fix/some bugs

app.py

@@ -1,5 +1,3 @@
-# app.py
-
 import os
 import time
 import uuid
@@ -11,7 +9,6 @@ import fitz
 import re
 import io
 from gtts import gTTS
-
 from langchain_core.documents import Document
 from langchain_community.document_loaders import (
     TextLoader,
@@ -26,12 +23,8 @@ from langchain_community.retrievers import BM25Retriever
 from langchain_community.chat_message_histories import ChatMessageHistory
 from langchain.storage import InMemoryStore
 from sentence_transformers.cross_encoder import CrossEncoder
-
-
 app = Flask(__name__)
 app.config['SECRET_KEY'] = os.urandom(24)
-
-
 class LocalReranker(BaseDocumentCompressor):
     model: Any
     top_n: int = 3
@@ -59,7 +52,7 @@ class LocalReranker(BaseDocumentCompressor):
             doc.metadata['rerank_score'] = float(score)
             top_docs.append(doc)
         return top_docs
-
+    
 is_hf_spaces = bool(os.getenv("SPACE_ID") or os.getenv("SPACES_ZERO_GPU"))
 if is_hf_spaces:
     app.config['UPLOAD_FOLDER'] = '/tmp/uploads'
@@ -68,12 +61,12 @@ else:
 
 try:
     os.makedirs(app.config['UPLOAD_FOLDER'], exist_ok=True)
-    print(f"✓ Upload folder ready: {app.config['UPLOAD_FOLDER']}")
+    print(f"Upload folder ready: {app.config['UPLOAD_FOLDER']}")
 except Exception as e:
-    print(f"✗ Failed to create upload folder {app.config['UPLOAD_FOLDER']}: {e}")
+    print(f"Failed to create upload folder {app.config['UPLOAD_FOLDER']}: {e}")
     app.config['UPLOAD_FOLDER'] = '/tmp/uploads'
     os.makedirs(app.config['UPLOAD_FOLDER'], exist_ok=True)
-    print(f"✓ Using fallback upload folder: {app.config['UPLOAD_FOLDER']}")
+    print(f"Using fallback upload folder: {app.config['UPLOAD_FOLDER']}")
 
 rag_chains = {}
 message_histories = {}
@@ -84,17 +77,17 @@ try:
         model_name="sentence-transformers/all-MiniLM-L6-v2",
         model_kwargs={'device': 'cpu'}
     )
-    print("✓ Embedding model loaded successfully.")
+    print("Embedding model loaded successfully.")
 except Exception as e:
-    print(f"✗ FATAL: Could not load embedding model. Error: {e}")
+    print(f"FATAL: Could not load embedding model. Error: {e}")
     raise e
 
 print("Loading local re-ranking model...")
 try:
     RERANKER_MODEL = CrossEncoder("cross-encoder/ms-marco-MiniLM-L-6-v2", device='cpu')
-    print("✓ Re-ranking model loaded successfully.")
+    print("Re-ranking model loaded successfully.")
 except Exception as e:
-    print(f"✗ FATAL: Could not load reranker model. Error: {e}")
+    print(f"FATAL: Could not load reranker model. Error: {e}")
     raise e
 
 def load_pdf_with_fallback(filepath):
@@ -112,12 +105,12 @@ def load_pdf_with_fallback(filepath):
                         }
                     ))
         if docs:
-            print(f"✓ Successfully loaded PDF with PyMuPDF: {filepath}")
+            print(f"Successfully loaded PDF with PyMuPDF: {filepath}")
             return docs
         else:
             raise ValueError("No text content found in PDF.")
     except Exception as e:
-        print(f"✗ PyMuPDF failed for {filepath}: {e}")
+        print(f"PyMuPDF failed for {filepath}: {e}")
         raise
 
 LOADER_MAPPING = {
@@ -200,7 +193,7 @@ def upload_files():
         
         store.mset(list(zip(doc_ids, parent_docs)))
         vectorstore.add_documents(child_docs)
-        print(f"✓ Stored {len(parent_docs)} parent docs and indexed {len(child_docs)} child docs.")
+        print(f"Stored {len(parent_docs)} parent docs and indexed {len(child_docs)} child docs.")
 
         bm25_retriever = BM25Retriever.from_documents(child_docs)
         bm25_retriever.k = 8
@@ -211,7 +204,7 @@ def upload_files():
             retrievers=[bm25_retriever, faiss_retriever],
             weights=[0.4, 0.6]
         )
-        print("✓ Created Hybrid Retriever for child documents.")
+        print("Created Hybrid Retriever for child documents.")
 
