Upload app.py with huggingface_hub

app.py

@@ -386,9 +386,6 @@ st.markdown("""
       background:linear-gradient(90deg,#a78bfa,#e879f9);
       -webkit-background-clip:text;-webkit-text-fill-color:transparent;
       background-clip:text;line-height:1.2">Mobile Lookup</div>
-    <div style="font-size:.72rem;color:rgba(180,180,220,.35);margin-top:3px;letter-spacing:.02em">
-      DuckDB · HuggingFace Parquet
-    </div>
   </div>
 </div>
 """, unsafe_allow_html=True)
@@ -405,71 +402,62 @@ if not submitted or not number.strip():
 
 target = number.strip()
 
-# ── Skeleton while loading ──────────────────────────────────────────────────────
-
-sk_placeholder = st.empty()
-sk_placeholder.markdown(SKELETONS_HTML, unsafe_allow_html=True)
-
-# ── Index lookup ────────────────────────────────────────────────────────────────
-
-t0          = time.perf_counter()
-idx_results = query_both_parallel(target)
-idx_ms      = (time.perf_counter() - t0) * 1000
-
-m1_df, m1_ms = idx_results["mobile1"]
-m2_df, m2_ms = idx_results["mobile2"]
-
-if m1_df.empty and m2_df.empty:
-    sk_placeholder.empty()
-    st.markdown("""
-    <div style="background:rgba(251,191,36,.08);border:1px solid rgba(251,191,36,.25);
-      border-radius:10px;padding:12px 16px;color:#fbbf24;font-size:.9rem">
-      No records found for <strong>{}</strong>
-    </div>""".format(target), unsafe_allow_html=True)
-    st.stop()
-
-# ── Row-group fetch ─────────────────────────────────────────────────────────────
-
-fetch_tasks = []
-for src_col, src_df in [("mobile1", m1_df), ("mobile2", m2_df)]:
-    if src_df.empty:
-        continue
-    for _, rg in src_df[["filename","row_group_id","row_start","row_end"]].drop_duplicates().iterrows():
-        fetch_tasks.append({
-            "filename":   str(rg["filename"]),
-            "row_start":  int(rg["row_start"]),
-            "row_end":    int(rg["row_end"]),
-            "search_col": src_col,
-        })
+# ── Search with progress ────────────────────────────────────────────────────────
 
 all_dfs, errors = [], []
-with ThreadPoolExecutor(max_workers=min(8, len(fetch_tasks))) as ex:
-    futures = {
-        ex.submit(_fetch_one, t["filename"], t["row_start"], t["row_end"], t["search_col"], target): t
-        for t in fetch_tasks
-    }
-    for f in as_completed(futures):
-        try:
-            df, _ = f.result()
-            if not df.empty:
-                all_dfs.append(df)
-        except Exception as e:
-            errors.append(str(e))
-
-fetch_ms = (time.perf_counter() - t0) * 1000
-
-# Clear skeleton
-sk_placeholder.empty()
+t0 = time.perf_counter()
+
+with st.status("Searching...", expanded=True) as status:
+
+    # Step 1 — index lookup
+    st.write("Querying mobile1 & mobile2 indexes in parallel...")
+    idx_results  = query_both_parallel(target)
+    idx_ms       = (time.perf_counter() - t0) * 1000
+    m1_df, m1_ms = idx_results["mobile1"]
+    m2_df, m2_ms = idx_results["mobile2"]
+    st.write(f"Index lookup done — {fmt_ms(idx_ms)}")
+
+    if m1_df.empty and m2_df.empty:
+        status.update(label="No records found", state="error", expanded=False)
+        st.warning(f"No records found for **{target}**")
+        st.stop()
+
+    # Step 2 — build fetch tasks
+    fetch_tasks = []
+    for src_col, src_df in [("mobile1", m1_df), ("mobile2", m2_df)]:
+        if src_df.empty:
+            continue
+        for _, rg in src_df[["filename","row_group_id","row_start","row_end"]].drop_duplicates().iterrows():
+            fetch_tasks.append({
+                "filename":   str(rg["filename"]),
+                "row_start":  int(rg["row_start"]),
+                "row_end":    int(rg["row_end"]),
+                "search_col": src_col,
+            })
+
+    st.write(f"Fetching {len(fetch_tasks)} row group(s) from HuggingFace...")
+
+    with ThreadPoolExecutor(max_workers=min(8, len(fetch_tasks))) as ex:
+        futures = {
+            ex.submit(_fetch_one, t["filename"], t["row_start"], t["row_end"], t["search_col"], target): t
+            for t in fetch_tasks
+        }
+        for f in as_completed(futures):
+            try:
+                df, _ = f.result()
+                if not df.empty:
+                    all_dfs.append(df)
+            except Exception as e:
+                errors.append(str(e))
+
+    fetch_ms = (time.perf_counter() - t0) * 1000
+    status.update(label=f"Done — {fmt_ms(fetch_ms)}", state="complete", expanded=False)
 
 for err in errors:
     st.error(err)
 
 if not all_dfs:
-    st.markdown("""
-    <div style="background:rgba(251,191,36,.08);border:1px solid rgba(251,191,36,.25);
-      border-radius:10px;padding:12px 16px;color:#fbbf24;font-size:.9rem">
-      No matching rows found in remote parquet.
-    </div>""", unsafe_allow_html=True)
+    st.warning("No matching rows found in remote parquet.")
     st.stop()
 
 combined = pd.concat(all_dfs, ignore_index=True).drop_duplicates()