update - v3

app.py

@@ -1,27 +1,27 @@
-from chromadb.config import Settingsimport os
+import os
 import shutil
 import streamlit as st
 import chromadb
-from chromadb.config import Settings
+from chromadb.config import Settings, DEFAULT_TENANT, DEFAULT_DATABASE
 from sentence_transformers import SentenceTransformer
 from transformers import pipeline
 from weather import get_weather_summary
 from travel import get_travel_spots
 import spacy
 
-# Force Chroma to use a valid persist directory
-os.environ["PERSIST_DIRECTORY"] = "/tmp/chroma"
-
-# Clean up any stale DB state (optional but avoids schema errors)
-if os.path.exists("/tmp/chroma"):
-    shutil.rmtree("/tmp/chroma")
-
-# Initialize Chroma with valid settings
-settings = Settings(
-    chroma_db_impl="duckdb+parquet",
-    persist_directory="/tmp/chroma"
+# Ensure a fresh Chroma DB directory (temporary path)
+CHROMA_PATH = "/tmp/chroma"
+if os.path.exists(CHROMA_PATH):
+    shutil.rmtree(CHROMA_PATH)
+os.makedirs(CHROMA_PATH, exist_ok=True)
+
+# Initialize Chroma using the new client API
+client = chromadb.PersistentClient(
+    path=CHROMA_PATH,
+    settings=Settings(),
+    tenant=DEFAULT_TENANT,
+    database=DEFAULT_DATABASE,
 )
-client = chromadb.Client(settings)
 db = client.get_or_create_collection("disaster_news")
 
 # Load models
@@ -29,11 +29,11 @@ embed_model = SentenceTransformer("all-MiniLM-L6-v2")
 qa_pipeline = pipeline("text2text-generation", model="google/flan-t5-base")
 nlp = spacy.load("en_core_web_sm")
 
-# Define keyword lists
+# Keyword lists
 weather_keywords = ["weather", "forecast", "rain", "snow", "temperature", "wind", "climate", "humid", "cold", "hot"]
 travel_keywords = ["visit", "travel", "tourist", "see", "go", "spots", "places", "explore", "attractions"]
 
-# Extract location and user intent
+# Extract location and intent
 def extract_location_and_intent(query):
     doc = nlp(query)
     locations = [ent.text for ent in doc.ents if ent.label_ in ("GPE", "LOC")]
@@ -62,16 +62,13 @@ if st.button("Submit") and user_input:
     response_parts = []
 
     if is_weather and location:
-        weather_summary = get_weather_summary(location)
-        response_parts.append(weather_summary)
+        response_parts.append(get_weather_summary(location))
 
     if is_travel and location:
-        travel_suggestions = get_travel_spots(location)
-        response_parts.append(travel_suggestions)
+        response_parts.append(get_travel_spots(location))
 
     if not is_weather and not is_travel:
-        rag_response = query_rag_system(user_input)
-        response_parts.append(rag_response)
+        response_parts.append(query_rag_system(user_input))
 
     if not response_parts:
         response_parts.append("❓ Couldn't understand your question or find relevant data.")
@@ -79,11 +76,3 @@ if st.button("Submit") and user_input:
     st.markdown("### 🔎 Response:")
     for part in response_parts:
         st.write(part)
-
-
-settings = Settings(
-    chroma_db_impl="duckdb+parquet",
-    persist_directory="/tmp/chroma"  # must not be None
-)
-
-client = chromadb.Client(settings)