Merge branch 'feature/add_steps_display' into feature/graph_recommandation

app.py

@@ -15,6 +15,8 @@ import time
 import re
 import json
 
+from gradio import ChatMessage
+
 # from gradio_modal import Modal
 
 from io import BytesIO
@@ -121,10 +123,8 @@ async def chat(query,history,audience,sources,reports,current_graphs):
         reports = []
     
     inputs = {"user_input": query,"audience": audience_prompt,"sources":sources}
-    print(f"\n\nInputs:\n {inputs}\n\n")
-    result = agent.astream_events(inputs,version = "v1") #{"callbacks":[MyCustomAsyncHandler()]})
-    # result = rag_chain.stream(inputs)
-
+    result = agent.astream_events(inputs,version = "v1") 
+    
     # path_reformulation = "/logs/reformulation/final_output"
     # path_keywords = "/logs/keywords/final_output"
     # path_retriever = "/logs/find_documents/final_output"
@@ -146,130 +146,166 @@ async def chat(query,history,audience,sources,reports,current_graphs):
         "transform_query":("🔄️ Thinking step by step to answer the question",True),
         "retrieve_documents":("🔄️ Searching in the knowledge base",False),
     }
-
+    
+    used_documents = []
+    answer_message_content = ""
     try:
         async for event in result:
 
-            if event["event"] == "on_chat_model_stream" and event["metadata"]["langgraph_node"] in ["answer_rag", "answer_rag_no_docs", "answer_chitchat", "answer_ai_impact"]:
-                if start_streaming == False:
-                    start_streaming = True
-                    history[-1] = (query,"")
-
-                new_token = event["data"]["chunk"].content
-                # time.sleep(0.01)
-                previous_answer = history[-1][1]
-                previous_answer = previous_answer if previous_answer is not None else ""
-                answer_yet = previous_answer + new_token
-                answer_yet = parse_output_llm_with_sources(answer_yet)
-                history[-1] = (query,answer_yet)
-
-                if docs_used is True and event["metadata"]["langgraph_node"] in ["answer_rag_no_docs", "answer_chitchat", "answer_ai_impact"]:
-                    docs_used = False
+            # if event["event"] == "on_chat_model_stream" and event["metadata"]["langgraph_node"] in ["answer_rag", "answer_rag_no_docs", "answer_chitchat", "answer_ai_impact"]:
+            #     if start_streaming == False:
+            #         start_streaming = True
+            #         history[-1] = (query,"")
+            if "langgraph_node" in event["metadata"]:
+                node = event["metadata"]["langgraph_node"]
+
+                if event["event"] == "on_chain_end" and event["name"] == "retrieve_documents" :# when documents are retrieved
+                    try:
+                        docs = event["data"]["output"]["documents"]
+                        docs_html = []
+                        for i, d in enumerate(docs, 1):
+                            docs_html.append(make_html_source(d, i))
+                        
+                        used_documents = used_documents + [d.metadata["name"] for d in docs]
+                        history[-1].content = "Adding sources :\n\n - " + "\n - ".join(np.unique(used_documents))
+                            
+                        docs_html = "".join(docs_html)
+                        
+                    except Exception as e:
+                        print(f"Error getting documents: {e}")
+                        print(event)
+ 
+                elif event["name"] in steps_display.keys() and event["event"] == "on_chain_start": #display steps
+                    event_description,display_output = steps_display[node]
+                    if not hasattr(history[-1], 'metadata') or history[-1].metadata["title"] != event_description: # if a new step begins
+                        history.append(ChatMessage(role="assistant", content = "", metadata={'title' :event_description}))
+ 
+                elif event["name"] != "transform_query" and event["event"] == "on_chat_model_stream" and node in ["answer_rag", "answer_search"]:# if streaming answer
+                    if start_streaming == False:
+                        start_streaming = True
+                        history.append(ChatMessage(role="assistant", content = ""))
+                    answer_message_content +=  event["data"]["chunk"].content
+                    answer_message_content = parse_output_llm_with_sources(answer_message_content)
+                    history[-1] = ChatMessage(role="assistant", content = answer_message_content)
+                    # history.append(ChatMessage(role="assistant", content = new_message_content))
+
+            #     if docs_used is True and event["metadata"]["langgraph_node"] in ["answer_rag_no_docs", "answer_chitchat", "answer_ai_impact"]:
+            #         docs_used = False
             
