Update UI aesthetics and add smart JSON scanning

app.py

@@ -43,6 +43,19 @@ def get_output_file(state: str) -> str:
 def get_excel_file(state: str) -> str:
     return os.path.join(EXCEL_DIR, f"{_state_prefix(state)}_Schools.xlsx")
 
+def check_missed_schools(json_path: str):
+    if not os.path.exists(json_path):
+        return None
+    try:
+        import json
+        with open(json_path, 'r', encoding='utf-8') as f:
+            data = json.load(f)
+        if not isinstance(data, list):
+            return None
+        return sum(1 for item in data if item.get("status") != "success")
+    except Exception:
+        return None
+
 
 # ── HTML builders ─────────────────────────────────────────────────────────────
 
@@ -264,6 +277,20 @@ def _stream(pytest_args: list, state: str, max_retries: int, mode: str):
         yield (_stopped_html(), _phase_html("Stopped by user."), result_rendered,
                "\n".join(log_lines[-80:]), out_file if os.path.exists(out_file) else None, gr.update(), gr.update(value="⏹ Stopped", variant="secondary"))
     else:
+        # Post-run scan of JSON file to guarantee state
+        failed_count = check_missed_schools(out_file)
+        if failed_count is not None:
+            if failed_count == 0:
+                is_done = True
+                has_failures = False
+                phase_text = "✅ All data retrieved successfully!"
+                result_rendered += "<br/><div style='color:green; font-weight:bold;'>File scanned: All schools retrieved.</div>"
+            else:
+                is_done = True
+                has_failures = True
+                phase_text = f"⚠️ {failed_count} combos missed. Click 'Run Retry'."
+                result_rendered += f"<br/><div style='color:orange; font-weight:bold;'>File scanned: {failed_count} combos still missing.</div>"
+
         yield _emit()
 
 
@@ -287,6 +314,12 @@ def ui_retry(state, max_retries):
     if not os.path.exists(out):
         yield _idle_html(), _phase_html(f"⚠️ No output file for {state}. Run Main Scrape first."), "", "", None, gr.update(), gr.update()
         return
+        
+    failed_count = check_missed_schools(out)
+    if failed_count == 0:
+        yield _idle_html(), _phase_html("✅ All data is already retrieved successfully! No need to retry."), "", "", out, gr.update(), gr.update(value="✅ ALL DATA RETRIEVED! You can download/export now.", variant="primary")
+        return
+        
     yield from _stream(
         ["pytest", "tests/test_retry_districts_by_category.py", "-v", "-s", "--state", state],
         state, max_retries, "retry",
@@ -354,9 +387,9 @@ custom_theme = gr.themes.Soft(
 
 with gr.Blocks(title="KYS Category Scraper", css=css, theme=custom_theme) as app:
 
-    gr.Markdown("# 🏫 KYS School Scraper — District × Category", elem_id="title")
+    gr.Markdown("# 🏫 KYS School Scraper", elem_id="title")
     gr.Markdown(
-        "Categories **3 · 5 · 6 · 7 · 10 · 11** applied at API call time. District-level.",
+        "<div style='text-align: center; color: var(--body-text-color-subdued);'>Categories 3 · 5 · 6 · 7 · 10 · 11</div>",
         elem_id="subtitle",
     )
 
@@ -370,10 +403,7 @@ with gr.Blocks(title="KYS Category Scraper", css=css, theme=custom_theme) as app
                 download_btn = gr.File(label="⬇️  Download JSON", interactive=False)
                 excel_dl_btn = gr.File(label="⬇️  Download Excel", interactive=False)
 
-    # ── Retries slider ────────────────────────────────────────────────────────
-    retries_slider = gr.Slider(minimum=1, maximum=10, value=5, step=1,
-                               label="🔁  Max captcha retry rounds per combo  (each round = 2 API calls)")
-
+    # (Settings moved to bottom)
     gr.Markdown("---")
 
     # ── Buttons ───────────────────────────────────────────────────────────────
@@ -385,13 +415,13 @@ with gr.Blocks(title="KYS Category Scraper", css=css, theme=custom_theme) as app
 
     with gr.Row():
         with gr.Column(scale=2):
-            gr.Markdown("_Starts fresh — deletes previous output for this state._")
+            gr.Markdown("<div style='text-align:center;'>_Starts fresh run_</div>")
         with gr.Column(scale=2):
-            gr.Markdown("_Merges into existing file. Only re-scrapes captcha-failed combos._")
+            gr.Markdown("<div style='text-align:center;'>_Retries failed captchas_</div>")
         with gr.Column(scale=2):
-            gr.Markdown("_Converts JSON to Schools.xlsx format._")
+            gr.Markdown("<div style='text-align:center;'>_Converts to Excel_</div>")
         with gr.Column(scale=1):
-            gr.Markdown("_Kills the running test immediately._")
+            gr.Markdown("<div style='text-align:center;'>_Stop process_</div>")
 
     gr.Markdown("---")
 
@@ -408,6 +438,11 @@ with gr.Blocks(title="KYS Category Scraper", css=css, theme=custom_theme) as app
                              show_label=False, elem_classes=["log-box"],
                              placeholder="Logs appear here during the run...")
 
+    # ── Settings ──────────────────────────────────────────────────────────────
+    with gr.Accordion("⚙️ Settings", open=False):
+        retries_slider = gr.Slider(minimum=1, maximum=10, value=5, step=1,
+                                   label="🔁  Max captcha retry rounds per combo")
+
     # ── Wiring ────────────────────────────────────────────────────────────────
     shared_outputs = [progress_html, phase_html, result_html, log_box, download_btn, retry_btn, status_indicator]