Nominatim/settings/env.defaults
Sarah Hoffmann af968d4903 introduce tokenizer modules
This adds the boilerplate for selecting configurable tokenizers.
A tokenizer can be chosen at import time and will then install
itself such that it is fixed for the given database import even
when the software itself is updated.

The legacy tokenizer implements Nominatim's traditional algorithms.
2021-04-30 11:29:57 +02:00

218 lines
8.6 KiB
Plaintext

# .env
# Default configuration settings for Nominatim.
# This file uses the dotenv format.
# Database connection string.
# Add host, port, user etc through additional semicolon-separated attributes.
# e.g. ;host=...;port=...;user=...;password=...
# Changing this variable requires to run 'nominatim refresh --website'.
NOMINATIM_DATABASE_DSN="pgsql:dbname=nominatim"
# Database web user.
# Nominatim sets up read-only access for this user during installation.
NOMINATIM_DATABASE_WEBUSER="www-data"
# Directory where to find the PostgreSQL server module.
# When empty the module is expected to be located in the 'module' subdirectory
# in the project directory.
# Changing this value requires to run 'nominatim refresh --functions'.
NOMINATIM_DATABASE_MODULE_PATH=
# Tokenizer used for normalizing and parsing queries and names.
# The tokenizer is set up during import and cannot be changed afterwards
# without a reimport.
# Currently available tokenizers: legacy
NOMINATIM_TOKENIZER="legacy"
# Number of occurances of a word before it is considered frequent.
# Similar to the concept of stop words. Frequent partial words get ignored
# or handled differently during search.
# Changing this value requires a reimport.
NOMINATIM_MAX_WORD_FREQUENCY=50000
# If true, admin level changes on places with many contained children are blocked.
NOMINATIM_LIMIT_REINDEXING=yes
# Restrict search languages.
# Normally Nominatim will include all language variants of name:XX
# in the search index. Set this to a comma separated list of language
# codes, to restrict import to a subset of languages.
# Currently only affects the initial import of country names and special phrases.
NOMINATIM_LANGUAGES=
# Rules for normalizing terms for comparisons.
# The default is to remove accents and punctuation and to lower-case the
# term. Spaces are kept but collapsed to one standard space.
# Changing this value requires a reimport.
NOMINATIM_TERM_NORMALIZATION=":: NFD (); [[:Nonspacing Mark:] [:Cf:]] >; :: lower (); [[:Punctuation:][:Space:]]+ > ' '; :: NFC ();"
# Search in the Tiger house number data for the US.
# Note: The tables must already exist or queries will throw errors.
# Changing this value requires to run ./utils/setup --create-functions --setup-website.
NOMINATIM_USE_US_TIGER_DATA=no
# Search in the auxilary housenumber table.
# Changing this value requires to run ./utils/setup --create-functions --setup-website.
NOMINATIM_USE_AUX_LOCATION_DATA=no
# Proxy settings
# The following settings allow to set a proxy to use when remotely downloading
# data. Host and port are required. Login and password are optional.
NOMINATIM_HTTP_PROXY=no
NOMINATIM_HTTP_PROXY_HOST=proxy.mydomain.com
NOMINATIM_HTTP_PROXY_PORT=3128
NOMINATIM_HTTP_PROXY_LOGIN=
NOMINATIM_HTTP_PROXY_PASSWORD=
# Also set these standard environment variables.
# HTTP_PROXY="http://user:pass@10.10.1.10:1080"
# HTTPS_PROXY="http://user:pass@10.10.1.10:1080"
# Location of the osm2pgsql binary.
# When empty, osm2pgsql is expected to reside in the osm2pgsql directory in
# the project directory.
# EXPERT ONLY. You should usually use the supplied osm2pgsql.
NOMINATIM_OSM2PGSQL_BINARY=
# Directory where to find US Tiger data files to import.
# OBSOLETE: use `nominatim add-data --tiger-data <dir>` to explicitly state
# the directory on import
NOMINATIM_TIGER_DATA_PATH=
# Directory where to find pre-computed Wikipedia importance files.
# When unset, the data is expected to be located in the project directory.
NOMINATIM_WIKIPEDIA_DATA_PATH=
# Configuration file for special phrase import.
# When unset, the internal default settings from 'settings/phrase-settings.json'
# are used.
NOMINATIM_PHRASE_CONFIG=
# Configuration file for rank assignments.
# When unset, the internal default settings from 'settings/address-levels.json'
# are used.
NOMINATIM_ADDRESS_LEVEL_CONFIG=
# Configuration file for OSM data import.
# This may either be the name of one of an internal style or point
# to a file with a custom style.
# Internal styles are: admin, street, address, full, extratags
NOMINATIM_IMPORT_STYLE=extratags
