PDF Analyzer Application

Overview

The PDF Analyzer Application is a FastAPI-based application designed to process PDF files, allowing users to upload PDFs, ask questions about their content, analyze sentiment, and generate new PDF documents based on user-defined instructions. The application utilizes the Ollama API to analyze and generate text responses, providing powerful capabilities for extracting insights from PDF documents.

Features

• Upload PDF: Users can upload a PDF file for analysis.
• Ask Questions: Users can ask questions related to the content of the uploaded PDF and receive answers generated by the Ollama API.
• Generate PDF: Users can instruct the application to create a new PDF document based on the analysis of the uploaded PDF content.
• Analyze Sentiment: Users can analyze the sentiment of the text extracted from the uploaded PDF, providing insights into the emotional tone of the content.
• Detect Fraud: Users can detect fraudulent activities within the PDF.
• Count Tokens: Users can count the number of tokens (words) in the text extracted from the PDF.

Installation

1. Clone the repository:

   git clone 
   cd 

2. Install the dependencies:

   pip install -r requirements.txt

3. Run the application:

   uvicorn pdf_analyzer_app.api:app --reload

API Endpoints

Upload PDF

• URL: /upload_pdf/
• Method: POST
• Description: Upload a PDF file for analysis.
• Parameters:
  • file: PDF file to upload.
• Response:

  {
    "summary": "Summary of the PDF"
  }

Ask Questions

• URL: /ask_question/
• Method: POST
• Description: Ask a question about the content of an uploaded PDF.
• Parameters:
  • question: Question to ask.
  • file: PDF file to upload.
• Response:

  {
    "answer": "Answer to the question"
  }

Generate PDF

• URL: /generate_pdf/
• Method: POST
• Description: Generate a new PDF document based on user-defined instructions.
• Parameters:
  • instruction: Instruction for generating the content.
  • file: PDF file to upload.
• Response:

  {
    "message": "PDF generated successfully",
    "filename": "generated_output.pdf"
  }

Analyze Sentiment

• URL: /analyze_sentiment/
• Method: POST
• Description: Analyze the sentiment of the text extracted from an uploaded PDF.
• Parameters:
  • file: PDF file to upload.
• Response:

  {
    "sentiment": "Result of sentiment analysis"
  }

Detect Fraud

• URL: /detect_fraud/
• Method: POST
• Description: Detect fraud in a PDF.
• Parameters:
  • file: PDF file to upload.
• Response:

  {
    "is_fraudulent": true | false
  }

Ask Questions with Different Parameter (/ask_question_2/)

• URL: /ask_question_2/
• Method: POST
• Description: Another method to ask questions about the content of a PDF.
• Parameters:
  • question: Question to ask.
  • file: PDF file to upload.
• Response:

  {
    "answer": "Answer to the question",
    "page_references": [1, 2, ...]
  }

Count Tokens

• URL: /count_tokens/
• Method: POST
• Description: Count tokens in the text extracted from the PDF.
• Parameters:
  • file: PDF file to upload.
• Response:

  {
    "tokens_per_page": [number of tokens per page],
    "total_tokens": total number of tokens
  }

Ask Questions with Parallel Processing

• URL: /ask_question_par/
• Method: POST
• Description: Ask questions about the content of a PDF using parallel processing.
• Parameters:
  • question: Question to ask.
  • file: PDF file to upload.
• Response:

  {
    "status": "Processing complete",
    "satisfaction_results": [{ "pages": [page], "result": percentage }],
    "general_result": general percentage
  }

Dev: [email protected]