prepare for 3.3.1 release

The class parameter was used as is, allowing for potential
SQL injection via the API.

Thanks to @bladeswords for finding this.

.travis.yml

@@ -1,6 +1,6 @@
 ---
 sudo: required
-dist: trusty
+dist: xenial
 language: python
 python:
   - "3.6"
@@ -11,21 +11,24 @@ git:
 env:
   - TEST_SUITE=tests
   - TEST_SUITE=monaco
+before_install:
+  - phpenv global 7.1
 install:
   - vagrant/install-on-travis-ci.sh
 before_script:
   - psql -U postgres -c "create extension postgis"
 script:
-  - cd $TRAVIS_BUILD_DIR/build
-  - if [[ $TEST_SUITE == "monaco" ]]; then wget --no-verbose --output-document=../data/monaco.osm.pbf http://download.geofabrik.de/europe/monaco-latest.osm.pbf; fi
-  - if [[ $TEST_SUITE == "monaco" ]]; then ./utils/setup.php --osm-file ../data/monaco.osm.pbf --osm2pgsql-cache 1000 --all 2>&1 | grep -v 'ETA (seconds)'; fi
-  - if [[ $TEST_SUITE == "monaco" ]]; then ./utils/specialphrases.php --wiki-import | psql -d test_api_nominatim >/dev/null; fi
+  - cd $TRAVIS_BUILD_DIR/
+  - if [[ $TEST_SUITE == "tests" ]]; then phpcs --report-width=120 . ; fi
   - cd $TRAVIS_BUILD_DIR/test/php
-  - if [[ $TEST_SUITE == "tests" ]]; then phpunit ./ ; fi
-  - if [[ $TEST_SUITE == "tests" ]]; then phpcs --report-width=120 */**.php ; fi
+  - if [[ $TEST_SUITE == "tests" ]]; then /usr/bin/phpunit ./ ; fi
   - cd $TRAVIS_BUILD_DIR/test/bdd
   - # behave --format=progress3 api
-  - if [[ $TEST_SUITE == "tests" ]]; then behave --format=progress3 db ; fi
+  - if [[ $TEST_SUITE == "tests" ]]; then behave -DREMOVE_TEMPLATE=1 --format=progress3 db ; fi
   - if [[ $TEST_SUITE == "tests" ]]; then behave --format=progress3 osm2pgsql ; fi
+  - cd $TRAVIS_BUILD_DIR/build
+  - if [[ $TEST_SUITE == "monaco" ]]; then wget --no-verbose --output-document=../data/monaco.osm.pbf http://download.geofabrik.de/europe/monaco-latest.osm.pbf; fi
+  - if [[ $TEST_SUITE == "monaco" ]]; then /usr/bin/env php ./utils/setup.php --osm-file ../data/monaco.osm.pbf --osm2pgsql-cache 1000 --all 2>&1 | grep -v 'ETA (seconds)'; fi
+  - if [[ $TEST_SUITE == "monaco" ]]; then /usr/bin/env php ./utils/specialphrases.php --wiki-import | psql -d test_api_nominatim >/dev/null; fi
 notifications:
   email: false

AUTHORS

@@ -3,16 +3,13 @@ Nominatim was written by:
   Brian Quinion
   Sarah Hoffmann
   Marc Tobias Metten
+
   markigail
+  gemo1011
   IrlJidel
   Frederik Ramm
-  Michael Spreng
-  Daniele Forsi
-  mfn
-  Grant Slater
-  Andree Klattenhoff
-  appelflap
-  b3nn0
-  Spin0us
-  Kurt Roeckx
-  Rodolphe Quiédeville
+
+and many more.
+
+For a full list of contributors see
+https://github.com/openstreetmap/Nominatim/graphs/contributors

CMakeLists.txt

@@ -19,8 +19,8 @@ list(APPEND CMAKE_MODULE_PATH "${CMAKE_SOURCE_DIR}/cmake")
 project(nominatim)
 
 set(NOMINATIM_VERSION_MAJOR 3)
-set(NOMINATIM_VERSION_MINOR 1)
-set(NOMINATIM_VERSION_PATCH 0)
+set(NOMINATIM_VERSION_MINOR 3)
+set(NOMINATIM_VERSION_PATCH 1)
 
 set(NOMINATIM_VERSION "${NOMINATIM_VERSION_MAJOR}.${NOMINATIM_VERSION_MINOR}.${NOMINATIM_VERSION_PATCH}")
 
