OpenXmlSdkTs — Open XML SDK for TypeScript

A TypeScript library for reading, writing, and manipulating Office Open XML documents (.docx, .xlsx, .pptx) in Node.js and browser environments. Inspired by the .NET Open-Xml-Sdk, this library brings the same familiar programming model to the TypeScript ecosystem.

Documentation

Full API reference and guides are published at ericwhitedev.github.io/OpenXmlSdkTs.

To build the documentation locally:

npm run docs
Copy

Output is written to docs/api/. Open docs/api/index.html in a browser.

Why OpenXmlSdkTs?

• Full document format support — Work with Word, Excel, and PowerPoint files at the XML level.
• Three I/O modes — Open and save documents as binary blobs (via JSZip), Flat OPC XML strings (the required format when building Office JavaScript/TypeScript add-ins), or Base64 strings.
• Friendly content & relationship types — All Open XML content types and relationship types are referenced using readable labels (e.g. RelationshipType.styles, ContentType.mainDocument) instead of long, error-prone URIs.
• Pre-initialized namespace, element, and attribute names — Static classes (W, S, P, A, etc.) provide pre-initialized XName and XNamespace objects for every element and attribute in the Open XML specification. Because XNamespace and XName objects are atomized (two objects with the same namespace and local name are the same object), this gives excellent performance when querying and modifying markup.
• Built on LINQ to XML for TypeScript — Powered by ltxmlts, a faithful TypeScript port of .NET's LINQ to XML. Query and transform XML with elements(), descendants(), attributes(), and the rest of the LINQ to XML API you already know.
• Intuitive part navigation — Navigate from the package to the main document part, then from part to part using typed methods that mirror the .NET SDK: mainDocumentPart(), styleDefinitionsPart(), worksheetParts(), slideParts(), and many more.
• Lightweight — Only two runtime dependencies: jszip and ltxmlts.
• MIT licensed — Free for commercial and open-source use. This is the same license as the C#/dotnet Open-Xml-Sdk.

Installation

npm install openxmlsdkts
Copy

Both openxmlsdkts and its companion library ltxmlts are available on npmjs.

Class Hierarchy

OpenXmlPackage                      Base class for all Office document packages
├── WmlPackage                      Word (.docx) packages
├── SmlPackage                      Excel (.xlsx) packages
└── PmlPackage                      PowerPoint (.pptx) packages

OpenXmlPart                         Base class for all parts within a package
├── WmlPart                         Word parts (document, styles, headers, footers, etc.)
├── SmlPart                         Excel parts (workbook, worksheets, charts, etc.)
└── PmlPart                         PowerPoint parts (presentation, slides, masters, etc.)

OpenXmlRelationship                 Represents a relationship between a package/part and a target

ContentType                         Static lookup of content type labels → URIs
RelationshipType                    Static lookup of relationship type labels → URIs

Namespace Classes (static)          Pre-initialized XNamespace / XName objects
├── W                               WordprocessingML
├── S                               SpreadsheetML
├── P                               PresentationML
├── A                               DrawingML
├── R, M, MC, W14, ...             30+ additional namespace classes
└── NoNamespace                     Attributes with no namespace
Copy

Quick Start

Load a Word Document, Modify It, and Save

import * as fs from "fs";
import { WmlPackage, W, WmlPart } from "openxmlsdkts";

async function capitalizeFirstWord() {
  // 1. Load the document from a file
  const buffer = fs.readFileSync("my-document.docx");
  const blob = new Blob([buffer]);
  const doc = await WmlPackage.open(blob);

  // 2. Get the main document part and its XML
  const mainPart: WmlPart | undefined = await doc.mainDocumentPart();
  if (!mainPart) throw new Error("No main document part found");

  const xDoc = await mainPart.getXDocument();

  // 3. Find the first word of the first run of text and capitalize it
  const body = xDoc.root!.element(W.body);
  if (body) {
    const firstParagraph = body.element(W.p);
    if (firstParagraph) {
      const firstRun = firstParagraph.element(W.r);
      if (firstRun) {
        const textElement = firstRun.element(W.t);
        if (textElement && textElement.value) {
          const words = textElement.value.split(" ");
          words[0] = words[0].toUpperCase();
          textElement.value = words.join(" ");
        }
      }
    }
  }

  // 4. Save the modified XML back to the part
  mainPart.putXDocument(xDoc);

  // 5. Save the document back to a file
  const savedBlob = await doc.saveToBlobAsync();
  const arrayBuffer = await savedBlob.arrayBuffer();
  fs.writeFileSync("my-document-modified.docx", Buffer.from(arrayBuffer));
}

capitalizeFirstWord();
Copy

Opening and Saving Documents

OpenXmlSdkTs supports three document formats. The open() method auto-detects the format from the input.