-            elif docs_used is True and event["name"] == "retrieve_documents" and event["event"] == "on_chain_end":
-                try:
-                    docs = event["data"]["output"]["documents"]
-                    docs_html = []
-                    for i, d in enumerate(docs, 1):
-                        docs_html.append(make_html_source(d, i))
-                    docs_html = "".join(docs_html)
-
-                except Exception as e:
-                    print(f"Error getting documents: {e}")
-                    print(event)
-
-            # elif event["name"] == "retrieve_documents" and event["event"] == "on_chain_start":
-            #     print(event)
-            #     questions = event["data"]["input"]["questions"]
-            #     questions = "\n".join([f"{i+1}. {q['question']} ({q['source']})" for i,q in enumerate(questions)])
-            #     answer_yet = "🔄️ Searching in the knowledge base\n{questions}"
-            #     history[-1] = (query,answer_yet)
-
-            elif event["name"] in ["retrieve_graphs", "retrieve_graphs_ai"] and event["event"] == "on_chain_end":
-                try:
-                    recommended_content = event["data"]["output"]["recommended_content"]
-                    # graphs = [
-                    #     {
-                    #         "embedding": x.metadata["returned_content"],
-                    #         "metadata": {
-                    #             "source": x.metadata["source"],
-                    #             "category": x.metadata["category"]
-                    #             }
-                    #             } for x in recommended_content if x.metadata["source"] == "OWID"
-                    #             ]
+            # elif docs_used is True and event["name"] == "retrieve_documents" and event["event"] == "on_chain_end":
+            #     try:
+            #         docs = event["data"]["output"]["documents"]
+            #         docs_html = []
+            #         for i, d in enumerate(docs, 1):
+            #             docs_html.append(make_html_source(d, i))
+            #         docs_html = "".join(docs_html)
+
+            #     except Exception as e:
+            #         print(f"Error getting documents: {e}")
+            #         print(event)
+
+            # # elif event["name"] == "retrieve_documents" and event["event"] == "on_chain_start":
+            # #     print(event)
+            # #     questions = event["data"]["input"]["questions"]
+            # #     questions = "\n".join([f"{i+1}. {q['question']} ({q['source']})" for i,q in enumerate(questions)])
+            # #     answer_yet = "🔄️ Searching in the knowledge base\n{questions}"
+            # #     history[-1] = (query,answer_yet)
+
+            # elif event["name"] in ["retrieve_graphs", "retrieve_graphs_ai"] and event["event"] == "on_chain_end":
+            #     try:
+            #         recommended_content = event["data"]["output"]["recommended_content"]
+            #         # graphs = [
+            #         #     {
+            #         #         "embedding": x.metadata["returned_content"],
+            #         #         "metadata": {
+            #         #             "source": x.metadata["source"],
+            #         #             "category": x.metadata["category"]
+            #         #             }
+            #         #             } for x in recommended_content if x.metadata["source"] == "OWID"
+            #         #             ]
                     
-                    unique_graphs = []
-                    seen_embeddings = set()
+            #         unique_graphs = []
+            #         seen_embeddings = set()
 
-                    for x in recommended_content:
-                        embedding = x.metadata["returned_content"]
+            #         for x in recommended_content:
+            #             embedding = x.metadata["returned_content"]
                         
-                        # Check if the embedding has already been seen
-                        if embedding not in seen_embeddings:
-                            unique_graphs.append({
-                                "embedding": embedding,
-                                "metadata": {
-                                    "source": x.metadata["source"],
-                                    "category": x.metadata["category"]
-                                }
-                            })
-                            # Add the embedding to the seen set
-                            seen_embeddings.add(embedding)
+            #             # Check if the embedding has already been seen
+            #             if embedding not in seen_embeddings:
+            #                 unique_graphs.append({
+            #                     "embedding": embedding,
+            #                     "metadata": {
+            #                         "source": x.metadata["source"],
+            #                         "category": x.metadata["category"]
+            #                     }
+            #                 })
+            #                 # Add the embedding to the seen set
+            #                 seen_embeddings.add(embedding)
 
     
-                    categories = {}
-                    for graph in unique_graphs:
-                        category = graph['metadata']['category']
-                        if category not in categories:
-                            categories[category] = []
-                        categories[category].append(graph['embedding'])
-
-                    # graphs_html = ""
-                    for category, embeddings in categories.items():
-                        # graphs_html += f"<h3>{category}</h3>"
-                        # current_graphs.append(f"<h3>{category}</h3>")
-                        for embedding in embeddings:
-                            current_graphs.append([embedding, category])
-                            # graphs_html += f"<div>{embedding}</div>"
+            #         categories = {}
+            #         for graph in unique_graphs:
+            #             category = graph['metadata']['category']
+            #             if category not in categories:
+            #                 categories[category] = []
+            #             categories[category].append(graph['embedding'])
+
+            #         # graphs_html = ""
+            #         for category, embeddings in categories.items():
+            #             # graphs_html += f"<h3>{category}</h3>"
+            #             # current_graphs.append(f"<h3>{category}</h3>")
+            #             for embedding in embeddings:
+            #                 current_graphs.append([embedding, category])
+            #                 # graphs_html += f"<div>{embedding}</div>"
                                                 
