ASCII Map Renderer

ascii_map_renderer.py renders the prefecture, city, and ward boundaries that intersect the visible map range, with the prefecture line brightest, the city line medium, and the ward line dimmest, over a muted JMA rain-radar background. The default center is Tokyo.

Run

pip install -r requirements.txt
python .\ascii_map_renderer.py

python .\ascii_map_renderer.py --help shows the CLI help without starting a render.

Options

• --lat N: map center latitude, default Tokyo
• --lon N: map center longitude, default Tokyo
• --marker-lat N: marker latitude, default Tokyo
• --marker-lon N: marker longitude, default Tokyo
• --cols N: map width in characters
• --rows N: map height in characters
• --scale M: meters per character cell, default 150
• --cell-aspect R: terminal cell height/width ratio, default 2.0
• --auto-cell-aspect: detect the terminal cell height/width ratio instead of using the fixed default
• --refresh-seconds N: redraw interval for the rain radar background, default 300
• --once: render once and exit
• --no-color: disable ANSI colors

Notes

• The script fetches boundary GeoJSON from OpenStreetMap Nominatim on first run.
• The script fetches only administrative boundary geometries with admin_level 4, 7, or 8 from the OpenStreetMap Overpass API for the current viewport.
• The script fetches JMA nowcast hrpns radar tiles and uses them as the background color field, dimmed slightly so the boundary strokes remain visible.
• Radar data is kept in memory for the lifetime of the process and refreshed periodically; it is not written to disk.
• The static map layers are built once per process and reused across radar refreshes.
• The top status line is rendered in Japanese as + 現在地：{Prefecture} {City} {Ward} | {天気} | 降水量：{降水量} | 風：{風向}{風速（m/s）} | 最終更新：{日時}.
• Downloaded boundary data is cached in cache/ next to the script.
• Boundary strokes are chosen by bitmap similarity against a monospace font atlas, so the renderer can pick from box drawing, punctuation, digits, and letters when that better matches the local boundary shape.
• The implementation uses a WGS84 ECEF/ENU local projection for coordinate placement, so the marker and boundaries share the same precise map space.
• Terminal cell aspect is applied to the geographic projection and the glyph raster height together, so the visible map stays proportioned correctly on tall terminal cells.
• The default cell aspect is fixed at 2.0 because automatic detection is environment-dependent; use --auto-cell-aspect only if you want to try detection.
• The implementation uses Pillow and numpy rather than a dedicated terminal canvas library because the goal here is glyph-shape matching, not just generic block rendering.
• A green + marker is placed at the configured marker point, which defaults to Tokyo.