API Reference

AnonDocs provides a RESTful API for anonymizing text and documents. You can interact with the API directly using HTTP requests or use our official SDK for TypeScript/JavaScript.

SDK (Recommended)

The easiest way to use AnonDocs is through our official SDK.

Installation

npm install @aismarttalk/anondocs-sdk

Quick Start

import { AnonDocsClient } from '@aismarttalk/anondocs-sdk';

// Initialize the client
const client = new AnonDocsClient({
  baseUrl: 'http://localhost:3000'
});

// Anonymize text
const result = await client.anonymizeText(
  'My name is John Smith and my email is john@example.com'
);

console.log(result.anonymizedText);
// Output: "My name is [NAME] and my email is [EMAIL]"

console.log(result.piiDetected);
// Output: { names: ['John Smith'], emails: ['john@example.com'], ... }

SDK Features

• 🔒 Privacy-first anonymization
• 📄 Document support (PDF, DOCX, TXT)
• ✨ NEW in v1.2.0: DOCX formatting preservation with downloadable files
• 🔄 NEW in v1.2.0: Precise PII replacement mappings
• 🌊 Streaming with real-time progress updates
• 🎯 Full TypeScript support
• 🚀 Works in Node.js and Browser
• ⚡ Automatic provider selection

SDK Methods

Anonymize Text

const result = await client.anonymizeText('Your text here', {
  provider: 'ollama' // Optional, omit to use default
});

console.log(result.anonymizedText);
console.log(result.piiDetected);
console.log(result.replacements); // NEW in v1.2.0 - precise PII mappings

Anonymize Document

Browser:

const fileInput = document.querySelector('input[type="file"]');
const file = fileInput.files[0];

const result = await client.anonymizeDocument(file);

// NEW in v1.2.0: DOCX files return download URL for formatted documents
if (result.downloadUrl) {
  // DOCX file with preserved formatting
  window.open(result.downloadUrl); // Download formatted DOCX
  console.log('Original:', result.originalFilename);
  console.log('Anonymized:', result.filename);
} else {
  // PDF or TXT - plain text only
  console.log(result.anonymizedText);
}

Node.js:

import { readFileSync } from 'fs';

const fileBuffer = readFileSync('./document.docx');
const result = await client.anonymizeDocument(fileBuffer);

// Download formatted DOCX (v1.2.0+)
if (result.downloadUrl) {
  const response = await fetch(result.downloadUrl);
  const buffer = await response.arrayBuffer();
  writeFileSync(result.filename, Buffer.from(buffer));
}

Streaming Anonymization

Get real-time progress updates:

await client.streamAnonymizeText(longText, {
  onProgress: (event) => {
    console.log(`${event.progress}% - ${event.message}`);
  },
  onComplete: (result) => {
    console.log('Done!', result.anonymizedText);
  },
  onError: (error) => {
    console.error('Error:', error.message);
  }
});

Health Check

const health = await client.health();
// Returns: { status: 'ok', timestamp: '2025-11-04T...' }

SDK Documentation

For complete SDK documentation, visit: npm package

---

REST API

If you prefer to use the REST API directly, here are the available endpoints.

Base URL

http://localhost:3000

Endpoints

Anonymize Text

POST /api/anonymize

Anonymize plain text input.

Request Body:

{
  "text": "Your text to anonymize",
  "provider": "ollama"
}

Response:

{
  "success": true,
  "data": {
    "anonymizedText": "Anonymized text",
    "piiDetected": {
      "names": ["John Doe"],
      "emails": ["john@example.com"],
      "phoneNumbers": [],
      "addresses": [],
      "dates": [],
      "organizations": [],
      "other": []
    },
    "replacements": [
      {"original": "John Doe", "anonymized": "[NAME]"},
      {"original": "john@example.com", "anonymized": "[EMAIL]"}
    ],
    "chunksProcessed": 1,
    "wordsPerMinute": 450,
    "processingTimeMs": 1234
  }
}

Anonymize Document

POST /api/document

Upload and anonymize PDF, DOCX, or TXT files.

Request: Multipart form data

• file: The document file
• provider: (optional) "ollama" or "openai"

Response (DOCX files):

{
  "success": true,
  "data": {
    "anonymizedText": "Anonymized text",
    "piiDetected": { ... },
    "replacements": [ ... ],
    "chunksProcessed": 1,
    "wordsPerMinute": 450,
    "processingTimeMs": 1234,
    "downloadUrl": "http://localhost:3000/api/document/download/anonymized-123.docx",
    "filename": "anonymized-123.docx",
    "originalFilename": "original.docx"
  }
}

Response (PDF/TXT files): Same as text anonymization (no downloadUrl)

Stream Text Anonymization

POST /api/stream/anonymize

Stream anonymization with real-time progress updates using Server-Sent Events (SSE).

Request Body:

{
  "text": "Your text to anonymize",
  "provider": "ollama"
}

Response: text/event-stream with progress events

Stream Document Anonymization

POST /api/stream/document

Stream document anonymization with real-time progress updates.

Request: Multipart form data

• file: The document file
• provider: (optional) "ollama" or "openai"

Response: text/event-stream with progress events

Health Check

GET /health

Check API health status.

Response:

{
  "status": "ok",
  "timestamp": "2024-11-03T12:00:00.000Z"
}

Error Responses

{
  "success": false,
  "error": "Error message"
}

Provider Options

• ollama: Use local Ollama instance (default)
• openai: Use OpenAI-compatible API (vLLM, LM Studio, etc.)