return `<html><body><table>${rows.join("")}</table></body></html>`;
+}
+function cellText(cell) {
+	if (cell.isMerged && cell.master !== cell) return "";
+	return cell.text;
+}
+function escapeHtml(value) {
+	return value.replace(/&/g, "&amp;").replace(/</g, "&lt;").replace(/>/g, "&gt;").replace(/"/g, "&quot;");
+}
 //#endregion
 //#region src/converters/xml-rss-atom.ts
 var RSSConverter = class {
@@ -500,7 +523,7 @@ var MarkItDown = class {
 		if (typeof source === "string" || source instanceof URL) return this.convertUrl(source.toString(), options);
 		if (isResponse(source)) return this.convertResponse(source, options);
 		const extensions = /* @__PURE__ */ new Set();
-		this.addExtension(extensions, options.file_extension);
+		this.addExtension(extensions, options.file_extension?.replace(/^\./, ""));
 		if (isBlob(source)) {
 			this.addExtension(extensions, extensionFromMime(source.type));
 			if ("name" in source && typeof source.name === "string") this.addExtension(extensions, extensionFromPathname(source.name));
@@ -553,6 +576,7 @@ var MarkItDown = class {
 				return result;
 			}
 		} catch (error) {
+			if (options.debug) console.error("Converter threw error:", error);
 			lastError = error;
 		}
 		const formats = [...extensions].join(", ");