@@ -53,7 +53,7 @@ link_directories(${PostgreSQL_LIBRARY_DIRS})
 
 find_program(PYOSMIUM pyosmium-get-changes)
 if (NOT EXISTS "${PYOSMIUM}")
-    set(PYOSMIUM_PATH "/nonexistent")
+    set(PYOSMIUM_PATH "")
         message(WARNING "pyosmium-get-changes not found (required for updates)")
 else()
     set(PYOSMIUM_PATH "${PYOSMIUM}")
@@ -77,14 +77,23 @@ find_package(BZip2 REQUIRED)
 find_package(LibXml2 REQUIRED)
 include_directories(${LIBXML2_INCLUDE_DIR})
 
+# Setting PHP binary variable as to command line (prevailing) or auto detect
+if (NOT PHP_BIN)
+     find_program (PHP_BIN php)
+endif()
+# sanity check if PHP binary exists
+if (NOT EXISTS ${PHP_BIN})
+    message(FATAL_ERROR "PHP binary not found. Install php or provide location with -DPHP_BIN=/path/php ")
+endif()
+message (STATUS "Using PHP binary " ${PHP_BIN})
+
 #-----------------------------------------------------------------------------
 #
 # Setup settings and paths
 #
 #-----------------------------------------------------------------------------
 
-set(CUSTOMFILES
-    settings/phrase_settings.php
+set(WEBSITESCRIPTS
     website/deletable.php
     website/details.php
     website/hierarchy.php
@@ -93,23 +102,31 @@ set(CUSTOMFILES
     website/reverse.php
     website/search.php
     website/status.php
-    utils/blocks.php
+)
+
+set(CUSTOMSCRIPTS
     utils/country_languages.php
-    utils/imports.php
     utils/importWikipedia.php
+    utils/export.php
     utils/query.php
-    utils/server_compare.php
     utils/setup.php
     utils/specialphrases.php
     utils/update.php
     utils/warm.php
    )
 
-foreach (cfile ${CUSTOMFILES})
-    configure_file(${PROJECT_SOURCE_DIR}/${cfile} ${PROJECT_BINARY_DIR}/${cfile})
+foreach (script_source ${CUSTOMSCRIPTS})
+    configure_file(${PROJECT_SOURCE_DIR}/cmake/script.tmpl
+                   ${PROJECT_BINARY_DIR}/${script_source})
 endforeach()
 
-configure_file(${PROJECT_SOURCE_DIR}/settings/defaults.php ${PROJECT_BINARY_DIR}/settings/settings.php)
+foreach (script_source ${WEBSITESCRIPTS})
+    configure_file(${PROJECT_SOURCE_DIR}/cmake/website.tmpl
+                   ${PROJECT_BINARY_DIR}/${script_source})
+endforeach()
+
+configure_file(${PROJECT_SOURCE_DIR}/settings/defaults.php
+               ${PROJECT_BINARY_DIR}/settings/settings.php)
 
 set(WEBPATHS css images js)
 

CONTRIBUTING.md

@@ -41,7 +41,7 @@ Please add the following information to your issue:
 
 ## Workflow for Pull Requests
 
-We love to get pull reuqests from you. We operate the "Fork & Pull" model
+We love to get pull requests from you. We operate the "Fork & Pull" model
 explained at
 
 https://help.github.com/articles/using-pull-requests
@@ -65,7 +65,7 @@ that duplicate work can be avoided.
 ## Coding style
 
 Nominatim historically hasn't followed a particular coding style but we
-are in process of consolodating the style. The following rules apply:
+are in process of consolidating the style. The following rules apply:
 
  * Python code uses the official Python style
  * indention
@@ -78,11 +78,13 @@ are in process of consolodating the style. The following rules apply:
    * no spaces after opening and before closing bracket
    * leave out space between a function name and bracket
      but add one between control statement(if, while, etc.) and bracket
+ * for PHP variables use CamelCase with a prefixing letter indicating the type
+   (i - integer, f - float, a - array, s - string, o - object)
 
 The coding style is enforced with PHPCS and can be tested with:
 
 ```
-  phpcs --report-width=120 --colors */**.php
+  phpcs --report-width=120 --colors .
 ```
 
