LangChain Document Embedding API

This is a Flask API for extracting text from PDFs, PowerPoints, XL Spreadsheets, Youtube Channels, and more.

Prerequisites

• Docker
• Docker Compose

Usage

1. Clone the repository and navigate to the root directory.
2. Run the following command to start the API:

docker-compose up

This will start the API on http://localhost:5000.

3. To extract text from a PDF, make a POST request to http://localhost:5000/getTextForPDF with the following JSON data:

{
"url": "https://example.com/path/to/pdf"
}

This will return the extracted text from the PDF as a JSON response.

4. To extract text from an HTML page, make a POST request to http://localhost:5000/getTextForURL with the following JSON data:

{
"url": "https://example.com"
}

This will return the extracted text and title from the HTML page as a JSON response.

Configuration

The following environment variables can be set to configure the API:

• PORT: The port on which the API should listen. Default is 5000.

Dependencies

The API uses the following dependencies:

• Flask==2.2.2
• Flask-Cors==3.0.10
• gunicorn==20.1.0
• selenium==4.7.2
• PyPDF2==3.0.1
• timeout-decorator==0.5.0
• beautifulsoup4==4.11.2