         reranker = LocalReranker(model=RERANKER_MODEL, top_n=4)
         
@@ -225,7 +218,7 @@ def upload_files():
         
         final_retriever = compression_retriever | get_parents
         
-        print("✓ Final retriever chain created: (Hybrid -> Rerank) -> Parent Fetch")
+        print("Final retriever chain created: (Hybrid -> Rerank) -> Parent Fetch")
 
         session_id = str(uuid.uuid4())
         rag_chain = create_rag_chain(final_retriever, get_session_history)
@@ -259,12 +252,7 @@ def chat():
     try:
         rag_chain = rag_chains[session_id]
         config = {"configurable": {"session_id": session_id}}
-        
-        # Invoke the chain, which will return a string
         answer_string = rag_chain.invoke({"question": question}, config=config)
-        
-        # --- THIS IS THE FIX ---
-        # Directly use the returned string in the JSON response.
         return jsonify({'answer': answer_string})
 
     except Exception as e:

rag_processor.py

@@ -1,16 +1,8 @@
-# rag_processor.py
-
 import os
 from dotenv import load_dotenv
-from operator import itemgetter # <--- ADD THIS IMPORT
-
-# LLM
+from operator import itemgetter
 from langchain_groq import ChatGroq
-
-# Prompting
 from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
-
-# Chains
 from langchain_core.runnables import RunnableParallel, RunnablePassthrough
 from langchain_core.output_parsers import StrOutputParser
 from langchain_core.runnables.history import RunnableWithMessageHistory
@@ -30,18 +22,12 @@ def create_rag_chain(retriever, get_session_history_func):
     Raises:
         ValueError: If the GROQ_API_KEY is missing.
     """
-    # Load environment variables from .env file (for local development)
     load_dotenv()
-    
-    # Get API key from environment (works for both HF Spaces secrets and local .env)
     api_key = os.getenv("GROQ_API_KEY")
     print("key loaded")
-    
     if not api_key or api_key == "your_groq_api_key_here":
         error_msg = "GROQ_API_KEY not found or not configured properly.\n"
         print("Not found key")
-        
-        # Detect if running in Hugging Face Spaces
         if os.getenv("SPACE_ID") or os.getenv("SPACES_ZERO_GPU"):
             error_msg += (
                 "For Hugging Face Spaces: Set GROQ_API_KEY in your Space's Settings > Repository Secrets.\n"
@@ -55,12 +41,8 @@ def create_rag_chain(retriever, get_session_history_func):
             )
         
         raise ValueError(error_msg)
-
-    # --- 1. Initialize the LLM ---
-    # Updated model_name to a standard, high-performance Groq model
     llm = ChatGroq(model_name="llama-3.1-8b-instant", api_key=api_key, temperature=0.1)
 
-    # --- 2. Create Query Rewriting Chain 🧠 ---
     print("\nSetting up query rewriting chain...")
     rewrite_template = """You are an expert at rewriting user questions for a vector database.
 You are here to help the user with their document.
@@ -79,8 +61,6 @@ Standalone Question:"""
         ("human", "Based on our conversation, reformulate this question to be a standalone query: {question}")
     ])
     query_rewriter = rewrite_prompt | llm | StrOutputParser()
-    
-    # --- 3. Create Main RAG Chain with Memory ---
     print("\nSetting up main RAG chain...")
     rag_template = """You are CogniChat, an expert document analysis assistant.
 IMPORTANT RULES:
@@ -95,35 +75,24 @@ Context:
         MessagesPlaceholder(variable_name="chat_history"),
         ("human", "{question}"),
     ])
-
-    # ============================ FIX IS HERE ============================
-    # Parallel process to fetch context and correctly pass through question and history.
-    # We use itemgetter to select the specific keys from the input dictionary.
     setup_and_retrieval = RunnableParallel({
     "context": RunnablePassthrough.assign(
         rewritten_question=query_rewriter
     ) | (lambda x: x["rewritten_question"]) | retriever,
     "question": itemgetter("question"),
     "chat_history": itemgetter("chat_history")})
-    # =====================================================================
-    
-    # The initial RAG chain
     conversational_rag_chain = (
         setup_and_retrieval
         | rag_prompt
         | llm
         | StrOutputParser()
     )
-
-    # Wrap the chain with memory management
     chain_with_memory = RunnableWithMessageHistory(
         conversational_rag_chain,
         get_session_history_func,
         input_messages_key="question",
         history_messages_key="chat_history",
     )
-
-    # --- 4. Create Answer Refinement Chain ✨ ---
     print("\nSetting up answer refinement chain...")
     refine_template = """You are an expert at editing and refining content.
 Your task is to take a given answer and improve its clarity, structure, and readability.
@@ -137,9 +106,6 @@ Refined Answer:"""
     refine_prompt = ChatPromptTemplate.from_template(refine_template)
     refinement_chain = refine_prompt | llm | StrOutputParser()
 