-                except Exception as e:
-                    print(f"Error getting graphs: {e}")
-
-            for event_name,(event_description,display_output) in steps_display.items():
-                if event["name"] == event_name:
-                    if event["event"] == "on_chain_start":
-                        # answer_yet = f"<p><span class='loader'></span>{event_description}</p>"
-                        # answer_yet = make_toolbox(event_description, "", checked = False)
-                        answer_yet = event_description
-
-                        history[-1] = (query,answer_yet)
-                    # elif event["event"] == "on_chain_end":
-                    #     answer_yet = ""
-                    #     history[-1] = (query,answer_yet)
-                        # if display_output:
-                        #     print(event["data"]["output"])
-
-            # if op['path'] == path_reformulation: # reforulated question
-            #     try:
-            #         output_language = op['value']["language"] # str
-            #         output_query = op["value"]["question"]
             #     except Exception as e:
-            #         raise gr.Error(f"ClimateQ&A Error: {e} - The error has been noted, try another question and if the error remains, you can contact us :)")
+            #         print(f"Error getting graphs: {e}")
+
+            # for event_name,(event_description,display_output) in steps_display.items():
+            #     if event["name"] == event_name:
+            #         if event["event"] == "on_chain_start":
+            #             # answer_yet = f"<p><span class='loader'></span>{event_description}</p>"
+            #             # answer_yet = make_toolbox(event_description, "", checked = False)
+            #             answer_yet = event_description
+
+            #             history[-1] = (query,answer_yet)
+            #         # elif event["event"] == "on_chain_end":
+            #         #     answer_yet = ""
+            #         #     history[-1] = (query,answer_yet)
+            #             # if display_output:
+            #             #     print(event["data"]["output"])
+
+            # # if op['path'] == path_reformulation: # reforulated question
+            # #     try:
+            # #         output_language = op['value']["language"] # str
+            # #         output_query = op["value"]["question"]
+            # #     except Exception as e:
+            # #         raise gr.Error(f"ClimateQ&A Error: {e} - The error has been noted, try another question and if the error remains, you can contact us :)")
             
-            # if op["path"] == path_keywords:
-            #     try:
-            #         output_keywords = op['value']["keywords"] # str
-            #         output_keywords = " AND ".join(output_keywords)
-            #     except Exception as e:
-            #         pass
+            # # if op["path"] == path_keywords:
+            # #     try:
+            # #         output_keywords = op['value']["keywords"] # str
+            # #         output_keywords = " AND ".join(output_keywords)
+            # #     except Exception as e:
+            # #         pass
 
 
 
-            history = [tuple(x) for x in history]
-            yield history,docs_html,output_query,output_language,gallery,current_graphs #,output_query,output_keywords
+            # history = [tuple(x) for x in history]
+            # yield history,docs_html,output_query,output_language,gallery,current_graphs #,output_query,output_keywords
 
 
+                if event["name"] == "transform_query" and event["event"] =="on_chain_end":
+                    if hasattr(history[-1],"content"):
+                        history[-1].content += "Decompose question into sub-questions: \n\n - " + "\n - ".join([q["question"] for q in event["data"]["output"]["remaining_questions"]])
+                        
+                if event["name"] == "categorize_intent" and event["event"] == "on_chain_start":
+                    print("X")
+            
+            yield history,docs_html,output_query,output_language,gallery #,output_query,output_keywords
+ 
     except Exception as e:
+        print(event, "has failed")
         raise gr.Error(f"{e}")
 
 
@@ -330,23 +366,24 @@ async def chat(query,history,audience,sources,reports,current_graphs):
         history[-1] = (history[-1][0],answer_yet)
         history = [tuple(x) for x in history]
 
-        print(f"\n\nImages:\n{gallery}")
+#         print(f"\n\nImages:\n{gallery}")
 