# Location of the flatnode file used by osm2pgsql to store node locations.
# When unset, osm2pgsql stores the location in the PostgreSQL database. This
# is especially useful for imports of larger areas, like continents or the
# full planet. The file needs at least 70GB storage.
NOMINATIM_FLATNODE_FILE=
### Tablespace settings
#
# The following settings allow to move parts of the database tables into
# different tablespaces. This is especially interesting if you have disks
# with different speeds. When unset, the default tablespace is used.
# Only has an effect during import.
# Tablespace used for tables used when searching.
NOMINATIM_TABLESPACE_SEARCH_DATA=
# Tablespace used for indexes used when searching.
NOMINATIM_TABLESPACE_SEARCH_INDEX=
# Tablespace used for the OSM data cache tables. Used for import and update only.
NOMINATIM_TABLESPACE_OSM_DATA=
# Tablespace used for the OSM data cache indexes. Used for import and update only.
NOMINATIM_TABLESPACE_OSM_INDEX=
# Tablespace used for place import table. Used for import and update only.
NOMINATIM_TABLESPACE_PLACE_DATA=
# Tablespace used for place import indexes. Used for import and update only.
NOMINATIM_TABLESPACE_PLACE_INDEX=
# Tablespace for tables used during address computation. Used for import and update only.
NOMINATIM_TABLESPACE_ADDRESS_DATA=
# Tablespace for indexes used during address computation. Used for import and update only.
NOMINATIM_TABLESPACE_ADDRESS_INDEX=
# Tablespace for tables for auxilary data, e.g. TIGER data, postcodes.
NOMINATIM_TABLESPACE_AUX_DATA=
# Tablespace for indexes for auxilary data, e.g. TIGER data, postcodes.
NOMINATIM_TABLESPACE_AUX_INDEX=
### Replication settings
#
# The following settings control where and how updates for the database are
# retrieved.
#
#
# Base URL of replication service.
# A replication service provides change files of OSM data at regular intervals.
# These are used to keep the database up to date. Per default it points to
# the minutely updates for the main OSM database. There are other services
# geared towards larger update intervals or data extracts.
# Changing this value requires to rerun 'nominatim replication --init'.
NOMINATIM_REPLICATION_URL="https://planet.openstreetmap.org/replication/minute"
# Maximum amount of data to download per batch.
# Size is in MB.
NOMINATIM_REPLICATION_MAX_DIFF=50
# Publication interval of the replication service.
# Determines when Nominatim will attempt again to download again a new
# update. The time is computed from the publication date of the last diff
# downloaded. Setting this to a slightly higher value than the actual
# publication interval avoids unnecessary rechecks.
NOMINATIM_REPLICATION_UPDATE_INTERVAL=75
# Wait time to recheck for a pending update.
# Time to wait after an expected update was not available on the server.
NOMINATIM_REPLICATION_RECHECK_INTERVAL=60
### API settings
#
# The following settings configure the API responses. You must rerun
# 'nominatim refresh --website' after changing any of them.
# Send permissive CORS access headers.
# When enabled, send CORS headers to allow access to everybody.
NOMINATIM_CORS_NOACCESSCONTROL=yes
# URL for static icon images.
# Set this to the /mapicon directory of your nominatim-ui to enable returning
# icon URLs with the results.
NOMINATIM_MAPICON_URL=
# Language to assume when no particular language is requested.
# When unset, the local language (i.e. the name tag without suffix) will be used.
NOMINATIM_DEFAULT_LANGUAGE=
# Enable a special batch query mode.
# This features is currently undocumented and potentially broken.
NOMINATIM_SEARCH_BATCH_MODE=no
# Threshold for searches by name only.
# Threshold where the lookup strategy in the database is switched. If there
# are less occurences of a tem than given, the search does the lookup only
# against the name, otherwise it uses indexes for name and address.
NOMINATIM_SEARCH_NAME_ONLY_THRESHOLD=500
# Maximum number of OSM ids accepted by /lookup.
NOMINATIM_LOOKUP_MAX_COUNT=50
# Number of different geometry formats that may be queried in parallel.
# Set to zero to disable polygon output.
NOMINATIM_POLYGON_OUTPUT_MAX_TYPES=1
### Log settings
#
# The following options allow to enable logging of API requests.
# You must rerun 'nominatim refresh --website' after changing any of them.
#
# Enable logging of requests into the DB.
# The request will be logged into the new_query_log table.
# You should set up a cron job that regularly clears out this table.
NOMINATIM_LOG_DB=no
# Enable logging of requests into a file.
# To enable logging set this setting to the file to log to.
NOMINATIM_LOG_FILE=