-    # --- 5. Combine Everything into the Final Chain ---
-    # The final chain passes the output of the memory-enabled chain to the refinement chain
-    # Note: We need to adapt the input for the refinement chain
     final_chain = (
         lambda input_dict: {"answer": chain_with_memory.invoke(input_dict, config=input_dict.get('config'))}
     ) | refinement_chain

test_dependencies.py

@@ -1,11 +1,7 @@
-#!/usr/bin/env python3
-"""
-Test script to verify all dependencies and PDF handling capabilities.
-"""
 import os
 import sys
 
-print("=== CogniChat Dependencies & PDF Handling Test ===")
+print("CogniChat Dependencies & PDF Handling Test")
 
 # Test imports
 try:
@@ -14,62 +10,50 @@ try:
     from langchain_community.retrievers import BM25Retriever
     from langchain.text_splitter import RecursiveCharacterTextSplitter
     from langchain_core.documents import Document
-    print("✓ Core LangChain imports successful!")
+    print("Core LangChain imports successful!")
     
 except ImportError as e:
-    print(f"✗ Import error: {e}")
+    print(f"Import error: {e}")
     if "rank_bm25" in str(e):
-        print("  → Missing dependency: pip install rank-bm25==0.2.2")
+        print("Missing dependency: pip install rank-bm25==0.2.2")
     sys.exit(1)
-
-# Test PDF loading capabilities
 try:
     print("\nTesting PDF loading capabilities...")
-    
-    # Test PyPDF availability
     try:
         from langchain_community.document_loaders import PyPDFLoader
-        print("✓ PyPDFLoader available")
+        print("PyPDFLoader available")
     except ImportError:
-        print("✗ PyPDFLoader not available")
-    
-    # Test PyMuPDF availability
+        print("PyPDFLoader not available")
+
     try:
         import fitz
-        print("✓ PyMuPDF (fitz) available - can handle corrupted PDFs")
+        print("PyMuPDF (fitz) available - can handle corrupted PDFs")
     except ImportError:
-        print("✗ PyMuPDF (fitz) not available")
+        print("PyMuPDF (fitz) not available")
     
-    # Test pdfplumber availability
     try:
         import pdfplumber
-        print("✓ pdfplumber available - additional PDF parsing method")
+        print("pdfplumber available - additional PDF parsing method")
     except ImportError:
-        print("✗ pdfplumber not available")
+        print("pdfplumber not available")
         
 except Exception as e:
-    print(f"✗ Error testing PDF capabilities: {e}")
-
-# Test BM25 Retriever
+    print(f"Error testing PDF capabilities: {e}")
 try:
     print("\nTesting BM25 Retriever...")
     
-    # Create some test documents
     test_docs = [
         Document(page_content="This is the first test document about machine learning."),
         Document(page_content="This is the second document discussing natural language processing."),
         Document(page_content="The third document covers artificial intelligence topics."),
     ]
     
-    # Create BM25 retriever
     bm25_retriever = BM25Retriever.from_documents(test_docs)
     bm25_retriever.k = 2
-    
-    # Test retrieval
     query = "machine learning"
     results = bm25_retriever.get_relevant_documents(query)
-    print(f"✓ BM25 retriever created and tested successfully!")
-    print(f"✓ Retrieved {len(results)} documents for query: '{query}'")
+    print(f"BM25 retriever created and tested successfully!")
+    print(f"Retrieved {len(results)} documents for query: '{query}'")
     
 except Exception as e:
     print(f"✗ Error testing BM25 retriever: {e}")
@@ -77,7 +61,7 @@ except Exception as e:
     traceback.print_exc()
     sys.exit(1)
 
-print("\n=== All tests completed successfully! ===")
+print("\nAll tests completed successfully!")
 print("\nThe application should now handle:")
 print("  • Regular file uploads and processing")
 print("  • Corrupted PDF files with multiple fallback methods")
@@ -86,8 +70,7 @@ print("  • Proper error messages for failed file processing")
 print("\nMake sure to install all dependencies with:")
 print("  pip install -r requirements.txt")
 
