File size: 4,948 Bytes
f9d300b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 |
import streamlit as st
from base64 import b64encode
import os
from dotenv import load_dotenv
from io import BytesIO
from PyPDF2 import PdfReader
from langchain_community.document_loaders import PyPDFLoader, UnstructuredPDFLoader, OnlinePDFLoader
from tempfile import NamedTemporaryFile
import google.generativeai as genai
load_dotenv() ## load all the environemnt variables
## Configure Genai Key
genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))
def prepare_prompt(question, context):
prompt= f"""
You are an expert in analyzing the context and providing accurate and comprehensive answers based on the context.
Use the context prvided below and answer comprehensively to the question at the end.
Context: {context}
Question:{question}
"""
return prompt
def get_gemini_response(prompt):
print(prompt)
model=genai.GenerativeModel('gemini-pro')
response=model.generate_content(prompt)
return response.text
def extract_text(uploaded_file):
"""Extracts text from each page of a PDF using fitz.
Args:
pdf_bytes (bytes): The PDF content in bytes format.
Returns:
list: A list containing the extracted text from each page.
"""
pages = []
if isinstance(uploaded_file, str): # Handle URL case
loader = OnlinePDFLoader(uploaded_file)
print("Fetching Url")
else:
pdf_reader = PdfReader(uploaded_file) # Handle uploaded file case
pages = []
for page in pdf_reader.pages:
pages.append(page.extract_text())
print(pages)
return pages
def display_pdf(pdf_data,col):
"""Displays the PDF using base64 encoding and an iframe."""
if pdf_data is not None:
base64_pdf = b64encode(pdf_data).decode('utf-8')
pdf_display = F'<iframe src="data:application/pdf;base64,{base64_pdf}" width="100%" height="600px" type="application/pdf"></iframe>'
with col:
st.markdown(pdf_display, unsafe_allow_html=True)
def main():
"""Streamlit app with PDF viewer and chat window in a two-row layout."""
st.set_page_config(page_title="PDF Viewer with Chat", page_icon="")
st.title("Upload or Enter a URL to View a PDF and Chat")
extracted_text=""
# Full page width layout
st.markdown("""
<style>
body {{ margin: 0; padding: 0; }}
</style>
""", unsafe_allow_html=True)
# Row 1 for file upload and URL
col1_1, col1_2 = st.columns(2)
with col1_1:
uploaded_file = st.file_uploader("Choose a PDF file", type="pdf")
with col1_2:
pdf_url = st.text_input("Enter a PDF URL (optional)")
# Row 2 for PDF viewer and chat
col2_1, col2_2 = st.columns([1, 1]) # 66.6% for PDF viewer, 33.3% for chat
# Chat window (right column, hidden initially)
with col2_2:
chat_container = st.empty() # Placeholder for chat area
chat_history = st.empty() # Placeholder for chat history (optional)
def show_chat(extracted_text):
chat_container.empty() # Clear previous chat elements
with chat_container:
# User input for chat
with col2_2:
user_input = st.text_input("Type your query...", key="chat_input")
# Send button (optional)
# You can add a button here to trigger actions based on user input
# send_button = st.button("Send")
# Update chat history (optional)
if user_input:
# Here, you can process the user input and update the chat history
# You can use a DataFrame (like `chat_history`) to store and display messages
# This example just shows a placeholder for future implementation
prompt=prepare_prompt(user_input,extracted_text)
res=get_gemini_response(prompt)
with col2_2:
chat_history.text_area(f"""User: {user_input}\nAnswer: {res}""")
user_input.text_input("")
if uploaded_file is not None:
pdf_data = uploaded_file.getvalue()
#display_pdf(pdf_data,col2_1)
extracted_text = extract_text(uploaded_file)
show_chat(extracted_text) # Show chat window only after upload
#chat_history.text("User: " + user_input)
elif pdf_url:
try:
import requests
response = requests.get(pdf_url, stream=True)
if response.status_code == 200:
pdf_data = response.content
#display_pdf(pdf_data,col2_1)
extracted_text = extract_text(pdf_url)
show_chat(extracted_text) # Show chat window only after successful URL fetch
else:
st.error(f"Error: Failed to fetch PDF from URL. Status code: {response.status_code}")
except Exception as e:
st.error(f"Error: An error occurred while fetching PDF from URL: {e}")
if __name__ == '__main__':
main()
|