remove special casing for legacy tokenizer from BDD tests

[nominatim.git] / settings / env.defaults

diff --git a/settings/env.defaults b/settings/env.defaults
index 4069270eb73161073fba7ed0af593a8c3730e69c..8aee4a0fe28481dc25e182d2a0023e2ad593633e 100644 (file)
--- a/settings/env.defaults
+++ b/settings/env.defaults
@@ -18,7 +18,13 @@ NOMINATIM_DATABASE_WEBUSER="www-data"
 # Changing this value requires to run 'nominatim refresh --functions'.
 NOMINATIM_DATABASE_MODULE_PATH=
 
 # Changing this value requires to run 'nominatim refresh --functions'.
 NOMINATIM_DATABASE_MODULE_PATH=
 

-# Number of occurances of a word before it is considered frequent.
+# 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: icu, legacy
+NOMINATIM_TOKENIZER="icu"
+
+# Number of occurrences 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.
 # Similar to the concept of stop words. Frequent partial words get ignored
 # or handled differently during search.
 # Changing this value requires a reimport.


@@ -40,12 +46,18 @@ NOMINATIM_LANGUAGES=
 # Changing this value requires a reimport.
 NOMINATIM_TERM_NORMALIZATION=":: NFD (); [[:Nonspacing Mark:] [:Cf:]] >;  :: lower (); [[:Punctuation:][:Space:]]+ > ' '; :: NFC ();"
 
 # Changing this value requires a reimport.
 NOMINATIM_TERM_NORMALIZATION=":: NFD (); [[:Nonspacing Mark:] [:Cf:]] >;  :: lower (); [[:Punctuation:][:Space:]]+ > ' '; :: NFC ();"
 

+# Configuration file for the tokenizer.
+# The content depends on the tokenizer used. If left empty the default settings
+# for the chosen tokenizer will be used. The configuration can only be set
+# on import and not be changed afterwards.
+NOMINATIM_TOKENIZER_CONFIG=
+

 # 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 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.
+# Search in the auxiliary housenumber table.

 # Changing this value requires to run ./utils/setup --create-functions --setup-website.
 NOMINATIM_USE_AUX_LOCATION_DATA=no
 
 # Changing this value requires to run ./utils/setup --create-functions --setup-website.
 NOMINATIM_USE_AUX_LOCATION_DATA=no
 


@@ -53,7 +65,7 @@ NOMINATIM_USE_AUX_LOCATION_DATA=no
 # 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
 # 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_HOST=

 NOMINATIM_HTTP_PROXY_PORT=3128
 NOMINATIM_HTTP_PROXY_LOGIN=
 NOMINATIM_HTTP_PROXY_PASSWORD=
 NOMINATIM_HTTP_PROXY_PORT=3128
 NOMINATIM_HTTP_PROXY_LOGIN=
 NOMINATIM_HTTP_PROXY_PASSWORD=


@@ -62,29 +74,15 @@ NOMINATIM_HTTP_PROXY_PASSWORD=
 # HTTPS_PROXY="http://user:pass@10.10.1.10:1080"
 
 # Location of the osm2pgsql binary.
 # 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=
 
 # 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=
 
 # 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.
 # Configuration file for rank assignments.

-# When unset, the internal default settings from 'settings/address-levels.json'
-# are used.
-NOMINATIM_ADDRESS_LEVEL_CONFIG=
+NOMINATIM_ADDRESS_LEVEL_CONFIG=address-levels.json

 
 # Configuration file for OSM data import.
 # This may either be the name of one of an internal style or point
 
 # Configuration file for OSM data import.
 # This may either be the name of one of an internal style or point


@@ -125,9 +123,9 @@ NOMINATIM_TABLESPACE_ADDRESS_DATA=
 # Tablespace for indexes used during address computation. Used for import and update only.
 NOMINATIM_TABLESPACE_ADDRESS_INDEX=
 
 # 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.
+# Tablespace for tables for auxiliary data, e.g. TIGER data, postcodes.

 NOMINATIM_TABLESPACE_AUX_DATA=
 NOMINATIM_TABLESPACE_AUX_DATA=

-# Tablespace for indexes for auxilary data, e.g. TIGER data, postcodes.
+# Tablespace for indexes for auxiliary data, e.g. TIGER data, postcodes.

 NOMINATIM_TABLESPACE_AUX_INDEX=
 
 
 NOMINATIM_TABLESPACE_AUX_INDEX=
 
 


@@ -150,14 +148,14 @@ NOMINATIM_REPLICATION_URL="https://planet.openstreetmap.org/replication/minute"
 # Size is in MB.
 NOMINATIM_REPLICATION_MAX_DIFF=50
 
 # Size is in MB.
 NOMINATIM_REPLICATION_MAX_DIFF=50
 

-# Publication interval of the replication service.
+# Publication interval of the replication service (in seconds).

 # 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
 
 # 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.
+# Wait time to recheck for a pending update (in seconds).

 # Time to wait after an expected update was not available on the server.
 NOMINATIM_REPLICATION_RECHECK_INTERVAL=60
 
 # Time to wait after an expected update was not available on the server.
 NOMINATIM_REPLICATION_RECHECK_INTERVAL=60
 


@@ -179,16 +177,6 @@ NOMINATIM_MAPICON_URL=
 # When unset, the local language (i.e. the name tag without suffix) will be used.
 NOMINATIM_DEFAULT_LANGUAGE=
 
 # 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
 
 # Maximum number of OSM ids accepted by /lookup.
 NOMINATIM_LOOKUP_MAX_COUNT=50
 


@@ -196,6 +184,32 @@ NOMINATIM_LOOKUP_MAX_COUNT=50
 # Set to zero to disable polygon output.
 NOMINATIM_POLYGON_OUTPUT_MAX_TYPES=1
 
 # Set to zero to disable polygon output.
 NOMINATIM_POLYGON_OUTPUT_MAX_TYPES=1
 

+# Offer backwards compatible PHP URLs.
+# When running one of the Python enignes, they will add endpoint aliases
+# under <endpoint>.php
+NOMINATIM_SERVE_LEGACY_URLS=yes
+
+# Maximum number of connection a single API object can use. (Python API only)
+# When running Nominatim as a server, then this is the maximum number
+# of connections _per worker_.
+NOMINATIM_API_POOL_SIZE=10
+
+# Timeout is seconds after which a single query to the database is cancelled.
+# The user receives a 503 response, when a query times out.
+# When empty, then timeouts are disabled.
+NOMINATIM_QUERY_TIMEOUT=10
+
+# Maximum time a single request is allowed to take. When the timeout is
+# exceeded, the available results are returned.
+# When empty, then timeouts are disabled.
+NOMINATIM_REQUEST_TIMEOUT=60
+
+# Search elements just within countries
+# If, despite not finding a point within the static grid of countries, it
+# finds a geometry of a region, do not return the geometry. Return "Unable
+# to geocode" instead.
+NOMINATIM_SEARCH_WITHIN_COUNTRIES=False
+

 ### Log settings
 #
 # The following options allow to enable logging of API requests.
 ### Log settings
 #
 # The following options allow to enable logging of API requests.


@@ -209,3 +223,7 @@ 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=
 # Enable logging of requests into a file.
 # To enable logging set this setting to the file to log to.
 NOMINATIM_LOG_FILE=

+
+# Echo raw SQL from SQLAlchemy statements.
+# EXPERT: Works only in command line/library use.
+NOMINATIM_DEBUG_SQL=no