 ## Testing

ChangeLog

@@ -1,3 +1,56 @@
+3.3.1
+ * security fix: fix possible SQL injection via details API
+
+3.3.0
+
+ * zoom 17 in reverse now zooms in on minor streets
+ * fix use of postcode relations in address
+ * support for housenumber 0 on interpolations
+ * replace database abstraction DB with PDO and switch to using exceptions
+ * exclude line features at rank 30 from reverse geocoding
+ * remove self-reference and country from place_addressline
+ * make json output more readable (less escaping)
+ * update conversion scripts for postcodes
+ * scripts in utils/ are no longer executable (always use scripts in build dir)
+ * remove Natural Earth country fallback (OSM is complete enough)
+ * make rank assignments configurable
+ * allow accept languages with underscore
+ * new reverse-only import mode (without search index table)
+ * rely on boundaries only for states and countries
+ * update osm2pgsql, now using a configurable style
+ * provide multiple import styles
+ * improve search when house number and postcodes are dropped
+ * overhaul of setup code
+ * add support for PHPUnit 6
+ * update test database
+ * various documentation improvements
+
+3.2.0
+
+ * complete rewrite of reverse search algorithm
+ * add new geojson and geocodejson output formats
+ * add simple export script to exprot addresses to CSV
+ * remove is_in terms from address computation
+ * remove unused search_name_country tables
+ * various smaller fixes to query parsing
+ * convert Tokens and token types to class types
+ * correctly handle update when boundary object type is changed
+ * improve debug output for /search endpoint
+ * update to latest osm2pgsql and leaflet.js
+ * overhaul of /details endpoint:
+     * new class parameter when using osmtype/osmid parameters
+     * permalink to instance-independent osmtype/osmid parameter format
+     * new json output format
+ * update CentOS vagrant machine to use SELinux
+ * add vagrant scripts for Ubuntu 18.04
+ * fix build process for BSD
+ * enable running the database on a different host than the setup scripts
+ * allow to configure use of custom PHP binaries (PHP_BIN)
+ * extensive coding style improvements to PHP code
+ * more PHP unit tests for new classes
+ * increase coverage for API tests
+ * add documentation for API
+
 3.1.0
 
  * rework postcode handling and introduce location_postcode table

README.md

@@ -7,25 +7,24 @@ Nominatim (from the Latin, 'by name') is a tool to search OpenStreetMap data
 by name and address (geocoding) and to generate synthetic addresses of
 OSM points (reverse geocoding). An instance with up-to-date data can be found
 at https://nominatim.openstreetmap.org. Nominatim is also used as one of the
-sources for the Search box on the OpenStreetMap home page and powers the search
-on the MapQuest Open Initiative websites.
+sources for the Search box on the OpenStreetMap home page.
 
 Documentation
 =============
 
-More information about Nominatim, including usage and installation instructions,
-can be found in the docs/ subdirectory and in the OSM wiki at:
-
-https://nominatim.org
+The documentation of the latest development version is in the
+`docs/` subdirectory. A HTML version can be found at
+https://nominatim.org/release-docs/develop/ .
 
 Installation
 ============
 
 The latest stable release can be downloaded from https://nominatim.org.
-There you can also find [installation instructions for the release](https://nominatim.org/release-docs/latest/Installation).
+There you can also find [installation instructions for the release](https://nominatim.org/release-docs/latest/admin/Installation).
 
 Detailed installation instructions for the development version can be
-found in the `/docs` directory, see [docs/Installation.md](docs/Installation.md).
+found at [nominatim.org](https://nominatim.org/release-docs/develop/admin/Installation)
+as well.
 
 A quick summary of the necessary steps:
 
@@ -48,11 +47,17 @@ License
 
 The source code is available under a GPLv2 license.
 
-Contact and Bug reports
-======================
 
-For questions you can join the geocoding mailinglist, see
+Contributing
+============
+
+Contributions are welcome. For details see [contribution guide](CONTRIBUTING.md).
+
+Both bug reports and pull requests are welcome.
+
+
+Mailing list
+============
+
+For questions you can join the geocoding mailing list, see
 https://lists.openstreetmap.org/listinfo/geocoding
-
-Bugs may be reported on the github project site:
-https://github.com/openstreetmap/Nominatim

VAGRANT.md

@@ -1,11 +1,11 @@
 # Install Nominatim in a virtual machine for development and testing
 
-This document describes how you can install Nominatim inside a Ubuntu 14
+This document describes how you can install Nominatim inside a Ubuntu 16
 virtual machine on your desktop/laptop (host machine). The goal is to give
 you a development environment to easily edit code and run the test suite
 without affecting the rest of your system. 
 
