This document describes how to manually configure your system for running OpenStreetMap Carto. If you prefer quick, platform independent setup for a development environment, without the need to install and configure tools by hand, follow a Docker installation guide in DOCKER.md.
You need OpenStreetMap data loaded into a PostGIS database (see below for dependencies). These stylesheets expect a database generated with osm2pgsql using the pgsql backend (table names of planet_osm_point
, etc), the default database name (gis
), and the lua transforms documented in the instructions below.
Start by creating a database
sudo -u postgres createuser -s $USER
createdb gis
Enable PostGIS and hstore extensions with
psql -d gis -c 'CREATE EXTENSION postgis; CREATE EXTENSION hstore;'
then grab some OSM data. It's probably easiest to grab an PBF of OSM data from Geofabrik. Once you've done that, import with osm2pgsql:
osm2pgsql -G --hstore --style openstreetmap-carto.style --tag-transform-script openstreetmap-carto.lua -d gis ~/path/to/data.osm.pbf
You can find a more detailed guide to setting up a database and loading data with osm2pgsql at switch2osm.org.
Custom indexes are not required, but will speed up rendering, particularly for full planet databases, heavy load, or other production environments. They will not be as helpful with development using small extracts.
psql -d gis -f indexes.sql
Some features are rendered using preprocessed shapefiles.
To download them and import them into the database you can run the following script
scripts/get-external-data.py
This script downloads necessary files, generates and populates the data directory with all needed shapefiles, including indexing them through shapeindex.
The stylesheet uses Noto, an openly licensed font family from Google with support for multiple scripts. The stylesheet uses Noto's "Sans" style where available. If not available, this stylesheet uses another appropriate style of the Noto family. The "UI" version is used where available, with its vertical metrics which fit better with Latin text.
DejaVu Sans is used as an optional fallback font for systems without Noto Sans. If all the Noto fonts are installed, it should never be used. Noto Naskh Arabic UI is used an an optional fallback font for systems without Noto Sans Arabic.
Hanazono is used a fallback for seldom used CJK characters that are not covered by Noto.
Unifont is used as a last resort fallback, with it's excellent coverage, common presence on machines, and ugly look. For compatibility reasons, we support two Linux-distributions-specific versions of Unifont, therefor it's expected that you always get a warning about a missing Unifont version.
If you do not install all the fonts, the rendering itself will not break, but missing glyphs will be ugly.
For more details, see the documentation at fonts.mss.
On Ubuntu 16.04 or Debian Testing you can download and install most of the required fonts
sudo apt-get install fonts-noto-cjk fonts-noto-hinted fonts-noto-unhinted fonts-hanazono ttf-unifont
Noto Emoji Regular (not Noto Color Emoji) can be downloaded from the Noto Emoji repository.
It might be useful to have a more recent version of the fonts for rare non-latin scripts. The current upstream font release has also some more scripts and style variants than in the Ubuntu package. It can be installed from source.
DejaVu is packaged as fonts-dejavu-core
.
The fonts can be downloaded here:
After the download, you have to install the font files in the usual way of your operating system.
For proper rendering of non-latin scripts, particularly those with complicated diacritics and tone marks the requirements are
-
FreeType 2.6.2 or later for CJK characters
-
A recent enough version of Noto with coverage for the scripts needed.
For development, a style design studio is needed.
- Kosmtik - Kosmtik can be launched with
node index.js serve path/to/openstreetmap-carto/project.mml
TileMill is not officially supported, but you may be able to use a recent TileMill version by copying or symlinking the project directly into your Mapbox/project directory.
To display any map a database containing OpenStreetMap data and some utilities are required
- PostgreSQL
- PostGIS
- osm2pgsql to import your data into a PostGIS database
- Python 3 with the psycopg2, yaml, and requests libraries (
python3-psycopg2
python3-yaml
python3-requests
packages on Debian-derived systems) ogr2ogr
for loading shapefiles into the database (gdal-bin
on Debian-derived systems)
Some colours, SVGs and other files are generated with helper scripts. Not all users will need these dependencies
- Python and Ruby to run helper scripts
- Color Math and numpy if running generate_road_colors.py helper script (may be obtained with
pip install colormath numpy
)
For deployment, CartoCSS and Mapnik are required.
With CartoCSS you compile these sources into a Mapnik compatible XML file. When running CartoCSS, specify the Mapnik API version you are using (at least 3.0.0: carto -a "3.0.0"
).
If you're calling Mapnik in your own program, remember to load the XML file in non strict mode. This way, fonts declared with alternative names will only generate warnings, not errors. For instance, using the Python bindings, this becomes:
mapnik.load_map(mapnik.Map(width, height), xml_filename, False) # False for non-strict mode