Update space

pages/main/scripts.py

@@ -1,13 +1,15 @@
 import faiss
 import gradio as gr
-from typing import Any, Generator
+from typing import Any, Generator, Iterator
 from sentence_transformers import SentenceTransformer
-from utils.llm_response import generate_response_with_llm # A função unificada agora trata as estratégias de RAG e LLM
+from utils.llm_response import generate_response_with_llm  # A função unificada agora trata as estratégias de RAG e LLM
 from utils.phrase_extractor import process_file_content
-#from utils.report_creation import generate_report
+
+# from utils.report_creation import generate_report
 from .strings import STRINGS
+
 # DEPRECATED: A função volta com a consolidação de um futuro OCR.
-#def extract_phrases_from_gradio_file(gradio_file: gr.File) -> gr.Textbox:
+# def extract_phrases_from_gradio_file(gradio_file: gr.File) -> gr.Textbox:
 #    """
 #    Utilizes the 'process_file' function from 'utils.phrase_extractor' to read the
 #    file content and extract phrases, returning them as a text block for Gradio.
@@ -18,20 +20,23 @@ from .strings import STRINGS
 #    try:
 #        # Chama a função unificada de processamento de arquivo que retorna uma lista de frases
 #        phrases = process_file_content(gradio_file.name)
-#        
+#
 #        phrases_text = "\n".join(phrases)
 #        return gr.Textbox(value=phrases_text, placeholder=STRINGS["TEXT_INPUT_PLACEHOLDER_LOADED"])
 #    except Exception as e:
 #        return gr.Textbox(value=f"Error: {e}", placeholder=STRINGS["TEXT_INPUT_PLACER_EMPTY"])
 
-def process_phrases_with_rag_llm(input_phrases_text: str, rag_docs:list[str], rag_index:faiss.Index, rag_embedder:SentenceTransformer) -> Generator[tuple[gr.Textbox, gr.Textbox, gr.Tabs, gr.TabItem]]:
+
+def process_phrases_with_rag_llm(
+    input_phrases_text: str, rag_docs: list[str], rag_index: faiss.Index, rag_embedder: SentenceTransformer
+) -> Iterator[tuple[gr.Textbox, gr.Textbox, gr.Tabs, gr.TabItem]]:
     """
     Receives a block of text (phrases separated by newlines) and processes it
     with the RAG+LLM API (`res_generate_API`) using a multiple-context strategy.
     Returns a status textbox, a formatted responses textbox, and updates tabs to switch to the results tab.
     """
-    print(f"Processando o bloco de frases para geração de resposta: \"{input_phrases_text[:100]}...\"")
-    current_symbol = " ♾️"  # Emojis para indicar status de processamento e sucesso    
+    print(f'Processando o bloco de frases para geração de resposta: "{input_phrases_text[:100]}..."')
+    current_symbol = " ♾️"  # Emojis para indicar status de processamento e sucesso
 
     # --- Ação 1: Mudar de aba IMEDIATAMENTE e mostrar mensagem de processamento ---
     # O 'yield' envia: (Status, Resultado, Tabs)
@@ -39,9 +44,9 @@ def process_phrases_with_rag_llm(input_phrases_text: str, rag_docs:list[str], ra
         gr.update(value=STRINGS["TXTBOX_STATUS_IDLE"], interactive=False),
         gr.update(value="", interactive=False),
         gr.update(selected=1),
-        gr.update(label=STRINGS["TAB_1_TITLE"]+current_symbol, interactive=True)
-        )
-    
+        gr.update(label=STRINGS["TAB_1_TITLE"] + current_symbol, interactive=True),
+    )
+
     # time.sleep(1)  # Simula um pequeno atraso para processamento
 
     try:
@@ -53,16 +58,16 @@ def process_phrases_with_rag_llm(input_phrases_text: str, rag_docs:list[str], ra
             documents=rag_docs,
             index=rag_index,
             embedder=rag_embedder,
-            llm_choice='gemini', # ou 'ollama', conforme a necessidade
-            rag_strategy='multiple' # A chave para usar a busca por múltiplos contextos
+            llm_choice="gemini",  # ou 'ollama', conforme a necessidade
+            rag_strategy="multiple",  # A chave para usar a busca por múltiplos contextos
         )
 
-#        with open("./sandbox/respostateste.txt", "r", encoding="utf-8") as arquivo:
-#            llm_response = arquivo.read() #TODO: Test Only
+        #        with open("./sandbox/respostateste.txt", "r", encoding="utf-8") as arquivo:
+        #            llm_response = arquivo.read() #TODO: Test Only
 
         status_message = STRINGS["TXTBOX_STATUS_OK"]
         formatted_output = f"--- Resposta Fornecida pela LLM ---\n{llm_response}\n"
-        current_symbol = " ✅" 
+        current_symbol = " ✅"
 
     except Exception as e:
         status_message = STRINGS["TXTBOX_STATUS_ERROR"]
@@ -75,5 +80,5 @@ def process_phrases_with_rag_llm(input_phrases_text: str, rag_docs:list[str], ra
         gr.update(value=status_message, interactive=False),
         gr.update(value=formatted_output, interactive=False),
         gr.update(),
-        gr.update(label=STRINGS["TAB_1_TITLE"]+current_symbol, interactive=True)
+        gr.update(label=STRINGS["TAB_1_TITLE"] + current_symbol, interactive=True),
     )