-The installation can run largely unsupervised. You should expect 1-2h from
+The installation can run largely unsupervised. You should expect 1h from
 start to finish depending on how fast your computer and download speed
 is.
 
@@ -19,7 +19,7 @@ is.
 
         git clone --recursive https://github.com/openstreetmap/Nominatim.git
 
-    If you haven't used `--recursive`, then you can load the submodules using
+    If you forgot `--recursive`, it you can later load the submodules using
     
         git submodule init
         git submodule update
@@ -37,18 +37,14 @@ is.
         vagrant ssh ubuntu
 
 3. Import a small country (Monaco)
-
-    You need to give the virtual machine more memory (2GB) for an import,
-    see `Vagrantfile`. Otherwise 1GB is enough.
     
     See the FAQ how to skip this step and point Nominatim to an existing database.
 
       ```
       # inside the virtual machine:
-      mkdir data
       cd build
-      wget --no-verbose --output-document=../data/monaco.osm.pbf http://download.geofabrik.de/europe/monaco-latest.osm.pbf
-      ./utils/setup.php --osm-file ../data/monaco.osm.pbf --osm2pgsql-cache 1000 --all 2>&1 | tee monaco.$$.log
+      wget --no-verbose --output-document=/tmp/monaco.osm.pbf http://download.geofabrik.de/europe/monaco-latest.osm.pbf
+      ./utils/setup.php --osm-file /tmp/monaco.osm.pbf --osm2pgsql-cache 1000 --all 2>&1 | tee monaco.$$.log
       ```
 
     To repeat an import you'd need to delete the database first
@@ -65,47 +61,64 @@ see Nominatim in action on [locahost:8089](http://localhost:8089/nominatim/).
 You edit code on your host machine in any editor you like. There is no need to
 restart any software: just refresh your browser window.
 
+Note that the webserver uses files from the /build directory. If you change
+files in Nominatim/website or Nominatim/utils for example you first need to
+copy them into the /build directory by running the `cmake` step from the
+installation.
+
 PHP errors are written to `/var/log/apache2/error.log`.
 
 With `echo` and `var_dump()` you write into the output (HTML/XML/JSON) when
 you either add `&debug=1` to the URL (preferred) or set
 `@define('CONST_Debug', true);` in `settings/local.php`.
 
+In the Python BDD test you can use `logger.info()` for temporary debug
+statements.
 
 
 
-## Running functional tests
-
-Tests in `/features/db` and `/features/osm2pgsql` have to pass 100%. Other
-tests might require full planet-wide data. Sadly even if you have your own
-planet-wide data there will be enough differences to the openstreetmap.org
-installation to cause false positives in the other tests (see FAQ). 
-
-To run the full test suite
-
-    cd ~/Nominatim/tests
-    NOMINATIM_SERVER=http://localhost:8089/nominatim lettuce features
-
-To run a single file
-
-    NOMINATIM_SERVER=http://localhost:8089/nominatim lettuce features/api/reverse.feature
-    
-To run specific tests you can add tags just before the `Scenario line`, e.g.
-
-    @bug-34
-    Scenario: address lookup for non-existing or invalid node, way, relation
-
-and then
-
-    NOMINATIM_SERVER=http://localhost:8089/nominatim lettuce -t bug-34
-
-
 ## Running unit tests
 
     cd ~/Nominatim/tests/php
     phpunit ./
 
 
+## Running PHP code style tests
+
+    cd ~/Nominatim
+    phpcs --colors .
+
+
+## Running functional tests
+
+Tests in `test/bdd/db` and `test/bdd/osm2pgsql` have to pass 100%. Other
+tests might require full planet-wide data. Sadly even if you have your own
+planet-wide data there will be enough differences to the openstreetmap.org
+installation to cause false positives in the other tests (see FAQ). 
+
+To run the full test suite
+
+    cd ~/Nominatim/test/bdd
+    behave -DBUILDDIR=/home/vagrant/build/ db osm2pgsql
+
+To run a single file
+
+    behave -DBUILDDIR=/home/vagrant/build/ api/lookup/simple.feature
+
+Or a single test by line number
+
+    behave -DBUILDDIR=/home/vagrant/build/ api/lookup/simple.feature:34
+    
+To run specific groups of tests you can add tags just before the `Scenario line`, e.g.
+
+    @bug-34
+    Scenario: address lookup for non-existing or invalid node, way, relation
+
+and then
+
+    behave -DBUILDDIR=/home/vagrant/build/ --tags @bug-34
+
+
 
