The ExtractCityData tool allows to process a 3DCityDB database in order to create a materialized view of buildings (encountered in the database) constituted by their id, their geometry and optionnally their year of construction and year of demolition.
Refer to the Install.md installation instructions
Activate venv if you created one:
. venv/bin/activate
You can run python extract_city_data.py -h to display help about this tool.
extract_city_data.py takes one mandatory argument and one optionnal argument.
The first one is a configuration file for the database such as ExtractCityData/db_config.yml:
PG_HOST: <server hosting your db>
PG_USER: <your db user name>
PG_PORT: <port>
PG_PASSWORD: <user password>
PG_NAME: <database name>
MATERIALIZED_VIEW_NAME: <name of the output materialized view>
You must fill these information. Note: MATERIALIZED_VIEW_NAME must start with
a letter.
It also takes an optionnal argument : -t or --temporal. If set, the materialized
view will also contain the years of constructions and years of demolitions of the
buildings.
Note: Once created, the view can be refreshed using REFRESH MATERIALIZED VIEW name.
More information
here
We had the choice between using a view, a materialized view or a new table (possibly in a new database) to store the output result.
We didn't choose a view because the output needs to be persitent in order to be used by another program (py3dtiles).
We chose a materialized view over a new table because the update process is easier and because we didn't want to modify the database schema of our 3DCityDB database. In addition, it is implemented in PostGIS since v9.3.