Update app.py

app.py

@@ -2,15 +2,15 @@ import gradio as gr
 import numpy as np
 import urllib.parse
 import traceback
-from ocr_engine import ocr_image
-from keyword_engine import extract_keywords
+from ocr_engine import ocr_items, ocr_text
+from keyword_engine import build_focus_text, extract_keywords_from_focus
 
 def build_link(keywords, full_text):
     if keywords:
         query = " ".join(keywords)
         shown = keywords
     else:
-        query = (full_text or "")[:100]
+        query = (full_text or "")[:120]
         shown = ["(dùng văn bản gốc)"]
 
     url = "https://www.google.com/search?q=" + urllib.parse.quote(query)
@@ -35,11 +35,16 @@ def process(image):
         if not isinstance(image, np.ndarray):
             return "Định dạng ảnh không hợp lệ.", "", ""
 
-        full_text = ocr_image(image)
+        full_text = ocr_text(image)
         if not full_text:
             return "Không trích xuất được nội dung.", "", ""
 
-        keywords = extract_keywords(full_text, top_n=5)
+        items = ocr_items(image)
+        focus_text = build_focus_text(items, img_h=image.shape[0], top_ratio=0.35, max_lines=30)
+        if not focus_text:
+            focus_text = full_text
+
+        keywords = extract_keywords_from_focus(focus_text, top_n=5)
         html_link, keywords_str = build_link(keywords, full_text)
         return full_text, keywords_str, html_link