-# Display dependency summary
-print("\n=== Key Dependencies Added/Updated ===")
+print("\nKey Dependencies Added/Updated")
 print("  • rank-bm25==0.2.2 (for BM25 retrieval)")
 print("  • pymupdf==1.23.26 (PDF fallback method)")
 print("  • pdfplumber==0.10.3 (additional PDF parsing)")

test_hf_spaces_session.py

@@ -1,18 +1,11 @@
-#!/usr/bin/env python3
-"""
-Quick test to diagnose the current HF Spaces chat issue.
-Run this after uploading a document to test the session state.
-"""
 import requests
 import json
-
-# Replace with your actual Space URL
-BASE_URL = "https://hyperxd-0-cognichat.hf.space"  # Update this to your actual Space URL
+BASE_URL = "https://huggingface.co/spaces/Zeri00/Cogni-chat-document-reader"
 
 def test_endpoints():
     """Test the debug and session endpoints to understand the issue."""
     
-    print("=== CogniChat HF Spaces Diagnostic ===\n")
+    print("CogniChat HF Spaces Diagnostic\n")
     
     # Test 1: Check debug endpoint
     print("1. Testing /debug endpoint...")
@@ -20,7 +13,7 @@ def test_endpoints():
         response = requests.get(f"{BASE_URL}/debug")
         if response.status_code == 200:
             data = response.json()
-            print("✓ Debug endpoint working")
+            print(" Debug endpoint working")
             print(f"   Environment: {data.get('environment')}")
             print(f"   GROQ API Key: {'Set' if data.get('groq_api_key_set') else 'NOT SET'}")
             print(f"   Sessions count: {data.get('sessions_count')}")
@@ -29,44 +22,39 @@ def test_endpoints():
             print(f"   Flask session ID: {data.get('flask_session_id')}")
             print(f"   Session keys: {data.get('flask_session_keys')}")
         else:
-            print(f"✗ Debug endpoint failed: {response.status_code}")
+            print(f"Debug endpoint failed: {response.status_code}")
     except Exception as e:
-        print(f"✗ Error accessing debug endpoint: {e}")
+        print(f"Error accessing debug endpoint: {e}")
     
     print()
     
-    # Test 2: Check session handling
     print("2. Testing /test-session endpoint...")
     try:
-        # Create a session with cookies
         session = requests.Session()
-        
-        # Test session write
         response = session.post(f"{BASE_URL}/test-session")
         if response.status_code == 200:
             data = response.json()
-            print("✓ Session write working")
-            print(f"   Test key: {data.get('test_key')}")
-            print(f"   Session keys: {data.get('session_keys')}")
+            print("Session write working")
+            print(f"Test key: {data.get('test_key')}")
+            print(f"Session keys: {data.get('session_keys')}")
         else:
-            print(f"✗ Session write failed: {response.status_code}")
+            print(f"Session write failed: {response.status_code}")
         
-        # Test session read
         response = session.get(f"{BASE_URL}/test-session")
         if response.status_code == 200:
             data = response.json()
-            print("✓ Session read working")
+            print("Session read working")
             print(f"   Test key persisted: {data.get('test_key')}")
             print(f"   Has session data: {data.get('has_session_data')}")
             
             if not data.get('test_key'):
-                print("⚠️  WARNING: Sessions are not persisting between requests!")
-                print("   This is likely the cause of the 400 chat error.")
+                print("WARNING: Sessions are not persisting between requests!")
+                print(" This is likely the cause of the 400 chat error.")
         else:
-            print(f"✗ Session read failed: {response.status_code}")
+            print(f"Session read failed: {response.status_code}")
             
     except Exception as e:
-        print(f"✗ Error testing sessions: {e}")
+        print(f"Error testing sessions: {e}")
     
     print()
     
@@ -78,16 +66,16 @@ def test_endpoints():
             data = response.json()
             session_ids = data.get('session_ids', [])
             if session_ids:
-                print(f"✓ Found {len(session_ids)} existing RAG sessions")
-                print(f"   Session IDs: {session_ids[:3]}{'...' if len(session_ids) > 3 else ''}")
+                print(f"Found {len(session_ids)} existing RAG sessions")
+                print(f" Session IDs: {session_ids[:3]}{'...' if len(session_ids) > 3 else ''}")
             else:
-                print("ℹ️  No RAG sessions found (normal if no documents were uploaded)")
+                print("No RAG sessions found (normal if no documents were uploaded)")
         
     except Exception as e:
-        print(f"✗ Error checking RAG sessions: {e}")
+        print(f"Error checking RAG sessions: {e}")
     
     print()
-    print("=== Diagnosis Complete ===")
+    print("Diagnosis Complete")
     print()
     print("LIKELY ISSUE:")
     print("If sessions are not persisting, this is a common issue in HF Spaces")
@@ -104,7 +92,4 @@ if __name__ == "__main__":
     print("3. Optionally upload a document first")
     print()
     
