Update app.py

app.py

@@ -49,4 +49,213 @@ def load_gdf_from_zip(zip_path: str) -> gpd.GeoDataFrame:
 
 def make_geocoder():
     geolocator = Nominatim(user_agent=USER_AGENT, timeout=10)
-    geocode = RateL
+    geocode = RateLimiter(geolocator.geocode, min_delay_seconds=GEOCODE_DELAY_SEC)
+    return geocode
+
+def geocode_with_cache(addresses, CFs, use_internet=True):
+    cache = load_cache()
+    cache_map = {row["address_input"]: (row["lat"], row["lon"], row["CF"]) for _, row in cache.iterrows()}
+    results = []
+    geocode = make_geocoder() if use_internet else None
+
+    for a, cf in zip(addresses, CFs):
+        a = str(a)
+        cf = "" if (cf is None or (isinstance(cf, float) and np.isnan(cf))) else str(cf)
+
+        # cache hit
+        if a in cache_map:
+            lat, lon, _cached_cf = cache_map[a]
+            if pd.notna(lat) and pd.notna(lon):
+                results.append({"address_input": a, "CF": cf, "lat": lat, "lon": lon})
+                continue
+
+        if not use_internet:
+            results.append({"address_input": a, "CF": cf, "lat": np.nan, "lon": np.nan})
+            continue
+
+        try:
+            loc = geocode(a, country_codes="jp", addressdetails=True)
+            if loc:
+                lat, lon = loc.latitude, loc.longitude
+            else:
+                lat, lon = np.nan, np.nan
+        except Exception:
+            lat, lon = np.nan, np.nan
+
+        cache = cache[cache["address_input"] != a]
+        cache = pd.concat(
+            [cache, pd.DataFrame([{"address_input": a, "lat": lat, "lon": lon, "CF": cf}])],
+            ignore_index=True
+        )
+        save_cache(cache)
+        results.append({"address_input": a, "CF": cf, "lat": lat, "lon": lon})
+
+    return pd.DataFrame(results)
+
+def plot_map_png(gdf_pref: gpd.GeoDataFrame, gdf_pts: gpd.GeoDataFrame,
+                 line_width: float = 0.6, figsize=(7, 7)) -> Image.Image:
+    fig, ax = plt.subplots(figsize=figsize)
+    gdf_pref.boundary.plot(ax=ax, linewidth=line_width, color="black")
+
+    gdf_pts_valid = gdf_pts[gdf_pts.geometry.notnull()]
+    if not gdf_pts_valid.empty:
+        cf_num = pd.to_numeric(gdf_pts_valid.get("CF", pd.Series([np.nan]*len(gdf_pts_valid))), errors="coerce")
+        gdf_pts_valid.assign(CF_num=cf_num).plot(
+            ax=ax, column="CF_num", cmap="OrRd", markersize=12, alpha=0.85, legend=True
+        )
+
+    ax.set_axis_off()
+    plt.tight_layout()
+    buf = io.BytesIO()
+    fig.savefig(buf, format="png", dpi=200)
+    plt.close(fig)
+    buf.seek(0)
+    return Image.open(buf)
+
+def make_folium_html(gdf_pref: gpd.GeoDataFrame, gdf_pts: gpd.GeoDataFrame):
+    gdf_pts_valid = gdf_pts[gdf_pts.geometry.notnull()]
+    if not gdf_pts_valid.empty:
+        center_lat = gdf_pts_valid.geometry.y.median()
+        center_lon = gdf_pts_valid.geometry.x.median()
+        zoom = 6
+    else:
+        center_lat, center_lon, zoom = 35.6812, 139.7671, 5
+
+    m = folium.Map(location=[center_lat, center_lon], zoom_start=zoom)
+
+    try:
+        folium.GeoJson(gdf_pref.to_json(), name="prefecture").add_to(m)
+    except Exception:
+        pass
+
+    for _, r in gdf_pts_valid.iterrows():
+        lat, lon = r.geometry.y, r.geometry.x
+        popup = f"{r.get('address_input','(no addr)')}<br>CF:{r.get('CF','')}"
+        folium.CircleMarker(
+            location=(float(lat), float(lon)),
+            radius=4,
+            fill=True,
+            fill_opacity=0.9,
+            popup=popup,
+        ).add_to(m)
+
+    return m._repr_html_()
+
+def _parse_indexer(x):
+    try:
+        return int(x)
+    except Exception:
+        return x
+
+def run(zip_file, excel_file, sheet_name, header_row, address_col, power_col,
+        use_inet, line_width):
+    # 1) Shapefile
+    if zip_file is not None and hasattr(zip_file, "name") and os.path.exists(zip_file.name):
+        zip_path = zip_file.name
+    elif os.path.exists(DEFAULT_ZIP):
+        zip_path = DEFAULT_ZIP
+    else:
+        empty_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