 
 
@@ -130,17 +143,18 @@ bug fixes) get added since those usually only get applied to new/changed data.
 Also this document skips the optional Wikipedia data import which affects ranking
 of search results. See [Nominatim installation](http://nominatim.org/release-docs/latest/Installation) for details.
 
-##### Why Ubuntu and CentOS, can I test CentOS/CoreOS/FreeBSD?
+##### Why Ubuntu? Can I test CentOS/Fedora/CoreOS/FreeBSD?
 
-There is a Vagrant script for CentOS available. Simply start your box
-with `vagrant up centos` and then log in with `vagrant ssh centos`.
-In general Nominatim will also run in the other environments. The installation steps
+There is a Vagrant script for CentOS available, but the Nominatim directory
+isn't symlinked/mounted to the host which makes development trickier. We used
+it mainly for debugging installation with SELinux.
+
+In general Nominatim will run in the other environments. The installation steps
 are slightly different, e.g. the name of the package manager, Apache2 package
 name, location of files. We chose Ubuntu because that is closest to the
 nominatim.openstreetmap.org production environment.
 
-You can configure/download other Vagrant boxes from [vagrantbox.es](http://www.vagrantbox.es/).
-
+You can configure/download other Vagrant boxes from [https://app.vagrantup.com/boxes/search](https://app.vagrantup.com/boxes/search).
 
 ##### How can I connect to an existing database?
 
@@ -148,7 +162,7 @@ Let's say you have a Postgres database named `nominatim_it` on server `your-serv
 
     pgsql://postgres@your-server.com:5432/nominatim_it
     
-No data import necessary, no restarting necessary.
+No data import necessary or restarting necessary.
 
 If the Postgres installation is behind a firewall, you can try
 
@@ -157,14 +171,15 @@ If the Postgres installation is behind a firewall, you can try
 inside the virtual machine. It will map the port to `localhost:9999` and then
 you edit `settings/local.php` with
 
-    pgsql://postgres@localhost:9999/nominatim_it
+    @define('CONST_Database_DSN', 'pgsql:host=localhost;port=9999;user=postgres;dbname=nominatim_it');
 
 To access postgres directly remember to specify the hostname, e.g. `psql --host localhost --port 9999 nominatim_it`
 
 
 ##### My computer is slow and the import takes too long. Can I start the virtual machine "in the cloud"?
 
-Yes. It's possible to start the virtual machine on [Amazon AWS (plugin)](https://github.com/mitchellh/vagrant-aws) or [DigitalOcean (plugin)](https://github.com/smdahlen/vagrant-digitalocean).
+Yes. It's possible to start the virtual machine on [Amazon AWS (plugin)](https://github.com/mitchellh/vagrant-aws)
+or [DigitalOcean (plugin)](https://github.com/smdahlen/vagrant-digitalocean).
 
 
 

Vagrantfile

@@ -15,6 +15,24 @@ Vagrant.configure("2") do |config|
   end
 
   config.vm.define "ubuntu", primary: true do |sub|
+      sub.vm.box = "bento/ubuntu-18.04"
+      sub.vm.provision :shell do |s|
+        s.path = "vagrant/Install-on-Ubuntu-18.sh"
+        s.privileged = false
+        s.args = [checkout]
+      end
+  end
+
+  config.vm.define "ubuntu18nginx" do |sub|
+      sub.vm.box = "bento/ubuntu-18.04"
+      sub.vm.provision :shell do |s|
+        s.path = "vagrant/Install-on-Ubuntu-18-nginx.sh"
+        s.privileged = false
+        s.args = [checkout]
+      end
+  end
+
+  config.vm.define "ubuntu16" do |sub|
       sub.vm.box = "bento/ubuntu-16.04"
       sub.vm.provision :shell do |s|
         s.path = "vagrant/Install-on-Ubuntu-16.sh"
@@ -32,38 +50,21 @@ Vagrant.configure("2") do |config|
       end
   end
 
