Update app.py

app.py

@@ -54,6 +54,7 @@ from langchain.schema import Document
 import gradio as gr
 import tempfile
 import timeit
+import textwrap
 
 FILE_LOADER_MAPPING = {
     "csv": (CSVLoader, {"encoding": "utf-8"}),
@@ -171,11 +172,16 @@ def create_chain(llm, prompt, db):
     #     condense_question_prompt=CONDENSE_QUESTION_PROMPT,
     #     memory=memory,
     # )
+    # chain = RetrievalQA.from_chain_type(llm=llm,
+    #                                    chain_type='stuff',
+    #                                    retriever=db.as_retriever(search_kwargs={'k': 3}),
+    #                                    return_source_documents=True,
+    #                                    chain_type_kwargs={'prompt': prompt}
+    #                                    )
     chain = RetrievalQA.from_chain_type(llm=llm,
                                        chain_type='stuff',
                                        retriever=db.as_retriever(search_kwargs={'k': 3}),
-                                       return_source_documents=True,
-                                       chain_type_kwargs={'prompt': prompt}
+                                       return_source_documents=True
                                        )
     return chain
 
@@ -215,6 +221,18 @@ def create_retrieval_qa_bot(loaded_documents):
 
     return qa
 
+def wrap_text_preserve_newlines(text, width=110):
+    # Split the input text into lines based on newline characters
+    lines = text.split('\n')
+
+    # Wrap each line individually
+    wrapped_lines = [textwrap.fill(line, width=width) for line in lines]
+
+    # Join the wrapped lines back together using newline characters
+    wrapped_text = '\n'.join(wrapped_lines)
+
+    return wrapped_text
+
 def retrieve_bot_answer(query, loaded_documents):
     """
     Retrieves the answer to a given query using a QA bot.
@@ -236,8 +254,13 @@ def retrieve_bot_answer(query, loaded_documents):
     #     raise KeyError("Expected 'answer' key in bot_response, but it was not found.")
     # result = bot_response['answer']
     
-    result = bot_response['result']
-    sources = []
+    # result = bot_response['result']
+    # sources = []
+    # for source in bot_response["source_documents"]:
+    #     sources.append(source.metadata['source'])
+    # return result, sources
+
+    result = wrap_text_preserve_newlines(bot_response['result'])
     for source in bot_response["source_documents"]:
         sources.append(source.metadata['source'])
     return result, sources