+        return None, None, "", empty_df, "Shapefile の ZIP をアップロードするか、data/japan_ver85.zip を配置してください。"
+
+    try:
+        gdf_pref = load_gdf_from_zip(zip_path)
+    except Exception as e:
+        empty_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
+        return None, None, "", empty_df, f"行政界の読み込みに失敗しました: {e}"
+
+    # 2) Excel→ジオコーディング
+    if excel_file is None or not hasattr(excel_file, "name"):
+        # 空のテーブルでも返す
+        gdf_pts = gpd.GeoDataFrame(columns=["address_input", "CF", "lat", "lon"], geometry=[], crs="EPSG:4326")
+        table_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
+    else:
+        try:
+            df = pd.read_excel(excel_file.name, sheet_name=sheet_name, header=int(header_row))
+        except Exception as e:
+            empty_df = pd.DataFrame(columns=["address_input", "CF", "lat", "lon"])
+            return None, None, "", empty_df, f"Excel の読み込みに失敗しました: {e}"
+
+        addr_series = df.iloc[:, address_col] if isinstance(address_col, int) else df[address_col]
+        cf_series = df.iloc[:, power_col] if isinstance(power_col, int) else df[power_col]
+
+        addresses = addr_series.astype(str).tolist()
+        cfs = cf_series.tolist()
+
+        geo_df = geocode_with_cache(addresses, cfs, use_internet=bool(use_inet))
+        table_df = geo_df[["address_input", "CF", "lat", "lon"]].copy()
+
+        geometry = [
+            Point(lon, lat) if (pd.notna(lat) and pd.notna(lon)) else None
+            for lat, lon in zip(geo_df["lat"], geo_df["lon"])
+        ]
+        gdf_pts = gpd.GeoDataFrame(geo_df, geometry=geometry, crs="EPSG:4326")
+
+    # 3) 図と地図
+    try:
+        img = plot_map_png(gdf_pref, gdf_pts, line_width=line_width)
+    except Exception as e:
+        return None, None, "", table_df, f"静的描画に失敗しました: {e}"
+
+    try:
+        html = make_folium_html(gdf_pref, gdf_pts)
+    except Exception as e:
+        html = f"<p>folium描画に失敗しました: {e}</p>"
+
+    # 4) 情報
+    info = []
+    info.append(f"都道府県レコード数: {len(gdf_pref)}")
+    if gdf_pref.crs:
+        info.append(f"PREF CRS: {gdf_pref.crs}")
+    info.append(f"ポイント数（有効座標）: {int(gdf_pts.geometry.notnull().sum())} / {len(gdf_pts)}")
+    if not gdf_pts.empty and gdf_pts.crs:
+        info.append(f"PTS CRS: {gdf_pts.crs}")
+
+    return img, html, "\n".join(info), table_df, ""
+
+with gr.Blocks(title="Japan Shapefile + Excel Geocoding Plotter") as demo:
+    gr.Markdown("## japan_ver85.shp（ZIP） + Excel住所 → 日本地図にプロット ＆ 表で確認")
+
+    with gr.Row():
+        zip_in = gr.File(label="Shapefile (ZIP)", file_count="single", file_types=[".zip"])
+        xlsx_in = gr.File(label="Excelファイル（住所付き）", file_count="single", file_types=[".xlsx", ".xls"])
+
+    with gr.Row():
+        sheet = gr.Textbox(label="シート名", value="認定設備")
+        header_row = gr.Number(label="ヘッダー行番号（0始まり）", value=2, precision=0)
+
+    with gr.Row():
+        address_col = gr.Textbox(label="住所列（列名 or 0始まり列番号）", value="発電設備の所在地")
+        power_col = gr.Textbox(label="数値列（任意：列名 or 0始まり列番号）", value="発電出力（kW）")
+
+    with gr.Row():
+        use_inet = gr.Checkbox(label="Nominatimに問い合わせ（オフでキャッシュのみ使用）", value=True)
+        line_width = gr.Slider(0.2, 2.0, value=0.6, step=0.1, label="境界線の太さ")
+
+    run_btn = gr.Button("描画")
+
+    out_img = gr.Image(label="静的地図（matplotlib）", type="pil")
+    out_html = gr.HTML(label="インタラクティブ地図（folium）")
+    out_info = gr.Textbox(label="メタ情報", lines=4)
+    out_table = gr.Dataframe(label="ジオコーディング結果（住所・緯度・経度・CF）", wrap=True)
+    out_err = gr.Markdown(label="エラー", visible=True)
+
+    def _parse(x):
+        try:
+            return int(x)
+        except Exception:
+            return x
+
+    def app_run(zipf, xls, s, h, a, p, inet, lw):
+        return run(
+            zipf, xls, s, int(h), _parse(a), _parse(p), inet, lw
+        )
+
+    run_btn.click(
+        fn=app_run,
+        inputs=[zip_in, xlsx_in, sheet, header_row, address_col, power_col, use_inet, line_width],
+        outputs=[out_img, out_html, out_info, out_table, out_err],
+    )
+
+if __name__ == "__main__":
+    demo.launch()