Update app.py

app.py

@@ -13,30 +13,7 @@ from llama_index.core.chat_engine.condense_plus_context import CondensePlusConte
 from llama_index.core.schema import Document
 
 # ===================================
-# 1️⃣ Fungsi Membaca Data PKB.json
-# ===================================
-def read_pkb_json():
-    try:
-        with open("pkb.json", "r", encoding="utf-8") as file:
-            data = json.load(file)
-        
-        pkb_text = "=== Perjanjian Kerja Bersama ===\n"
-        for bab, content in data["perjanjian_kerja_bersama"].items():
-            pkb_text += f"\n## {content['judul']} ##\n"
-            for pasal, pasal_data in content.items():
-                if pasal != "judul":
-                    pkb_text += f"\n### {pasal_data['judul']} ###\n"
-                    for item in pasal_data["isi"]:
-                        if isinstance(item, dict):
-                            pkb_text += f"- {item['istilah']}: {item['definisi']}\n"
-                        else:
-                            pkb_text += f"- {item}\n"
-        return pkb_text.strip()
-    except Exception as e:
-        return f"❌ ERROR membaca PKB.json: {str(e)}"
-
-# ===================================
-# 2️⃣ Fungsi Membaca Data Google Spreadsheet
+# 1️⃣ Fungsi Membaca Data Google Spreadsheet
 # ===================================
 def read_google_sheets():
     try:
@@ -45,7 +22,7 @@ def read_google_sheets():
         client = gspread.authorize(creds)
         
         SPREADSHEET_ID = "1e_cNMhwF-QYpyYUpqQh-XCw-OdhWS6EuYsoBUsVtdNg"
-        sheet_names = ["datatarget", "datacuti", "dataabsen", "datalembur"]
+        sheet_names = ["datatarget", "datacuti", "dataabsen", "datalembur", "pkb"]
 
         all_data = []
         spreadsheet = client.open_by_key(SPREADSHEET_ID)
@@ -69,7 +46,7 @@ def read_google_sheets():
         return f"❌ ERROR: {str(e)}"
 
 # ===================================
-# 3️⃣ Inisialisasi Model Llama
+# 2️⃣ Inisialisasi Model Llama
 # ===================================
 def initialize_llama_model():
     model_path = hf_hub_download(
@@ -80,18 +57,18 @@ def initialize_llama_model():
     return model_path
 
 # ===================================
-# 4️⃣ Inisialisasi Pengaturan Model
+# 3️⃣ Inisialisasi Pengaturan Model
 # ===================================
 def initialize_settings(model_path):
     Settings.llm = LlamaCPP(model_path=model_path, temperature=0.7)
 
 # ===================================
-# 5️⃣ Inisialisasi Index & Chat Engine
+# 4️⃣ Inisialisasi Index & Chat Engine
 # ===================================
 def initialize_index():
-    text_data = read_google_sheets() + "\n" + read_pkb_json()
+    text_data = read_google_sheets()
     document = Document(text=text_data)
-    parser = SentenceSplitter(chunk_size=150, chunk_overlap=10)
+    parser = SentenceSplitter(chunk_size=100, chunk_overlap=30)
     nodes = parser.get_nodes_from_documents([document])
     
     embedding = HuggingFaceEmbedding("sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2")
@@ -109,7 +86,7 @@ def initialize_chat_engine(index):
     return chat_engine
 
 # ===================================
-# 6️⃣ Fungsi untuk Merapikan Jawaban Chatbot
+# 5️⃣ Fungsi untuk Merapikan Jawaban Chatbot
 # ===================================
 def clean_response(response):
     text = "".join(response.response_gen)  # Gabungkan teks yang dihasilkan
@@ -118,7 +95,7 @@ def clean_response(response):
     return text
 
 # ===================================
-# 7️⃣ Fungsi untuk Menghasilkan Respons Chatbot
+# 6️⃣ Fungsi untuk Menghasilkan Respons Chatbot
 # ===================================
 def generate_response(message, history, chat_engine):
     if history is None:
@@ -143,7 +120,7 @@ def generate_response(message, history, chat_engine):
     return cleaned_text
 
 # ===================================
-# 8️⃣ Fungsi Utama untuk Menjalankan Aplikasi
+# 7️⃣ Fungsi Utama untuk Menjalankan Aplikasi
 # ===================================
 def main():
     model_path = initialize_llama_model()