-    # gallery = [x.metadata["image_path"] for x in docs if (len(x.metadata["image_path"]) > 0 and "IAS" in x.metadata["image_path"])]
-    # if len(gallery) > 0:
-    #     gallery = list(set("|".join(gallery).split("|")))
-    #     gallery = [get_image_from_azure_blob_storage(x) for x in gallery]
+#     # gallery = [x.metadata["image_path"] for x in docs if (len(x.metadata["image_path"]) > 0 and "IAS" in x.metadata["image_path"])]
+#     # if len(gallery) > 0:
+#     #     gallery = list(set("|".join(gallery).split("|")))
+#     #     gallery = [get_image_from_azure_blob_storage(x) for x in gallery]
 
-        yield history,docs_html,output_query,output_language,gallery,current_graphs #,output_query,output_keywords
+#         yield history,docs_html,output_query,output_language,gallery,current_graphs #,output_query,output_keywords
 
 
 
-#     else:
-#         docs_string = "No relevant passages found in the climate science reports (IPCC and IPBES)"
-#         complete_response = "**No relevant passages found in the climate science reports (IPCC and IPBES), you may want to ask a more specific question (specifying your question on climate issues).**"
-#         messages.append({"role": "assistant", "content": complete_response})
-#         gradio_format = make_pairs([a["content"] for a in messages[1:]])
-#         yield gradio_format, messages, docs_string
+# #     else:
+# #         docs_string = "No relevant passages found in the climate science reports (IPCC and IPBES)"
+# #         complete_response = "**No relevant passages found in the climate science reports (IPCC and IPBES), you may want to ask a more specific question (specifying your question on climate issues).**"
+# #         messages.append({"role": "assistant", "content": complete_response})
+# #         gradio_format = make_pairs([a["content"] for a in messages[1:]])
+# #         yield gradio_format, messages, docs_string
+    yield history,docs_html,output_query,output_language,gallery#,output_query,output_keywords
 
 
 def save_feedback(feed: str, user_id):
@@ -392,56 +429,6 @@ papers_cols_widths = {
 papers_cols = list(papers_cols_widths.keys())
 papers_cols_widths = list(papers_cols_widths.values())
 
-# async def find_papers(query, keywords,after):
-
-#     summary = ""
-    
-#     df_works = oa.search(keywords,after = after)
-#     df_works = df_works.dropna(subset=["abstract"])
-#     df_works = oa.rerank(query,df_works,reranker)
-#     df_works = df_works.sort_values("rerank_score",ascending=False)
-#     G = oa.make_network(df_works)
-
-#     height = "750px"
-#     network = oa.show_network(G,color_by = "rerank_score",notebook=False,height = height)
-#     network_html = network.generate_html()
-
-#     network_html = network_html.replace("'", "\"")
-#     css_to_inject = "<style>#mynetwork { border: none !important; } .card { border: none !important; }</style>"
-#     network_html = network_html + css_to_inject
-
-    
-#     network_html = f"""<iframe style="width: 100%; height: {height};margin:0 auto" name="result" allow="midi; geolocation; microphone; camera; 
-#     display-capture; encrypted-media;" sandbox="allow-modals allow-forms 
-#     allow-scripts allow-same-origin allow-popups 
-#     allow-top-navigation-by-user-activation allow-downloads" allowfullscreen="" 
-#     allowpaymentrequest="" frameborder="0" srcdoc='{network_html}'></iframe>"""
-
-
-#     docs = df_works["content"].head(15).tolist()
-
-#     df_works = df_works.reset_index(drop = True).reset_index().rename(columns = {"index":"doc"})
-#     df_works["doc"] = df_works["doc"] + 1
-#     df_works = df_works[papers_cols]
-
-#     yield df_works,network_html,summary
-
-#     chain = make_rag_papers_chain(llm)
-#     result = chain.astream_log({"question": query,"docs": docs,"language":"English"})
-#     path_answer = "/logs/StrOutputParser/streamed_output/-"
-
-#     async for op in result:
-
-#         op = op.ops[0]
-
-#         if op['path'] == path_answer: # reforulated question
-#             new_token = op['value'] # str
-#             summary += new_token
-#         else:
-#             continue
-#         yield df_works,network_html,summary
-    
-
 
 # --------------------------------------------------------------------
 # Gradio
@@ -478,23 +465,28 @@ def save_graph(saved_graphs_state, embedding, category):
     return saved_graphs_state, gr.Button("Graph Saved")
 
 
-with gr.Blocks(title="Climate Q&A", css="style.css", theme=theme,elem_id = "main-component") as demo:
-    user_id_state = gr.State([user_id])
+# with gr.Blocks(title="Climate Q&A", css="style.css", theme=theme,elem_id = "main-component") as demo:
+#     user_id_state = gr.State([user_id])
 
