main.py

from fastapi import FastAPI, Form
from fastapi.middleware.cors import CORSMiddleware
import subprocess
import json
import openai
import os
import subprocess
import asyncio
from typing import List, Tuple
import json
import googleapiclient.discovery
import googleapiclient.errors
import os
from dotenv import load_dotenv
import subprocess



app = FastAPI()

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

def gpt_request(prompt, api_key):
    response = openai.Completion.create(
        engine="gpt-3.5-turbo-instruct",
        prompt=prompt,
        max_tokens=1000,
        api_key=api_key
    )
    return response.choices[0].text.strip()

prompts = [
    ("I am trying to learn {text} and I AM A BEGINEER. Now, tell me more about it and break it down into three levels for me: Beginner. Make sure your output is bulleted with three bullet points for each level and DO NOT display anything else aside of the bullet points. Keep them short and simple as well. Precede level with a # and the three bullets with a comma(,).", "System"),
    ("I am trying to learn {text} and I AM AT INTERMEDIATE EXPERTISE. Now, tell me more about it and break it down into three levels for me: Intermediate. Make sure your output is bulleted with three bullet points for each level and DO NOT display anything else aside of the bullet points. Keep them short and simple as well. Precede level with a # and the three bullets with a comma(,).", "System"),
    ("I am trying to learn {text} and I AM AT ADVANCED EXPERTISE. Now, tell me more about it and break it down into three levels for me: Advanced. Make sure your output is bulleted with three bullet points for each level and DO NOT display anything else aside of the bullet points. Keep them short and simple as well. Precede level with a # and the three bullets with a comma(,).", "System")
]

# Route to receive the single input from the frontend
def generate_replaced_prompts(text: str) -> List[Tuple[str, str]]:
    replaced_prompts = []
    for prompt, system in prompts:
        replaced_prompt = prompt.format(text=text)
        replaced_prompts.append((replaced_prompt, system))
    return replaced_prompts

@app.post("/input")
async def receive_input(text: str = Form(...)):
    try:
        # Validate or sanitize the input if necessary

        # Generate replaced prompts
        replaced_prompts = generate_replaced_prompts(text)

        # # Return the replaced prompts
        # return {"replaced_prompts": replaced_prompts}
    except Exception as e:
        return {"error": str(e)}

# @app.post("/input")
# async def receive_input(text: str = Form(...)):
    # try:
    #     # Process the input
    #     return {"message": "Received input: " + text}
    # except Exception as e:
    #     return {"error": str(e)}
# Initialize an empty array to store the outputs
    outputs = []
    # Set the OpenAI API key as an environment variable
    os.environ['OPENAI_API_KEY'] = 'YOUR_API_KEY'

    # Define the topic


    # Split the prompts equally
    # num_prompts = 9
    # prompts = [
    #     f"{topic} {i+1}" for i in range(num_prompts)
    # ]

    # Initialize an empty array to store the outputs
    outputs = []

    # Run the prompts/instances and store the outputs
    for prompt, user in replaced_prompts:
        response = openai.ChatCompletion.create(
            model="gpt-3.5-turbo",
            messages=[
                {"role": "system", "content": "You are a super helpful assistant and you need to help a student with three bullet points for each prompt. Do not give me extra fluff and text just give me topics in particular"},
                {"role": "system", "content": user},
                {"role": "assistant", "content": prompt}
            ]
        )
        output = response.choices[0].message.content
        outputs.append(output)

    # Print the outputs
    # return(outputs)

    def formatOutputs(outputs):
        formatted = []
        for element in outputs:
            formatted.append(list(element.split("\n")))
        return formatted

    def removeFluff(outputs):
        for element in outputs:
            element.pop(0)
            for item in range(len(element)):
                element[item] = element[item][2::]
        return outputs

    outputs = removeFluff(formatOutputs(outputs))
    # Create a dictionary to store the outputs
    output_dict = {
        "outputs": outputs
    }

    # Define the path to the JSON file
    json_file_path = "C:/Users/Vansh/Desktop/hackbeanpot2024/topics.json"

    # Write the dictionary to the JSON file
    with open(json_file_path, "w") as json_file:
        json.dump(output_dict, json_file)

    # Set up the YouTube Data API client
    # await asyncio.sleep(3)

    # Call VideoSearch.py to search for the video IDs
    subprocess.run(["python", "VideoSearch.py"])
    # Run the previous processes here

    # Wait for all processes to complete
    # await asyncio.sleep(2.4)

    # Continue with the instructions below

    # Call VideoPlaylist.py to create a new playlist
    subprocess.run(["python", "VideoPlaylist.py"])

    # Wait for all processes to complete


    # Return the GPT-3.5 response
    

    # gpt_response2 = gpt_request(custom_prompt2, api_key)
    # level = []
    # bullets = []

    # schedule_text2 = gpt_response
    
    # lines = schedule_text2.split('\n')

    # for line in lines:
    #     line = line.strip()
    #     if line.startswith('#'):
    #         level.append(line)
    #     elif line.startswith('*'):
    #         bullets.append(line)

    # for i in range(len(level)):
    #     title2 = level[i].lstrip('#').strip()
    #     task2 = bullets[i].lstrip('*').strip()

    # # Return the GPT-3.5 response
    # gpt_response3 = gpt_request(custom_prompt3, api_key)
    # level = []
    # bullets = []

    # schedule_text3 = gpt_response3
    
    # lines = schedule_text3.split('\n')

    # for line in lines:
    #     line = line.strip()
    #     if line.startswith('#'):
    #         level.append(line)
    #     elif line.startswith('*'):
    #         bullets.append(line)

    # for i in range(len(level)):
    #     title3 = level[i].lstrip('#').strip()
    #     task3 = bullets[i].lstrip('*').strip()

    # return {title1, task1, title2, task2, title3, task3}


# if __name__ == "__main__":
#     import uvicorn
#     uvicorn.run(app, host="0.0.0.0", port=8000)