-   config.vm.define "centos" do |sub|
-      sub.vm.box = "bento/centos-7.2"
+  config.vm.define "centos" do |sub|
+      sub.vm.box = "centos/7"
       sub.vm.provision :shell do |s|
         s.path = "vagrant/Install-on-Centos-7.sh"
         s.privileged = false
-        s.args = [checkout]
+        s.args = "yes"
       end
+      sub.vm.synced_folder ".", "/home/vagrant/Nominatim", disabled: true
+      sub.vm.synced_folder ".", "/vagrant", disabled: true
   end
 
-  # configure shared package cache if possible
-  #if Vagrant.has_plugin?("vagrant-cachier")
-  #  config.cache.enable :apt
-  #  config.cache.scope = :box
-  #end
-
-
   config.vm.provider "virtualbox" do |vb|
     vb.gui = false
-    vb.customize ["modifyvm", :id, "--memory", "2048"]
+    vb.memory = 2048
+    vb.customize ["setextradata", :id, "VBoxInternal2/SharedFoldersEnableSymlinksCreate//vagrant","0"]
   end
 
-
-  # config.vm.provider :digital_ocean do |provider, override|
-  #   override.ssh.private_key_path = '~/.ssh/id_rsa'
-  #   override.vm.box = 'digital_ocean'
-  #   override.vm.box_url = "https://github.com/smdahlen/vagrant-digitalocean/raw/master/box/digital_ocean.box"
-
-  #   provider.token = ''
-  #   # provider.token = 'YOUR TOKEN'
-  #   provider.image = 'ubuntu-14-04-x64'
-  #   provider.region = 'nyc2'
-  #   provider.size = '512mb'
-  # end
-
 end

cmake/script.tmpl

@@ -0,0 +1,4 @@
+#!@PHP_BIN@ -Cq
+<?php
+require_once(dirname(dirname(__FILE__)).'/settings/settings.php');
+require_once(CONST_BasePath.'/@script_source@');

cmake/website.tmpl

@@ -0,0 +1,3 @@
+<?php
+require_once(dirname(dirname(__FILE__)).'/settings/settings.php');
+require_once(CONST_BasePath.'/@script_source@');

data-sources/country-grid/README.md

@@ -0,0 +1,77 @@
+# Fallback Country Boundaries
+
+Each place is assigned a `country_code` and partition. Partitions derive from `country_code`.
+
+Nominatim imports two pre-generated files
+
+   * `data/country_name.sql` (country code, name, default language, partition)
+   * `data/country_osm_grid.sql` (country code, geometry)
+
+before creating places in the database. This helps with fast lookups and missing data (e.g. if the data the user wants to import doesn't contain any country places).
+
+The number of countries in the world can change (South Sudan created 2011, Germany reunification), so can their boundaries. This document explain how the pre-generated files can be updated.
+
+
+
+## Country code
+
+Each place is assigned a two letter country_code based on its location, e.g. `gb` for Great Britain. Or `NULL` if no suitable country is found (usually it's in open water then).
+
+In `sql/functions.sql: get_country_code(geometry)` the place's center is checked against
+
+   1. country places already imported from the user's data file. Places are imported by rank low-to-high. Lowest rank 2 is countries so most places should be matched. Still the data file might be incomplete.
+   2. if unmatched: OSM grid boundaries
+   3. if still unmatched: OSM grid boundaries, but allow a small distance
+
+
+
+## Partitions
+
+Each place is assigned partition, which is a number 0..250. 0 is fallback/other.
+
+During place indexing (`sql/functions.sql: placex_insert()`) a place is assigned the partition based on its country code (`sql/functions.sql: get_partition(country_code)`). It checks in the `country_name` table.
+
+Most countries have their own partition, some share a partition. Thus partition counts vary greatly.
+
+Several database tables are split by partition to allow queries to run against less indices and improve caching.
+
+   * `location_area_large_<partition>`
+   * `search_name_<partition>`
+   * `location_road_<partition>`
+
+
+
+
+
+## Data files
+
+### data/country_name.sql
+
+Export from existing database table plus manual changes. `country_default_language_code` most taken from [https://wiki.openstreetmap.org/wiki/Nominatim/Country_Codes](), see `utils/country_languages.php`.
+
+
+
+### data/country_osm_grid.sql
+
+`country_grid.sql` merges territories by country. Then uses `function.sql: quad_split_geometry` to split each country into multiple [Quadtree](https://en.wikipedia.org/wiki/Quadtree) polygons for faster point-in-polygon lookups.
+
+To visualize one country as geojson feature collection, e.g. for loading into [geojson.io](http://geojson.io/):
+
+```
+-- http://www.postgresonline.com/journal/archives/267-Creating-GeoJSON-Feature-Collections-with-JSON-and-PostGIS-functions.html
+
+SELECT row_to_json(fc)
+FROM (
+  SELECT 'FeatureCollection' As type, array_to_json(array_agg(f)) As features
+  FROM (
+    SELECT 'Feature' As type,
+    ST_AsGeoJSON(lg.geometry)::json As geometry,
+    row_to_json((country_code, area)) As properties
+    FROM country_osm_grid As lg where country_code='mx'
+  ) As f
+) As fc;
+```
+
+`cat /tmp/query.sql | psql -At nominatim > /tmp/mexico.quad.geojson`
+
+![mexico](mexico.quad.png)