-    chat_completed_state = gr.State(0)
-    current_graphs = gr.State([])
-    saved_graphs = gr.State({})
+#     chat_completed_state = gr.State(0)
+#     current_graphs = gr.State([])
+#     saved_graphs = gr.State({})
+with gr.Blocks(title="Climate Q&A", css_paths=os.getcwd()+ "/style.css", theme=theme,elem_id = "main-component") as demo:
 
     with gr.Tab("ClimateQ&A"):
 
         with gr.Row(elem_id="chatbot-row"):
             with gr.Column(scale=2):
-                state = gr.State([system_template])
+                # state = gr.State([system_template])
                 chatbot = gr.Chatbot(
-                    value=[(None,init_prompt)],
-                    show_copy_button=True,show_label = False,elem_id="chatbot",layout = "panel",
+                    value = [ChatMessage(role="assistant", content=init_prompt)],
+                    type = "messages",
+                    show_copy_button=True,
+                    show_label = False,
+                    elem_id="chatbot",
+                    layout = "panel",
                     avatar_images = (None,"https://i.ibb.co/YNyd5W2/logo4.png"),
-                )#,avatar_images = ("assets/logo4.png",None))
+                )
                 
                 # bot.like(vote,None,None)
 
@@ -502,8 +494,7 @@ with gr.Blocks(title="Climate Q&A", css="style.css", theme=theme,elem_id = "main
 
                 with gr.Row(elem_id = "input-message"):
                     textbox=gr.Textbox(placeholder="Ask me anything here!",show_label=False,scale=7,lines = 1,interactive = True,elem_id="input-textbox")
-                    # submit = gr.Button("",elem_id = "submit-button",scale = 1,interactive = True,icon = "https://static-00.iconduck.com/assets.00/settings-icon-2048x2046-cw28eevx.png")
-
+                 
 
             with gr.Column(scale=1, variant="panel",elem_id = "right-panel"):
 
@@ -685,8 +676,9 @@ with gr.Blocks(title="Climate Q&A", css="style.css", theme=theme,elem_id = "main
 
 
     def start_chat(query,history):
-        history = history + [(query,None)]
-        history = [tuple(x) for x in history]
+        # history = history + [(query,None)]
+        # history = [tuple(x) for x in history]
+        history = history + [ChatMessage(role="user", content=query)]
         return (gr.update(interactive = False),gr.update(selected=1),history)
     
     def finish_chat():
@@ -722,8 +714,6 @@ with gr.Blocks(title="Climate Q&A", css="style.css", theme=theme,elem_id = "main
 
     dropdown_samples.change(change_sample_questions,dropdown_samples,samples)
 
-    # query_papers.submit(generate_keywords,[query_papers], [keywords_papers])
-    # search_papers.click(find_papers,[query_papers,keywords_papers,after], [papers_dataframe,citations_network,papers_summary])
 
     demo.queue()
 

style.css

@@ -2,6 +2,14 @@
 /* :root {
     --user-image: url('https://ih1.redbubble.net/image.4776899543.6215/st,small,507x507-pad,600x600,f8f8f8.jpg');
   } */
+.avatar-container.svelte-1x5p6hu:not(.thumbnail-item) img {
+    width: 100%;
+    height: 100%;
+    object-fit: cover;
+    border-radius: 50%;
+    padding: 0px;
+    margin: 0px;
+}
 
   .warning-box {
     background-color: #fff3cd;
@@ -57,6 +65,7 @@ body.dark .tip-box * {
 
 .message{
     font-size:14px !important;
+    
 }
 
 
@@ -65,6 +74,10 @@ a {
     color: inherit;
 }
 
+.doc-ref sup{
+    color:#dc2626!important;
+    /* margin-right:1px; */
+}
 .card {
     background-color: white;
     border-radius: 10px;
@@ -426,7 +439,7 @@ span.chatbot > p > img{
   
   .loader {
     border: 1px solid #d0d0d0 !important; /* Light grey background */
-    border-top: 1px solid #3498db !important; /* Blue color */
+    border-top: 1px solid #db3434 !important; /* Blue color */
     border-right: 1px solid #3498db !important; /* Blue color */
     border-radius: 50%;
     width: 20px;
@@ -492,4 +505,7 @@ div#tab-saved-graphs {
         max-height: 50vh; /* Reduce height for smaller screens */
         overflow-y: auto;
     }
-}
+}
+.message-buttons-left.panel.message-buttons.with-avatar {
+    display: none;
+}