-    # Uncomment the next line and update the URL to run the test
-    # test_endpoints()
-    
     print("Update the BASE_URL variable above and uncomment the test_endpoints() call")

test_upload_permissions.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python3
+
 """
 Test script to verify upload folder permissions and file operations.
 """
@@ -6,7 +6,7 @@ import os
 import tempfile
 from pathlib import Path
 
-print("=== Upload Folder Permission Test ===")
+print("Upload Folder Permission Test")
 
 # Detect environment
 is_hf_spaces = bool(os.getenv("SPACE_ID") or os.getenv("SPACES_ZERO_GPU"))
@@ -20,7 +20,7 @@ test_folders = [
     '/tmp/cognichat_uploads'  # Alternative temp location
 ]
 
-print("\n=== Testing Upload Folder Options ===")
+print("\nTesting Upload Folder Options")
 
 for folder in test_folders:
     print(f"\nTesting: {folder}")
@@ -28,35 +28,35 @@ for folder in test_folders:
     try:
         # Try to create the directory
         os.makedirs(folder, exist_ok=True)
-        print(f"  ✓ Directory created/exists")
+        print(f"Directory created/exists")
         
         # Test write permissions
         test_file = os.path.join(folder, 'test_write.txt')
         with open(test_file, 'w') as f:
             f.write('test content')
-        print(f"  ✓ Write permission verified")
+        print(f"Write permission verified")
         
         # Test read permissions  
         with open(test_file, 'r') as f:
             content = f.read()
-        print(f"  ✓ Read permission verified")
+        print(f"Read permission verified")
         
         # Clean up test file
         os.remove(test_file)
-        print(f"  ✓ File deletion works")
+        print(f" File deletion works")
         
         # Get absolute path
         abs_path = os.path.abspath(folder)
-        print(f"  → Full path: {abs_path}")
-        print(f"  → Writable: {os.access(folder, os.W_OK)}")
+        print(f"Full path: {abs_path}")
+        print(f"Writable: {os.access(folder, os.W_OK)}")
         
     except PermissionError as e:
-        print(f"  ✗ Permission denied: {e}")
+        print(f"Permission denied: {e}")
     except Exception as e:
-        print(f"  ✗ Error: {e}")
+        print(f"Error: {e}")
 
 # Recommended configuration
-print(f"\n=== Recommended Configuration ===")
+print(f"\nRecommended Configuration")
 if is_hf_spaces:
     recommended_folder = '/tmp/uploads'
     print(f"For Hugging Face Spaces: {recommended_folder}")
@@ -68,17 +68,15 @@ print(f"\nUse this in your Flask app:")
 print(f"app.config['UPLOAD_FOLDER'] = '{recommended_folder}'")
 
 # Test the current working directory permissions
-print(f"\n=== Current Directory Info ===")
+print(f"\nCurrent Directory Info")
 cwd = os.getcwd()
 print(f"Current working directory: {cwd}")
 print(f"CWD is writable: {os.access(cwd, os.W_OK)}")
-
-# Show environment variables related to paths
-print(f"\n=== Path Environment Variables ===")
+print(f"\nPath Environment Variables")
 path_vars = ['HOME', 'TMPDIR', 'TEMP', 'TMP', 'SPACE_ID', 'SPACES_ZERO_GPU']
 for var in path_vars:
     value = os.getenv(var)
     if value:
         print(f"{var}: {value}")
 
-print(f"\n=== Test Complete ===")
+print(f"\nTest Complete")

verify_hf_spaces_ready.py

@@ -16,7 +16,7 @@ def print_header(text):
 
 def print_check(condition, message):
     """Print a check result."""
-    status = "✅ PASS" if condition else "❌ FAIL"
+    status = "PASS" if condition else "FAIL"
     print(f"{status}: {message}")
     return condition
 
@@ -247,8 +247,8 @@ def main():
         print("5. Wait for build to complete")
         return 0
     else:
-        print(f"\n❌ SOME CHECKS FAILED ({total - passed}/{total} issues)")
-        print("\n⚠️  Please fix the issues above before deploying.")
+        print(f"\n SOME CHECKS FAILED ({total - passed}/{total} issues)")
+        print("\n Please fix the issues above before deploying.")
         print("\nFor detailed guidance, see:")
         print("- HF_SPACES_FILE_STORAGE_GUIDE.md")
         print("- DEPLOYMENT.md")