import streamlit as st st.set_page_config(page_title="Status Law Assistant", page_icon="⚖️") import os import time from dotenv import load_dotenv from langchain_groq import ChatGroq from langchain_huggingface import HuggingFaceEmbeddings from langchain_community.vectorstores import FAISS from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_community.document_loaders import WebBaseLoader from langchain_core.prompts import PromptTemplate from langchain_core.output_parsers import StrOutputParser from langchain_core.runnables import RunnableLambda import requests import json from datetime import datetime from huggingface_hub import HfApi, upload_file, upload_folder, create_repo, Repository from huggingface_hub.utils import RepositoryNotFoundError import shutil # Add these to your secrets or environment variables try: HF_TOKEN = st.secrets["HF_TOKEN"] HF_USERNAME = "Rulga" DATASET_NAME = "LS_chat" DATASET_REPO = f"{HF_USERNAME}/{DATASET_NAME}" # Добавим проверку значения токена if not HF_TOKEN or HF_TOKEN.strip() == "": st.error("HF_TOKEN пустой или отсутствует в secrets") st.stop() st.write("DEBUG: HF credentials loaded successfully") except Exception as e: st.error(f"Ошибка загрузки HuggingFace credentials: {str(e)}") st.stop() # Define base directory and absolute paths BASE_DIR = os.path.dirname(os.path.abspath(__file__)) VECTOR_STORE_PATH = os.path.join(BASE_DIR, "vector_store") CHAT_HISTORY_DIR = os.path.join(BASE_DIR, "chat_history") # Create required directories with absolute paths REQUIRED_DIRS = [CHAT_HISTORY_DIR, VECTOR_STORE_PATH] for dir_path in REQUIRED_DIRS: os.makedirs(dir_path, exist_ok=True) gitkeep_path = os.path.join(dir_path, '.gitkeep') if not os.path.exists(gitkeep_path): with open(gitkeep_path, 'w') as f: pass # Knowledge base info in session_state if 'kb_info' not in st.session_state: st.session_state.kb_info = { 'build_time': None, 'size': None } # Initialize chat_history in session_state if 'chat_history' not in st.session_state: st.session_state.chat_history = [] # Initialize messages if not exists if 'messages' not in st.session_state: st.session_state.messages = [] # Create history folder if not exists #if not os.path.exists("chat_history"): # os.makedirs("chat_history") # Display title and knowledge base info # st.title("www.Status.Law Legal Assistant") st.markdown( '''

⚖️ Status.Law Legal Assistant

''', unsafe_allow_html=True ) if st.session_state.kb_info['build_time'] and st.session_state.kb_info['size']: st.caption(f"(Knowledge base build time: {st.session_state.kb_info['build_time']:.2f} seconds, " f"size: {st.session_state.kb_info['size']:.2f} MB)") # Path to store vector database # VECTOR_STORE_PATH = "vector_store" # Website URLs urls = [ "https://status.law", "https://status.law/about", "https://status.law/careers", "https://status.law/challenging-sanctions", "https://status.law/tariffs-for-services-against-extradition-en", "https://status.law/law-firm-contact-legal-protection", "https://status.law/cross-border-banking-legal-issues", "https://status.law/extradition-defense", "https://status.law/international-prosecution-protection", "https://status.law/interpol-red-notice-removal", "https://status.law/practice-areas", "https://status.law/reputation-protection", "https://status.law/faq" ] # Load secrets try: GROQ_API_KEY = st.secrets["GROQ_API_KEY"] except Exception as e: st.error("Error loading secrets. Please check your configuration.") st.stop() # Initialize models @st.cache_resource def init_models(): llm = ChatGroq( model_name="llama-3.3-70b-versatile", temperature=0.6, api_key=GROQ_API_KEY ) embeddings = HuggingFaceEmbeddings( #model_name="intfloat/multilingual-e5-large-instruct" model_name="sentence-transformers/all-MiniLM-L6-v2" ) return llm, embeddings # Build knowledge base def build_knowledge_base(embeddings): start_time = time.time() documents = [] with st.status("Loading website content...") as status: for url in urls: try: loader = WebBaseLoader(url) docs = loader.load() documents.extend(docs) status.update(label=f"Loaded {url}") except Exception as e: st.error(f"Error loading {url}: {str(e)}") text_splitter = RecursiveCharacterTextSplitter( chunk_size=500, chunk_overlap=100 ) chunks = text_splitter.split_documents(documents) vector_store = FAISS.from_documents(chunks, embeddings) # Immediately save vector store after creation force_save_vector_store(vector_store) end_time = time.time() build_time = end_time - start_time # Calculate knowledge base size total_size = 0 for path, dirs, files in os.walk(VECTOR_STORE_PATH): for f in files: fp = os.path.join(path, f) total_size += os.path.getsize(fp) size_mb = total_size / (1024 * 1024) # Save knowledge base info st.session_state.kb_info['build_time'] = build_time st.session_state.kb_info['size'] = size_mb st.success(f""" Knowledge base created successfully: - Time taken: {build_time:.2f} seconds - Size: {size_mb:.2f} MB - Number of chunks: {len(chunks)} """) return vector_store # Function to save chat history def save_chat_to_file(chat_history): """Save chat history to file using absolute path""" current_date = datetime.now().strftime("%Y-%m-%d") filename = os.path.join(CHAT_HISTORY_DIR, f"chat_history_{current_date}.json") try: with open(filename, 'w', encoding='utf-8') as f: json.dump(chat_history, f, ensure_ascii=False, indent=2) except Exception as e: st.error(f"Error saving chat history: {e}") # Function to load chat history def load_chat_history(): """Load chat history from file using absolute path""" current_date = datetime.now().strftime("%Y-%m-%d") filename = os.path.join(CHAT_HISTORY_DIR, f"chat_history_{current_date}.json") if os.path.exists(filename): try: with open(filename, 'r', encoding='utf-8') as f: return json.load(f) except Exception as e: st.error(f"Error loading chat history: {e}") return [] return [] def check_directory_permissions(directory): """Check if directory has proper read/write permissions""" try: # Check if directory exists and create if not os.makedirs(directory, exist_ok=True) # Try to create a test file test_file = os.path.join(directory, "write_test.txt") with open(test_file, "w") as f: f.write("test") f.flush() os.fsync(f.fileno()) # Force write to disk # Try to read the test file with open(test_file, "r") as f: content = f.read() if content != "test": raise Exception("File content verification failed") # Clean up os.remove(test_file) return True, None except Exception as e: permissions = oct(os.stat(directory).st_mode)[-3:] if os.path.exists(directory) else "N/A" error_msg = f"Permission error: {str(e)} (Directory permissions: {permissions})" return False, error_msg def sync_with_hf(local_path, repo_path, commit_message): """Sync local files with Hugging Face dataset""" try: st.write(f"DEBUG: Starting sync with HF for {repo_path}") api = HfApi() # Ensure the repository exists try: api.repo_info(repo_id=DATASET_REPO, repo_type="dataset") st.write("DEBUG: Repository exists") except RepositoryNotFoundError: st.write("DEBUG: Creating new repository") create_repo(DATASET_REPO, repo_type="dataset", token=HF_TOKEN) # Upload directory content st.write(f"DEBUG: Uploading folder {local_path} to {repo_path}") api.upload_folder( folder_path=local_path, path_in_repo=repo_path, repo_id=DATASET_REPO, repo_type="dataset", commit_message=commit_message, token=HF_TOKEN ) st.toast(f"✅ Synchronized with Hugging Face: {repo_path}", icon="🤗") st.write("DEBUG: Sync completed successfully") except Exception as e: error_msg = f"Failed to sync with Hugging Face: {str(e)}" st.error(error_msg) st.write(f"DEBUG: Sync error details: {str(e)}") raise Exception(error_msg) def force_save_vector_store(vector_store): """Save vector store locally and sync with HF""" try: # Local save vector_store.save_local(VECTOR_STORE_PATH) # Sync with HF sync_with_hf( local_path=VECTOR_STORE_PATH, repo_path="vector_store", commit_message=f"Update vector store: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}" ) except Exception as e: error_msg = f"Failed to save vector store: {str(e)}" st.error(error_msg) raise Exception(error_msg) def force_save_chat_history(chat_entry): """Save chat history locally and sync with HF""" try: current_date = datetime.now().strftime("%Y-%m-%d") filename = os.path.join(CHAT_HISTORY_DIR, f"chat_history_{current_date}.json") # Load existing history existing_history = [] if os.path.exists(filename): with open(filename, 'r', encoding='utf-8') as f: existing_history = json.load(f) # Add new entry existing_history.append(chat_entry) # Save locally with open(filename, 'w', encoding='utf-8') as f: json.dump(existing_history, f, ensure_ascii=False, indent=2) # Sync with HF sync_with_hf( local_path=CHAT_HISTORY_DIR, repo_path="chat_history", commit_message=f"Update chat history: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}" ) except Exception as e: error_msg = f"Failed to save chat history: {str(e)}" st.error(error_msg) raise Exception(error_msg) # Main function def main(): # Initialize models llm, embeddings = init_models() # Check if knowledge base exists if not os.path.exists(os.path.join(VECTOR_STORE_PATH, "index.faiss")): st.warning("Knowledge base not found. Please create it first.") if st.button("Create Knowledge Base"): with st.spinner("Creating knowledge base... This may take a few minutes."): try: vector_store = build_knowledge_base(embeddings) st.session_state.vector_store = vector_store st.success("Knowledge base created successfully!") st.rerun() except Exception as e: st.error(f"Error creating knowledge base: {e}") return # Load existing knowledge base if 'vector_store' not in st.session_state: try: st.session_state.vector_store = FAISS.load_local( VECTOR_STORE_PATH, embeddings, allow_dangerous_deserialization=True ) except Exception as e: st.error(f"Error loading knowledge base: {e}") return with st.sidebar: st.write(f"Working directory: {BASE_DIR}") st.write(f"Vector store: {VECTOR_STORE_PATH}") st.write(f"Chat history: {CHAT_HISTORY_DIR}") # Chat mode if 'vector_store' in st.session_state: if 'messages' not in st.session_state: st.session_state.messages = [] # Load chat history on startup if not st.session_state.chat_history: st.session_state.chat_history = load_chat_history() # Display chat history for message in st.session_state.messages: st.chat_message("user").write(message["question"]) st.chat_message("assistant").write(message["answer"]) # User input if question := st.chat_input("Ask your question"): st.chat_message("user").write(question) # Retrieve context and generate response with st.chat_message("assistant"): with st.spinner("Thinking..."): context = st.session_state.vector_store.similarity_search(question) context_text = "\n".join([doc.page_content for doc in context]) prompt = PromptTemplate.from_template(""" You are a helpful and polite legal assistant at Status Law, an international law firm specializing in extradition cases. Answer in the language in which the question was asked. Use the following information to answer questions: - Primary context: {context} - Services and pricing page: https://status.law/tariffs-for-services-against-extradition-en When asked about service prices or specific legal services: 1. Search for the specific service on our website 2. Provide a brief description of how Status Law can help with this specific issue 3. Explain the key benefits or features of this service 4. Only share the direct link to pricing (https://status.law/tariffs-for-services-against-extradition-en) if the question is specifically about prices 5. For general service inquiries without price questions, focus on service descriptions without sharing the pricing page link For example: - If asked "How much does legal representation in court cost?", describe the service briefly and provide the pricing page link - If asked "Can you help with document preparation?", explain the service without sharing the pricing link If you cannot answer based on the available information, say so politely and offer to contact Status Law directly via the following channels: - For all users: +32465594521 (landline phone) - For English and Swedish speakers only: +46728495129 (available on WhatsApp, Telegram, Signal, IMO) - Provide a link to the contact form: [Contact Form](https://status.law/law-firm-contact-legal-protection/) Question: {question} Response Guidelines: 1. Answer in the user's language 2. Be concise but informative 3. Cite specific service details when relevant 4. Emphasize our international expertise in extradition law 5. Share pricing page link ONLY when questions are specifically about costs 6. Offer contact options if the question requires detailed legal advice """) chain = prompt | llm | StrOutputParser() response = chain.invoke({ "context": context_text, "question": question }) st.write(response) # Create chat entry chat_entry = { "timestamp": datetime.now().strftime("%Y-%m-%d %H:%M:%S"), "question": question, "answer": response, "context": context_text } # Force save chat history force_save_chat_history(chat_entry) # Update session state if "chat_history" not in st.session_state: st.session_state.chat_history = [] st.session_state.chat_history.append(chat_entry) st.session_state.messages.append({ "question": question, "answer": response }) if __name__ == "__main__": main()