Upload 2 files

law.py

@@ -0,0 +1,171 @@
+import streamlit as st
+import os
+import pickle
+import time
+import g4f
+import tempfile
+import PyPDF2
+from pdf2image import convert_from_path
+import pytesseract
+
+st.set_page_config(page_title="LEGAL ASSISTANT")
+
+st.markdown(
+    """
+    <style>
+        .title {
+            text-align: center;
+            font-size: 2em;
+            font-weight: bold;
+        }
+    </style>
+    <div class="title"> ⚖️ LEGAL ASSISTANT ⚖️</div>
+    """,
+    unsafe_allow_html=True
+)
+# Load and Save Conversations
+conversations_file = "conversations.pkl"
+
+
+@st.cache_data
+def load_conversations():
+    try:
+        with open(conversations_file, "rb") as f:
+            return pickle.load(f)
+    except (FileNotFoundError, EOFError):
+        return []
+
+
+def save_conversations(conversations):
+    temp_conversations_file = conversations_file
+    with open(temp_conversations_file, "wb") as f:
+        pickle.dump(conversations, f)
+    os.replace(temp_conversations_file, conversations_file)
+
+
+if 'conversations' not in st.session_state:
+    st.session_state.conversations = load_conversations()
+
+if 'current_conversation' not in st.session_state:
+    st.session_state.current_conversation = [{"role": "assistant", "content": "How may I assist you today?"}]
+
+
+def truncate_string(s, length=30):
+    return s[:length].rstrip() + "..." if len(s) > length else s
+
+
+def display_chats_sidebar():
+    with st.sidebar.container():
+        st.header('Settings')
+        col1, col2 = st.columns([1, 1])
+
+        with col1:
+            if col1.button('Start New Chat', key="new_chat"):
+                st.session_state.current_conversation = []
+                st.session_state.conversations.append(st.session_state.current_conversation)
+
+        with col2:
+            if col2.button('Clear All Chats', key="clear_all"):
+                st.session_state.conversations = []
+                st.session_state.current_conversation = []
+
+    if st.sidebar.button('Summarize Agreements', key="summarize_bills", use_container_width=True):
+        st.session_state.page = "summarize_bills"
+
+    with st.sidebar.container():
+        st.header('Conversations')
+        for idx, conversation in enumerate(st.session_state.conversations):
+            if conversation:
+                chat_title_raw = next((msg["content"] for msg in conversation if msg["role"] == "user"), "New Chat")
+                chat_title = truncate_string(chat_title_raw)
+                if st.sidebar.button(f"{chat_title}", key=f"chat_button_{idx}"):
+                    st.session_state.current_conversation = st.session_state.conversations[idx]
+
+
+def summarize_bill():
+    st.header("📜 Summarize Agreements 📜")
+
+    if st.button("Back to Chat"):
+        st.session_state.page = "chat"
+
+    uploaded_file = st.file_uploader("Upload an Agreement", type=['pdf'])
+    if uploaded_file is not None:
+        with tempfile.NamedTemporaryFile(delete=False) as tmp_file:
+            tmp_file.write(uploaded_file.read())
+            extracted_text = extract_text_from_pdf(tmp_file.name)
+
+        if st.button('Summarize'):
+            # Assuming g4f.ChatCompletion can be used for summarization
+            # Replace with appropriate summarization logic if needed
+            summary = g4f.ChatCompletion.create(
+                model="gpt-3.5-turbo",
+                messages=[{"role": "user", "content": "Please Summarize this Agreement: \n" +extracted_text}],
+                temperature=0.5,  # You can adjust parameters as needed
+                max_tokens=150  # Adjust the token limit as needed
+            )
+            st.text_area("Summary", summary, height=400)
+
+
+def extract_text_from_pdf(file_path: str) -> str:
+    try:
+        with open(file_path, 'rb') as file:
+            reader = PyPDF2.PdfReader(file)
+            text = ''
+            for page_number in range(len(reader.pages)):
+                page = reader.pages[page_number]
+                text += page.extract_text()
+        return text
+    except Exception as e:
+        try:
+            images = convert_from_path(file_path)
+            extracted_texts = [pytesseract.image_to_string(image) for image in images]
+            return "\n".join(extracted_texts)
+        except Exception as e:
+            raise ValueError(f"Failed to process {file_path} using PDF Reader and OCR. Error: {e}")
+
+
+def main_app():
+    for message in st.session_state.current_conversation:
+        with st.chat_message(message["role"]):
+            st.write(message["content"])
+
+    def generate_response(prompt_input):
+        string_dialogue = "You are a helpful Medical Assistant. You will Only respond to Medical related Queries. Say Sorry to any other Type of Queries."
+        for dict_message in st.session_state.current_conversation:
+            string_dialogue += dict_message["role"].capitalize() + ": " + dict_message["content"] + "\\n\\n"
+
+        prompt = f"{string_dialogue}\n  {prompt_input} Assistant: "
+        response_generator = g4f.ChatCompletion.create(
+            model="gpt-3.5-turbo",
+            messages=[{"role": "user", "content": prompt}],
+            stream=True,
+        )
+        return response_generator
+
+    if prompt := st.chat_input('Send a Message'):
+        st.session_state.current_conversation.append({"role": "user", "content": prompt})
+        with st.chat_message("user"):
+            st.write(prompt)
+
+        with st.chat_message("assistant"):
+            with st.spinner("Thinking..."):
+                response = generate_response(prompt)
+                placeholder = st.empty()
+                full_response = ''
+                for item in response:
+                    full_response += item
+                    time.sleep(0.003)
+                    placeholder.markdown(full_response)
+                placeholder.markdown(full_response)
+                st.session_state.current_conversation.append({"role": "assistant", "content": full_response})
+                save_conversations(st.session_state.conversations)
+
+
+display_chats_sidebar()
+if st.session_state.get('page') == "summarize_bills":
+    summarize_bill()
+elif st.session_state.get('page') == "chat":
+    main_app()
+else:
+    # Default page when the app starts or when the state is not set
+    main_app()

requirements.txt

@@ -0,0 +1,7 @@
+gpt4all==1.0.8
+streamlit
+g4f
+docx2txt
+pypdf2
+pytesseract
+pdf2image