Update app.py

app.py

@@ -1,4 +1,4 @@
-# Import Library yang Diperlukan
+import json
 import gradio as gr
 import gspread
 from oauth2client.service_account import ServiceAccountCredentials
@@ -17,49 +17,40 @@ from llama_index.core.schema import Document
 # ===================================
 def read_google_sheets():
     try:
-        # Tentukan scope akses ke Google Sheets & Drive
         scope = ["https://www.googleapis.com/auth/spreadsheets", "https://www.googleapis.com/auth/drive"]
-        
-        # Load kredensial dari file credentials.json
         creds = ServiceAccountCredentials.from_json_keyfile_name("credentials.json", scope)
         client = gspread.authorize(creds)
-        
-        # ID Spreadsheet (tetap sama untuk semua sheet)
         SPREADSHEET_ID = "1e_cNMhwF-QYpyYUpqQh-XCw-OdhWS6EuYsoBUsVtdNg"
-
-        # 📌 Daftar nama worksheet yang akan dibaca
         sheet_names = ["datatarget", "datacuti", "dataabsen", "datalembur"]
 
-        all_data = []  # 🔹 List untuk menyimpan semua data
-
-        # 📌 Loop untuk membaca setiap worksheet
+        all_data = []
         spreadsheet = client.open_by_key(SPREADSHEET_ID)
         for sheet_name in sheet_names:
             try:
                 sheet = spreadsheet.worksheet(sheet_name)
                 data = sheet.get_all_values()
-                
-                # Tambahkan header nama sheet sebelum data untuk membedakan
                 all_data.append(f"=== Data dari {sheet_name.upper()} ===")
                 all_data.extend([" | ".join(row) for row in data])
-                all_data.append("\n")  # Pisahkan tiap sheet dengan newline
-                
+                all_data.append("\n")
             except gspread.exceptions.WorksheetNotFound:
                 all_data.append(f"❌ ERROR: Worksheet {sheet_name} tidak ditemukan.")
+        return "\n".join(all_data)
+    except Exception as e:
+        return f"❌ ERROR: {str(e)}"
 
-        # Gabungkan semua data menjadi satu string panjang
-        formatted_text = "\n".join(all_data)
-        
-        return formatted_text
-    
-    except gspread.exceptions.SpreadsheetNotFound:
-        return "❌ ERROR: Spreadsheet tidak ditemukan. Pastikan ID/nama benar!"
-    
+# ===================================
+# 2️⃣ Fungsi untuk Membaca File PKB JSON
+# ===================================
+def read_pkb_json():
+    try:
+        with open("pkb.json", "r", encoding="utf-8") as file:
+            pkb_data = json.load(file)
+        return json.dumps(pkb_data, indent=2, ensure_ascii=False)
     except Exception as e:
         return f"❌ ERROR: {str(e)}"
 
 # ===================================
-# 2️⃣ Fungsi untuk Mengunduh Model Llama
+# 3️⃣ Fungsi untuk Mengunduh Model Llama
 # ===================================
 def initialize_llama_model():
     model_path = hf_hub_download(
@@ -70,7 +61,7 @@ def initialize_llama_model():
     return model_path
 
 # ===================================
-# 3️⃣ Inisialisasi Model dan Pengaturan
+# 4️⃣ Inisialisasi Model dan Pengaturan
 # ===================================
 def initialize_settings(model_path):
     Settings.llm = LlamaCPP(
@@ -79,24 +70,24 @@ def initialize_settings(model_path):
     )
 
 # ===================================
-# 4️⃣ Inisialisasi Index dari Data Spreadsheet
+# 5️⃣ Inisialisasi Index dari Data Spreadsheet & PKB JSON
 # ===================================
 def initialize_index():
     text_data = read_google_sheets()
-    document = Document(text=text_data)
+    pkb_data = read_pkb_json()
+    combined_text = text_data + "\n\n" + pkb_data
+
+    document = Document(text=combined_text)
     documents = [document]
-    
     parser = SentenceSplitter(chunk_size=150, chunk_overlap=10)
     nodes = parser.get_nodes_from_documents(documents)
-    
     embedding = HuggingFaceEmbedding("sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2")
     Settings.embed_model = embedding
-    
     index = VectorStoreIndex(nodes)
     return index
 
 # ===================================
-# 5️⃣ Inisialisasi Mesin Chatbot
+# 6️⃣ Inisialisasi Mesin Chatbot
 # ===================================
 def initialize_chat_engine(index):
     retriever = index.as_retriever(similarity_top_k=3)
@@ -107,58 +98,28 @@ def initialize_chat_engine(index):
     return chat_engine
 
 # ===================================
-# 6️⃣ Fungsi untuk Menghasilkan Respons Chatbot
+# 7️⃣ Fungsi untuk Menghasilkan Respons Chatbot
 # ===================================
 def generate_response(message, history, chat_engine):
     if history is None:
         history = []
-    
-    text_data = read_google_sheets()
-    document = Document(text=text_data)
-    documents = [document]
-    
-    parser = SentenceSplitter(chunk_size=150, chunk_overlap=10)
-    nodes = parser.get_nodes_from_documents(documents)
-    index = VectorStoreIndex(nodes)
-    retriever = index.as_retriever(similarity_top_k=3)
-    
-    chat_engine = CondensePlusContextChatEngine.from_defaults(
-        retriever=retriever,
-        verbose=True,
-    )
-    
-    chat_messages = [
-    ChatMessage(
-        role="system",
-        content=(
-            "Anda adalah chatbot yang dirancang khusus untuk berbicara dalam Bahasa Indonesia. "
-            "Anda tidak diperbolehkan menjawab dalam bahasa lain, termasuk Inggris. "
-            "Gunakan gaya bahasa profesional tetapi tetap ramah. "
-            "Jika informasi tidak tersedia dalam dokumen, katakan dengan sopan bahwa Anda tidak tahu. "
-            "Pastikan setiap jawaban diberikan secara ringkas, jelas, dan sesuai konteks."
-        ),
-    ),
-]
-    
     response = chat_engine.stream_chat(message)
     text = "".join(response.response_gen)
-    
     history.append((message, text))
     return history
 
 # ===================================
-# 7️⃣ Fungsi Utama untuk Menjalankan Aplikasi
+# 8️⃣ Fungsi Utama untuk Menjalankan Aplikasi
 # ===================================
 def main():
     model_path = initialize_llama_model()
-    initialize_settings(model_path)  
-
+    initialize_settings(model_path)
     index = initialize_index()
-    chat_engine = initialize_chat_engine(index)  
-
+    chat_engine = initialize_chat_engine(index)
+    
     def chatbot_response(message, history):
-        return generate_response(message, history, chat_engine)  
-
+        return generate_response(message, history, chat_engine)
+    
     gr.Interface(
         fn=chatbot_response,
         inputs=["text"],