Binary (Blob)

The standard .docx / .xlsx / .pptx ZIP-based format.

// Open
const buffer = fs.readFileSync("report.docx");
const doc = await WmlPackage.open(new Blob([buffer]));

// Save
const blob = await doc.saveToBlobAsync();
Copy

Flat OPC (XML String)

A single-file XML representation of the entire package. This is the format you must work with when building Office JavaScript/TypeScript add-in applications.

// Open
const flatOpcXml = fs.readFileSync("report.xml", "utf-8");
const doc = await WmlPackage.open(flatOpcXml);

// Save
const flatOpc: string = await doc.saveToFlatOpcAsync();
Copy

Base64 String

Convenient for APIs, data URIs, and serialization scenarios.

// Open
const base64 = fs.readFileSync("report.txt", "utf-8"); // base64-encoded docx
const doc = await WmlPackage.open(base64);

// Save
const savedBase64: string = await doc.saveToBase64Async();
Copy

Navigating Parts

Each package type provides typed methods to access its constituent parts.

Word Documents (WmlPackage / WmlPart)

const doc = await WmlPackage.open(blob);

// Package-level
const mainPart = await doc.mainDocumentPart();
const contentParts = await doc.contentParts(); // main + headers + footers + footnotes + endnotes

// Part-level navigation
const styles = await mainPart!.styleDefinitionsPart();
const theme = await mainPart!.themePart();
const numbering = await mainPart!.numberingDefinitionsPart();
const headers = await mainPart!.headerParts();
const footers = await mainPart!.footerParts();
const comments = await mainPart!.wordprocessingCommentsPart();
const fonts = await mainPart!.fontTablePart();
const settings = await mainPart!.documentSettingsPart();
Copy

Excel Spreadsheets (SmlPackage / SmlPart)

const doc = await SmlPackage.open(blob);
const workbook = await doc.workbookPart();

const worksheets = await workbook!.worksheetParts();
const sharedStrings = await workbook!.sharedStringTablePart();
const wbStyles = await workbook!.workbookStylesPart();
const charts = await workbook!.chartsheetParts();
const pivotTables = await workbook!.pivotTableParts();
Copy

PowerPoint Presentations (PmlPackage / PmlPart)

const doc = await PmlPackage.open(blob);
const presentation = await doc.presentationPart();

const slides = await presentation!.slideParts();
const masters = await presentation!.slideMasterParts();
const layout = await presentation!.slideLayoutPart();
const notesMaster = await presentation!.notesMasterPart();
Copy

Working with XML

All XML manipulation uses the LINQ to XML API from ltxmlts. The pre-initialized namespace classes make queries concise and performant.

import { WmlPackage, W, NoNamespace, XElement } from "openxmlsdkts";

const doc = await WmlPackage.open(blob);
const mainPart = await doc.mainDocumentPart();
const xDoc = await mainPart!.getXDocument();

// Query all paragraphs
const paragraphs = xDoc.root!.element(W.body)!.elements(W.p);

// Find paragraphs with a specific style
for (const para of paragraphs) {
  const pPr = para.element(W.pPr);
  const pStyle = pPr?.element(W.pStyle);
  const styleVal = pStyle?.attribute(NoNamespace.val)?.value;
  if (styleVal === "Heading1") {
    console.log("Found Heading1 paragraph");
  }
}

// Add a new paragraph
const newPara = new XElement(W.p,
  new XElement(W.r,
    new XElement(W.t, "Hello from OpenXmlSdkTs!")
  )
);
xDoc.root!.element(W.body)!.add(newPara);

// Write the modified XML back to the part
mainPart!.putXDocument(xDoc);
Copy

Content Types and Relationship Types

Instead of using long URIs, use the ContentType and RelationshipType static lookups.

import { ContentType, RelationshipType } from "openxmlsdkts";

