Document AI Processor with Node.js

This project demonstrates how to use Google Cloud's Document AI API to process a document (e.g., an image) and extract text and paragraph information. The script is written in Node.js and uses the @google-cloud/documentai library.

Features

• Process an image file (e.g., JPEG) using Google Cloud's Document AI API.
• Extract and display text and paragraphs from the document.
• Provides a starting point for integrating Document AI into your Node.js projects.

---

Prerequisites

1. Google Cloud Setup

1. Enable the Document AI API on your Google Cloud Console.
2. Create a processor in the Document AI section of the Cloud Console. Note the PROJECT_ID, LOCATION, and PROCESSOR_ID values.
3. Set up a service account with the required permissions for Document AI.
4. Download the service account key JSON file and set the path in the GOOGLE_APPLICATION_CREDENTIALS environment variable.

2. Node.js Setup

1. Install Node.js (v14 or later is recommended).
2. Clone this repository:

   git clone <repository-url>
   cd <repository-name>

3. Install dependencies:

   npm install

---

Environment Variables

Create a .env file in the root directory with the following variables:

PROJECT_ID=<Your Google Cloud Project ID>
LOCATION=<Processor Location, e.g., 'us' or 'eu'>
PROCESSOR_ID=<Your Processor ID>
GOOGLE_APPLICATION_CREDENTIALS=<Path to your service account key JSON file>

---

Usage

1. Prepare the document you want to process (e.g., an image file) and place it in the data/ folder. Update the filePath in the script if necessary.

2. Run the script:

   node index.js

3. The extracted text and paragraph information will be logged to the console.

---

Code Structure

• index.js: Main script for processing documents using the Document AI API.
• data/two.jpeg: Sample image file for testing (you can replace this with your own image).
• .env: Environment variables file (excluded from version control).

---

Output Example

Here’s an example of the output:

Starting the quickstart function...
Processor name: projects/<PROJECT_ID>/locations/<LOCATION>/processors/<PROCESSOR_ID>
File read successfully: data/two.jpeg
Image file encoded to base64.
Request object created.
Document processed successfully.
Text extracted from the document.
The document contains the following paragraphs:
Paragraph text:
<Extracted text from the document>

---

Dependencies

• dotenv
• @google-cloud/documentai
• Node.js fs module for file handling.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contributing

Feel free to open issues or create pull requests if you have suggestions or improvements.

---

References

• Google Cloud Document AI Documentation
• Google Cloud Node.js Client Libraries

---

Author

Asutosh Sidhya

Check out my GitHub profile.

---

Disclaimer

Ensure you handle sensitive data, such as API keys and documents, securely and avoid committing them to your repository.

Instructions

1. Save this content into a README.md file in your repository.
2. Replace placeholders like <Your Google Cloud Project ID> and <repository-url> with your specific details.
3. Include a .gitignore file to exclude sensitive files such as .env and GOOGLE_APPLICATION_CREDENTIALS.

Let me know if you need additional adjustments or further enhancements!