tinydocx

Minimal DOCX/ODT creation library. <1K LOC, zero dependencies.

npm install tinydocx

Why tinydocx?

| | tinydocx | docx | | --- | --- | --- | | Size (gzip) | 8.1 KB | 108.6 KB | | Dependencies | 0 | 5 |

~13x smaller. Everything you need: text, tables, lists, images, links, headers/footers. DEFLATE-compressed output.

Features

| Feature | Description | | --- | --- | | Text | Any size, bold/italic/underline, hex colors, custom fonts | | Paragraphs | Alignment (left/center/right), multi-line joining | | Headings | H1–H6 | | Lists | Bullet and numbered, multi-level (5 levels), nested | | Tables | With borders and column widths | | Hyperlinks | External links | | Images | PNG, JPEG, GIF, WebP — DOCX and ODT | | Headers/Footers | With page numbers — DOCX and ODT | | Page Numbers | DOCX and ODT | | Markdown | Convert to DOCX/ODT | | ODT | Full feature parity with DOCX | | Compression | DEFLATE (ZIP method 8), memory-optimized |

Quick start

import { docx } from 'tinydocx'
import { writeFileSync } from 'fs'

const doc = docx()
doc.content((ctx) => {
  ctx.heading('Hello World', 1)
  ctx.paragraph('This is a paragraph.')
  ctx.paragraph('Bold text', { bold: true })
  ctx.paragraph('Centered', { align: 'center' })
})

writeFileSync('output.docx', doc.build())

ODT

import { odt } from 'tinydocx'

const doc = odt()
doc.header((ctx) => ctx.paragraph('Company Name', { bold: true }))
doc.footer((ctx) => ctx.pageNumber())
doc.content((ctx) => {
  ctx.heading('Hello World', 1)
  ctx.paragraph('Same API, different format.')
  ctx.image(readFileSync('logo.png'), { width: 2, height: 1 })
})

writeFileSync('output.odt', doc.build())

API

import { docx, odt, markdownToDocx, markdownToOdt } from 'tinydocx'

const doc = docx() // or odt() — same API

doc.content((ctx) => {
  ctx.heading(str, level)
  ctx.paragraph(str, opts?)
  ctx.text(str, size, opts?)
  ctx.lineBreak()
  ctx.horizontalRule()
  ctx.list(items, ordered?)
  ctx.table(rows, opts?)
  ctx.link(text, url, opts?)
  ctx.image(data, opts)
  ctx.pageNumber()
})

doc.header((ctx) => { ... })
doc.footer((ctx) => { ... })

doc.build() // returns Uint8Array (DEFLATE-compressed ZIP)

TextOptions

{
  align?: 'left' | 'center' | 'right'
  bold?: boolean
  italic?: boolean
  underline?: boolean
  strikethrough?: boolean
  code?: boolean
  color?: string
  font?: string
  size?: number
}

Lists

ctx.list(['Item 1', 'Item 2', 'Item 3'])
ctx.list(['First', 'Second', 'Third'], true)

Tables

ctx.table([
  ['Name', 'Age'],
  ['Alice', '30'],
  ['Bob', '25']
])

ctx.table([['Wide', 'Narrow']], { colWidths: [4000, 2000] })

Images

const imageData = readFileSync('photo.png')
ctx.image(imageData, { width: 4, height: 3 })

Hyperlinks

ctx.link('Click here', 'https://example.com')
ctx.link('Styled', 'https://example.com', { bold: true })

Headers/Footers

doc.header((ctx) => ctx.paragraph('Company', { bold: true }))
doc.footer((ctx) => ctx.pageNumber())

Markdown

import { markdownToDocx, markdownToOdt } from 'tinydocx'

const md = `
# Heading 1
## Heading 2

This is **bold**, *italic*, ~~strikethrough~~, and \`inline code\`.

- Bullet list
  - Nested item
- Another item

1. Numbered list
2. Second item

> Blockquote with **formatting**

| Name | Age |
|------|-----|
| Alice | 30 |

\`\`\`javascript
const code = 'block'
\`\`\`

[Link text](https://example.com)

---
`

writeFileSync('output.docx', markdownToDocx(md))
writeFileSync('output.odt', markdownToOdt(md))

Supported syntax:

| Feature | Syntax | |---------|--------| | Headings | # ## ### #### ##### ###### | | Bold | **text** or __text__ | | Italic | *text* or _text_ | | Strikethrough | ~~text~~ | | Inline code | `code` | | Links | [text](url) | | Images | ![alt](url) | | Bullet lists | - item or * item | | Numbered lists | 1. item | | Nested lists | Indent with spaces | | Tables | \| col \| col \| | | Code blocks | ``` | | Blockquotes | > text | | Horizontal rule | --- or *** or ___ | | Multi-line paragraphs | Consecutive lines joined automatically |

Full example

import { docx } from 'tinydocx'

const doc = docx()

doc.header((ctx) => ctx.paragraph('ACME CORP', { bold: true }))
doc.footer((ctx) => ctx.pageNumber())

doc.content((ctx) => {
  ctx.heading('INVOICE', 1)
  ctx.text('#INV-2025-001', 10, { color: '#666666' })
  ctx.lineBreak()

  ctx.table([
    ['Description', 'Amount'],
    ['Website Development', '$5,000.00'],
    ['Hosting (Annual)', '$200.00']
  ], { colWidths: [6000, 3000] })
  ctx.lineBreak()

  ctx.paragraph('Total: $5,200.00', { bold: true, align: 'right' })
  ctx.link('View online', 'https://example.com/invoice')
})

writeFileSync('invoice.docx', doc.build())

How it works

DOCX and ODT files are ZIP archives containing XML. tinydocx generates these directly using DEFLATE compression (node:zlib) with a memory-optimized single-pass buffer allocation.

DOCX:                       ODT:
[Content_Types].xml         mimetype
_rels/.rels                 META-INF/manifest.xml
word/document.xml           content.xml
word/styles.xml             styles.xml
word/numbering.xml          Pictures/
word/header1.xml
word/footer1.xml
word/media/

License

MIT