// Navigate parts by relationship type
const stylePart = await mainPart!.getPartByRelationshipType(RelationshipType.styles);
const themePart = await mainPart!.getPartByRelationshipType(RelationshipType.theme);
const imageParts = await mainPart!.getPartsByRelationshipType(RelationshipType.image);

// Query relationships
const rels = await mainPart!.getRelationships();
for (const rel of rels) {
  console.log(rel.getType());    // e.g. the styles relationship URI
  console.log(rel.getTarget());  // e.g. "styles.xml"
}
Copy

Pre-Initialized Namespace Classes

Over 40 static classes cover every namespace in the Open XML specification:

Class	Namespace	Description
W	wordprocessingml/2006/main	WordprocessingML elements and attributes
S	spreadsheetml/2006/main	SpreadsheetML elements and attributes
P	presentationml/2006/main	PresentationML elements and attributes
A	drawingml/2006/main	DrawingML elements and attributes
R	officeDocument/2006/relationships	Relationship attributes
M	officeDocument/2006/math	Office Math
W14	microsoft.com/office/word/2010/wordml	Word 2010 extensions
MC	markup-compatibility/2006	Markup Compatibility
NoNamespace	(none)	Attributes with no namespace (e.g. val, id)
...	...	30+ additional namespace classes

Each class exposes a namespace property and XName properties for every element and attribute:

W.namespace  // XNamespace for WordprocessingML
W.body       // XName for <w:body>
W.p          // XName for <w:p>
W.r          // XName for <w:r>
W.t          // XName for <w:t>
Copy

Because XName and XNamespace objects are atomized, equality checks are identity checks (===), giving excellent query performance.

Use with Office Add-ins

Flat OPC is the document format you must use when building Word, Excel, and PowerPoint JavaScript/TypeScript add-ins:

// In an Office Add-in — get the document as Flat OPC via the Office.js API
// then manipulate with OpenXmlSdkTs
const doc = await WmlPackage.open(flatOpcString);
const mainPart = await doc.mainDocumentPart();
// ... modify the document ...
const modifiedFlatOpc = await doc.saveToFlatOpcAsync();
// Set the modified Flat OPC back via the Office.js API
Copy

API Reference

Packages

Class	Description	Key Methods
WmlPackage	Word documents	open(), mainDocumentPart(), contentParts(), saveToBase64Async(), saveToBlobAsync(), saveToFlatOpcAsync()
SmlPackage	Excel spreadsheets	open(), workbookPart(), saveToBase64Async(), saveToBlobAsync(), saveToFlatOpcAsync()
PmlPackage	PowerPoint presentations	open(), presentationPart(), saveToBase64Async(), saveToBlobAsync(), saveToFlatOpcAsync()

Parts

Class	Notable Navigation Methods
WmlPart	headerParts(), footerParts(), styleDefinitionsPart(), themePart(), numberingDefinitionsPart(), fontTablePart(), documentSettingsPart(), endnotesPart(), footnotesPart(), wordprocessingCommentsPart(), glossaryDocumentPart()
SmlPart	worksheetParts(), chartsheetParts(), sharedStringTablePart(), workbookStylesPart(), calculationChainPart(), pivotTableParts(), tableDefinitionParts()
PmlPart	slideParts(), slideMasterParts(), slideLayoutPart(), notesMasterPart(), notesSlidePart(), handoutMasterPart(), commentAuthorsPart(), presentationPropertiesPart(), viewPropertiesPart()

Common Methods (All Parts)

Method	Description
getXDocument()	Get the part's XML as an XDocument
putXDocument(xDoc)	Write modified XML back to the part
getRelationships()	Get all relationships from this part
getPartByRelationshipType(type)	Get a single related part by relationship type
getPartsByRelationshipType(type)	Get all related parts by relationship type
getUri()	Get the part's URI within the package
getContentType()	Get the part's content type

Dependencies

Package	Purpose
ltxmlts	LINQ to XML for TypeScript — XML querying and manipulation
jszip	ZIP compression for binary Open XML packages

Requirements

• Node.js 18+ or modern browser with Blob support
• TypeScript 5.0+ (recommended)

License

MIT License. Copyright (c) 2026 Eric White.

Author

Eric White

• ericwhite.com
• LinkedIn
• eric@ericwhite.com