Update app.py

app.py

@@ -5,30 +5,27 @@ import docx
 from langchain.chat_models import ChatOpenAI
 from langchain.schema import SystemMessage, HumanMessage
 from rapidfuzz import fuzz
+import concurrent.futures
 
 # ---------- استایل ----------
 st.markdown("""
     <style>
     @import url('https://fonts.googleapis.com/css2?family=Vazirmatn:wght@400;700&display=swap');
-
     html, body, [class*="css"] {
         font-family: 'Vazirmatn', Tahoma, sans-serif;
         direction: rtl;
         text-align: right;
     }
-
     .stApp {
         background: linear-gradient(to left, #4b5e40, #2e3b2e);
         color: #ffffff;
     }
-
     [data-testid="stSidebar"] {
         width: 260px !important;
         background-color: #1a2b1e;
         border: none !important;
         padding-top: 20px;
     }
-
     .menu-item {
         display: flex;
         align-items: center;
@@ -39,17 +36,14 @@ st.markdown("""
         cursor: pointer;
         transition: background-color 0.3s ease;
     }
-
     .menu-item:hover {
         background-color: #2e3b2e;
         color: #b8860b;
     }
-
     .menu-item img {
         width: 24px;
         height: 24px;
     }
-
     .stButton>button {
         background-color: #b8860b !important;
         color: #1a2b1e !important;
@@ -63,13 +57,11 @@ st.markdown("""
         width: 100%;
         margin: 10px 0;
     }
-
     .stButton>button:hover {
         background-color: #8b6508 !important;
         transform: translateY(-2px);
         box-shadow: 0 4px 8px rgba(0,0,0,0.3);
     }
-
     .header-text {
         text-align: center;
         margin: 20px 0;
@@ -78,20 +70,17 @@ st.markdown("""
         border-radius: 15px;
         box-shadow: 0 6px 12px rgba(0,0,0,0.4);
     }
-
     .header-text h1 {
         font-size: 42px;
         color: #b8860b;
         margin: 0;
         font-weight: 700;
     }
-
     .subtitle {
         font-size: 18px;
         color: #d4d4d4;
         margin-top: 10px;
     }
-
     .chat-message {
         background-color: rgba(26, 43, 30, 0.95);
         border: 2px solid #b8860b;
@@ -106,12 +95,10 @@ st.markdown("""
         align-items: center;
         gap: 15px;
     }
-
     @keyframes fadeIn {
         from { opacity: 0; transform: translateY(10px); }
         to { opacity: 1; transform: translateY(0); }
     }
-
     .stTextInput>div>input, .stTextArea textarea {
         background-color: rgba(26, 43, 30, 0.95) !important;
         border-radius: 10px !important;
@@ -121,12 +108,10 @@ st.markdown("""
         font-size: 16px;
         color: #d4d4d4 !important;
     }
-
     hr {
         border: 1px solid #b8860b;
         margin: 15px 0;
     }
-
     [data-testid="stSidebar"] > div {
         border: none !important;
     }
@@ -186,33 +171,41 @@ llm = ChatOpenAI(
     model="meta-llama/Llama-3.3-70B-Instruct-Turbo-Free"
 )
 
-# ---------- پردازش فایل‌ها ----------
-folder_path = 'C:/Users/ici/Downloads/Telegram Desktop/45/46'
-texts = []
-
-for filename in os.listdir(folder_path):
-    if filename.endswith(".docx"):
-        full_path = os.path.join(folder_path, filename)
-        doc = docx.Document(full_path)
-        file_text = "\n".join([para.text for para in doc.paragraphs])
-        if file_text.strip():
-            texts.append(file_text)
-
+# ---------- پردازش فایل‌ها با کش و موازی ----------
+folder_path = '46'
 normalizer = Normalizer()
 sentence_tokenizer = SentenceTokenizer()
 
-all_sentences = []
-for text in texts:
-    normalized = normalizer.normalize(text)
-    sentences = sentence_tokenizer.tokenize(normalized)
-    all_sentences.extend(sentences)
+@st.cache_data(show_spinner="در حال پردازش اسناد... لطفاً صبور باشید.")
+def load_and_process_documents(path):
+    def process_docx(filename):
+        try:
+            full_path = os.path.join(path, filename)
+            doc = docx.Document(full_path)
+            file_text = "\n".join([para.text for para in doc.paragraphs])
+            if file_text.strip():
+                normalized = normalizer.normalize(file_text)
+                return sentence_tokenizer.tokenize(normalized)
+        except:
+            return []
+        return []
+
+    all_sentences = []
+    with concurrent.futures.ThreadPoolExecutor() as executor:
+        results = executor.map(process_docx, [f for f in os.listdir(path) if f.endswith(".docx")])
+        for sentences in results:
+            if sentences:
+                all_sentences.extend(sentences)
+    return all_sentences
+
+all_sentences = load_and_process_documents(folder_path)
 
 # ---------- ورودی جستجو ----------
 query = st.text_input("🔎 کلمه یا عبارت موردنظر خود را وارد کنید:")
 
 if query:
     found = False
-    threshold = 80
+    threshold = 70
 
     for idx, sentence in enumerate(all_sentences):
         similarity = fuzz.partial_ratio(query, sentence)
@@ -240,4 +233,4 @@ if query:
             HumanMessage(content=prompt)
         ])
         rewritten = response.content.strip()
-        st.markdown(f'<div class="chat-message">{rewritten}</div>', unsafe_allow_html=True)
+        st.markdown(f'<div class="chat-message">{rewritten}</div>', unsafe_allow_html=True)