@vertesia/converters

Image and document conversion utilities for Node.js. Provides functions for image transformation, PDF to text extraction, and document to Markdown conversion.

Features

• Image Transformation: Resize and convert images using Sharp
• PDF to Text: Extract text from PDF files using MuTool
• Document to Markdown: Convert various document formats to Markdown using Pandoc

Installation

npm install @vertesia/converters
# or
pnpm add @vertesia/converters

System Dependencies

Some converters require external tools to be installed:

• Image conversion: No external dependencies (uses Sharp)
• PDF to text: Requires MuTool (mutool command)
• Document to Markdown: Requires Pandoc (pandoc command)

Usage

Image Transformation

Transform images with resizing and format conversion:

import {
  transformImage,
  transformImageToBuffer,
  transformImageToFile
} from '@vertesia/converters';

// Transform image to a stream
import { createReadStream, createWriteStream } from 'fs';

const input = createReadStream('input.jpg');
const output = createWriteStream('output.webp');

await transformImage(input, output, {
  max_hw: 1024,      // Max width/height (maintains aspect ratio)
  format: 'webp'     // Output format
});

// Transform image to buffer
const buffer = await transformImageToBuffer(inputBuffer, {
  max_hw: 800,
  format: 'png'
});

// Transform image to file
await transformImageToFile(inputBuffer, 'output.jpg', {
  max_hw: 1200,
  format: 'jpeg'
});

PDF to Text

Extract text content from PDF files:

import { pdfToText, pdfToTextBuffer, pdfFileToText } from '@vertesia/converters';

// From buffer to string
const text = await pdfToText(pdfBuffer);

// From buffer to buffer
const textBuffer = await pdfToTextBuffer(pdfBuffer);

// From file to file
await pdfFileToText('input.pdf', 'output.txt');

Document to Markdown

Convert documents to Markdown format using Pandoc:

import { manyToMarkdown } from '@vertesia/converters';
import { createReadStream } from 'fs';

// Convert DOCX to Markdown
const stream = createReadStream('document.docx');
const markdown = await manyToMarkdown(stream, 'docx');

// Convert HTML to Markdown
const htmlStream = createReadStream('page.html');
const md = await manyToMarkdown(htmlStream, 'html');

Supported input formats include all formats supported by Pandoc: docx, html, latex, rst, textile, org, mediawiki, and many more.

API Reference

Image Functions

| Function | Description | |----------|-------------| | transformImage(input, output, opts) | Transform image from stream to stream | | transformImageToBuffer(input, opts) | Transform image to buffer | | transformImageToFile(input, output, opts) | Transform image to file |

TransformOptions

| Option | Type | Description | |--------|------|-------------| | max_hw | number | Maximum width/height (maintains aspect ratio, no upscaling) | | format | string | Output format (jpeg, png, webp, avif, etc.) |

PDF Functions

| Function | Description | |----------|-------------| | pdfToText(buffer) | Convert PDF buffer to text string | | pdfToTextBuffer(buffer) | Convert PDF buffer to text buffer | | pdfFileToText(input, output) | Convert PDF file to text file |

Document Functions

| Function | Description | |----------|-------------| | manyToMarkdown(stream, format) | Convert document stream to Markdown |

Requirements

• Node.js 18+
• MuTool (for PDF conversion)
• Pandoc (for document conversion)

License

Apache-2.0