create FAISS content page with search functionality (#115)

Signed-off-by: sancheet230 <sancheet.p24@medhaviskillsuniversity.edu.in>

Author: sancheet230

src/mvt/menu.py

@@ -22,6 +22,7 @@ def authenticated_menu():
         st.sidebar.page_link("pages/prompt_management.py", label="Prompt Management")
         st.sidebar.page_link("pages/user_management.py", label="User Management")
         st.sidebar.page_link("pages/admin_responses.py", label="Admin Responses")
+        st.sidebar.page_link("pages/vector_db_content.py", label="DB Content")
     st.sidebar.page_link("app.py", label="About")
 
 def unauthenticated_menu():

src/mvt/pages/vector_db_content.py

@@ -0,0 +1,80 @@
+import streamlit as st
+import os
+from utils import load_yaml_file_with_db_prompts
+from dotenv import load_dotenv, find_dotenv
+from langchain_community.vectorstores import FAISS
+from langchain_mistralai.embeddings import MistralAIEmbeddings
+from menu import menu_with_redirect
+
+menu_with_redirect()
+
+# only allow admin users
+if st.session_state.user_type not in ["admin"]:
+    st.warning("You do not have permission to view this page.")
+    st.stop()
+
+st.markdown("# Vector DB Content (FAISS)")
+
+# load the configuration and environment variables
+config_data = load_yaml_file_with_db_prompts("config.yaml")
+load_dotenv(find_dotenv())
+
+# set up the embeddings model based on the provider
+if config_data["llm_provider"] == "mistral":
+    embeddings = MistralAIEmbeddings(
+        model=config_data["embedding_model"],
+        mistral_api_key=os.getenv("MISTRALAI_API_KEY")
+    )
+else:
+    from langchain_openai import OpenAIEmbeddings
+    embeddings = OpenAIEmbeddings(
+        openai_api_key=os.getenv("OPENAI_API_KEY")
+    )
+
+# load the FAISS vector database from disk
+persist_dir = config_data["persist_directory"]
+try:
+    vectordb = FAISS.load_local(
+        persist_dir,
+        embeddings,
+        allow_dangerous_deserialization=True
+    )
+    st.success(f"Loaded FAISS content from '{persist_dir}'")
+except Exception as e:
+    st.error(f"Could not load FAISS content: {e}")
+    st.stop()
+
+# get all documents from the vector database (may be slow if many documents)
+try:
+    all_docs = list(vectordb.docstore._dict.values())
+except Exception as e:
+    st.error(f"Could not access documents in FAISS index: {e}")
+    st.stop()
+
+search_term = st.text_input("Search in vector DB content:", "")
+
+# Check if a document matches the search term 
+def doc_matches(doc, term):
+    if not term:
+        return True
+    terms = [t.strip().lower() for t in term.split() if t.strip()]
+    content = getattr(doc, 'page_content', '').lower()
+    metadata_values = [str(v).lower() for v in getattr(doc, 'metadata', {}).values()]
+    for t in terms:
+        if t not in content and not any(t in v for v in metadata_values):
+            return False
+    return True
+
+filtered_docs = [doc for doc in all_docs if doc_matches(doc, search_term)]
+
+st.markdown(f"### Showing {len(filtered_docs)} / {len(all_docs)} documents")
+
+# show each document with its metadata and a snippet of its content
+for i, doc in enumerate(filtered_docs):
+    with st.expander(f"Document {i+1}"):
+        st.write("**Metadata:**", doc.metadata)
+        snippet = doc.page_content[:500] + ("..." if len(doc.page_content) > 500 else "")
+        st.write("**Content snippet:**")
+        st.code(snippet)
+        with st.expander("Show full content"):
+            st.write(doc.page_content)