add: greenlights to showcase widgets

qa/qa.py

@@ -49,6 +49,8 @@ def qa_main():
     index = None
     doc = None
 
+    upload_document_greenlight = False
+    uploaded_processed_document_greenlight = False
     # OpenAI API Key - TODO: consider adding a key valid for everyone
     st.header("Configure OpenAI API Key")
     st.warning('Please enter your OpenAI API Key!', icon='⚠️')
@@ -62,59 +64,80 @@ def qa_main():
     if user_secret:
         if set_openai_api_key(user_secret):
             st.success('OpenAI API key successfully provided!', icon='✅')
+            upload_document_greenlight = True
+    
+    if upload_document_greenlight:
+        # File that needs to be queried
+        st.header("Upload a file")
+        uploaded_file = st.file_uploader(
+            "Upload a pdf, docx, or txt file (scanned documents not supported)",
+            type=["pdf", "docx", "txt", "py", "json", "html", "css", "md"],
+            help="Scanned documents are not supported yet 🥲",
+            on_change=clear_submit, 
+            accept_multiple_files=multiple_files,
+        )
+            
+        # reading the uploaded file
+        if uploaded_file is not None:
+            # toggle internal file submission state to True
+            st.session_state["file_submitted"] = True
+            # parse the file using custom parsers
+            doc = file_to_doc(uploaded_file)
+            # converts the files into a list of documents
+            text = text_to_docs(text=tuple(doc))
 
-    # File that needs to be queried
-    st.header("Upload a file")
-    uploaded_file = st.file_uploader(
-        "Upload a pdf, docx, or txt file (scanned documents not supported)",
-        type=["pdf", "docx", "txt", "py", "json", "html", "css", "md"],
-        help="Scanned documents are not supported yet 🥲",
-        on_change=clear_submit, 
-        accept_multiple_files=multiple_files,
-    )
+            try:
+                with st.spinner("Indexing the document... This might take a while!"):
+                    index = embed_docs(tuple(text))
+                    st.session_state["api_key_configured"] = True
+            except OpenAIError as e:
+                st.error("OpenAI error encountered: ", e._message)
         
-    # reading the uploaded file
-    if uploaded_file is not None:
-        # toggle internal file submission state to True
-        st.session_state["file_submitted"] = True
-        # parse the file using custom parsers
-        doc = file_to_doc(uploaded_file)
-        # converts the files into a list of documents
-        text = text_to_docs(text=tuple(doc))
-
-        try:
-            with st.spinner("Indexing the document... This might take a while!"):
-                index = embed_docs(tuple(text))
-                st.session_state["api_key_configured"] = True
-        except OpenAIError as e:
-            st.error("OpenAI error encountered: ", e._message)
-
-    if "messages" not in st.session_state:
-        st.session_state["messages"] = []
-
-    for message in st.session_state.messages:
-        with st.chat_message(message["role"]):
-            st.markdown(message["content"])
-
-    if prompt := st.chat_input("Ask the document something..."):
-        st.session_state.messages.append({"role": "user", "content": prompt})
-        with st.chat_message("user"):
-            st.markdown(prompt)
-
-        with st.chat_message("assistant"):
-            message_placeholder = st.empty()
-            # retrieving the most relevant sources
-            sources = search_docs(index, prompt)
-            # producing the answer, live
-            full_response = ""
-            for answer_bit in get_answer(sources, prompt)["output_text"]:
-                full_response += answer_bit
-                message_placeholder.markdown(full_response + "▌")
-            
-            message_placeholder.markdown(full_response)
-            # answer = get_answer(sources, prompt)
-            # message_placeholder.markdown(answer["output_text"])
+            uploaded_processed_document_greenlight = True
         
-        # st.session_state.messages.append({"role": "assistant", "content": answer["output_text"]})
-        st.session_state.messages.append({"role": "assistant", "content": full_response})
+    if uploaded_processed_document_greenlight: 
+        if "messages" not in st.session_state:
+            st.session_state["messages"] = []
+
+        for message in st.session_state.messages:
+            with st.chat_message(message["role"]):
+                st.markdown(message["content"])
+
+        if prompt := st.chat_input("Ask the document something..."):
+            st.session_state.messages.append({"role": "user", "content": prompt})
+            with st.chat_message("user"):
+                st.markdown(prompt)
+
+            with st.chat_message("assistant"):
+                message_placeholder = st.empty()
+                # retrieving the most relevant sources
+                sources = search_docs(index, prompt)
+                # producing the answer, live
+                full_response = ""
+                for answer_bit in get_answer(sources, prompt)["output_text"]:
+                    full_response += answer_bit
+                    message_placeholder.markdown(full_response + "▌")
+                
+                message_placeholder.markdown(full_response)
+
+                # answer = get_answer(sources, prompt)
+                # message_placeholder.markdown(answer["output_text"])
+            
+            # st.session_state.messages.append({"role": "assistant", "content": answer["output_text"]})
+            st.session_state.messages.append({"role": "assistant", "content": full_response})
+
+# This might be useful to add memory to the chatbot harnessing a more low-level approach
+# llm = ChatOpenAI(temperature=0, model_name="gpt-3.5-turbo")
+
+# memory = ConversationBufferMemory(memory_key='chat_history', return_messages=True, output_key='answer')
+# retriever = your_vector_store.as_retriever()
+
+# # Create the multipurpose chain
+# qachat = ConversationalRetrievalChain.from_llm(
+#     llm=ChatOpenAI(temperature=0),
+#     memory=memory,
+#     retriever=retriever, 
+#     return_source_documents=True
+# )
 
+# qachat("Ask your question here...")