This repository provides a simple script to generate TopoJSON files from MLIT of Japan's National Land Numerical Information download service.
See jp-atlas-example and bl.ocks.org/nobuf.
Approval to be obtained from the Geographical Survey Institute Head in order to reproduce this data. Acceptance and acknowledgment of [About National Land Numerical Information download service] The source of the data [digital national land information] and name of the parties processing the data is indicated.
git clone https://github.com/nobuf/jp-atlas.git
cd jp-atlas
# If you have the latest Docker:
env MLIT_GO_JP_ENDPOINT="http://nlftp.mlit.go.jp/...hidden..." \
TARGET_YEAR=2017 \
docker-compose up
# Or, if you have nodejs locally:
env MLIT_GO_JP_ENDPOINT="..." \
TARGET_YEAR=2017 \
npm install
It will create JSON files in the jp
folder.
TARGET_YEAR
could be 2017
, 2016
etc. Since MLIT_GO_JP_ENDPOINT
is probably not intended to receive programmatic access, you might want to download them manually. Place the downloaded zip file (ex. N03-170101_26_GML.zip
) as tmp/[2 digit prefecture id]_[target year].zip
. You'd need 47 files as Japan has 47 prefectures.
# jp/japan-2017-topo.json
A TopoJSON topology containing one geometry collection: cities. The geometry is quantized and simplified.
# jp/japan-2017-topo.low.json
Same as the above file, but for lower zoom level. While it has 1/3 of file size, it is not suitable for city-level zoom.
# jp.objects.cities
id
— 5 digits code that represents the city/town (市区町村); For instance,13104
is the code for Shinjuku, Tokyo; See also list-of-cities-in-japanproperties
— currently no property is included
If you have any ideas let us know by opening an issue. Pull requests are warmly welcome.
MIT
This repository's format is based on es-atlas and us-atlas :)