Update main.py

main.py

@@ -1,6 +1,7 @@
 import os
 import json
 import bcrypt
+import pandas as pd
 from typing import List
 from pathlib import Path
 from langchain_huggingface import HuggingFaceEmbeddings
@@ -16,7 +17,7 @@ from langchain.memory import ConversationBufferMemory
 from langchain.schema.runnable import Runnable, RunnablePassthrough, RunnableConfig, RunnableLambda
 from langchain.callbacks.base import BaseCallbackHandler
 from langchain.chains import (
-    StuffDocumentsChain, ConversationalRetrievalChain
+    StuffDocumentsChain, ConversationalRetrievalChain, create_extraction_chain
 )
 
 import chainlit as cl
@@ -52,6 +53,29 @@ async def Retriever(categorie):
         retriever = vectorstore.as_retriever(search_type="similarity_score_threshold", search_kwargs={"score_threshold": .7, "k": 250,"filter": {"title": {"$eq": "videos-table-rondeia"}, "time": {"$gte": 1320}}})
     return retriever
 
+@cl.step(type="tool")
+async def OtherRequest(answer):
+    schema = {
+        "properties": {
+            "Questions en relation avec le contexte": {"type": "string"},
+        },
+        "required": ["Questions en relation avec le contexte"],
+    }
+    llm = await LLMistral()
+    chainExtraction = create_extraction_chain(schema, llm)
+    dataframe = chainExtraction.invoke(GoogleTranslator(source='auto', target='fr').translate(answer))
+    actRequest = dataframe['text']
+    df_actRequest = pd.DataFrame(actRequest)
+    allRequest = pd.DataFrame(df_actRequest['Questions en relation avec le contexte'])
+    allRequest.drop_duplicates(keep = 'first', inplace=True)
+    allRequestArray = allRequest.values.tolist()
+    RequestArray = []
+    for act in allRequestArray:
+        RequestArray.append(cl.Starter(label=act[0], message=act[0], icon="/public/request-theme.svg",),)
+    print(RequestArray)
+    await cl.Message(content=RequestArray).send()
+    
+    
 @cl.step(type="embedding")
 async def Search(input, categorie):
     vectorstore = await VectorDatabase(categorie)
@@ -167,7 +191,8 @@ async def on_message(message: cl.Message):
 
     await cl.Message(content=GoogleTranslator(source='auto', target='fr').translate(answer)).send()
     
-    #search = vectorstore.similarity_search(message.content,k=50, filter={"categorie": {"$eq": "bibliographie-OPP-DGDIN"}})
+    await OtherRequest(answer)
+    
     search = await Search(message.content, "videosTC")
     
     sources = [