data-sources/gb-postcodes/README.md

@@ -0,0 +1,56 @@
+# GB Postcodes
+
+
+The server [importing instructions](https://www.nominatim.org/release-docs/latest/admin/Import-and-Update/) allow optionally download [`gb_postcode_data.sql.gz`](https://www.nominatim.org/data/gb_postcode_data.sql.gz). This document explains how the file got created.
+
+## GB vs UK
+
+GB (Great Britain) is more correct as the Ordnance Survey dataset doesn't contain postcodes from Northern Ireland.
+
+## Importing separately after the initial import
+
+If you forgot to download the file, or have a new version, you can import it separately:
+
+1. Import the downloaded `gb_postcode_data.sql.gz` file.
+
+2. Run the SQL query `SELECT count(getorcreate_postcode_id(postcode)) FROM gb_postcode;`. This will update the search index.
+
+3. Run `utils/setup.php --calculate-postcodes` from the build directory. This will copy data form the `gb_postcode` table to the `location_postcodes` table.
+
+
+
+## Converting Code-Point Open data
+
+1. Download from [Code-Point® Open](https://www.ordnancesurvey.co.uk/business-and-government/products/code-point-open.html). It requires an email address where a download link will be send to.
+
+2. `unzip codepo_gb.zip`
+
+    Unpacked you'll see a directory of CSV files.
+
+        $ more codepo_gb/Data/CSV/n.csv
+        "N1 0AA",10,530626,183961,"E92000001","E19000003","E18000007","","E09000019","E05000368"
+        "N1 0AB",10,530559,183978,"E92000001","E19000003","E18000007","","E09000019","E05000368"
+
+    The coordinates are "Northings" and "Eastings" in [OSGB 1936](http://epsg.io/1314) projection. They can be projected to WGS84 like this
+
+        SELECT ST_AsText(ST_Transform(ST_SetSRID('POINT(530626 183961)'::geometry,27700), 4326));
+        POINT(-0.117872733220225 51.5394424719303)
+
+    [-0.117872733220225 51.5394424719303 on OSM map](https://www.openstreetmap.org/?mlon=-0.117872733220225&mlat=51.5394424719303&zoom=16)
+
+
+
+3. Create database, import CSV files, add geometry column, dump into file
+
+        DBNAME=create_gb_postcode_file
+        createdb $DBNAME
+        echo 'CREATE EXTENSION postgis' | psql $DBNAME
+
+        cat data/gb_postcode_table.sql | psql $DBNAME      
+        cat codepo_gb/Data/CSV/*.csv | ./data-sources/gb-postcodes/convert_codepoint.php | psql $DBNAME
+        cat codepo_gb/Doc/licence.txt | iconv -f iso-8859-1 -t utf-8 | dos2unix | sed 's/^/-- /g' > gb_postcode_data.sql
+        pg_dump -a -t gb_postcode $DBNAME | grep -v '^--' >> gb_postcode_data.sql
+      
+        gzip -9 -f gb_postcode_data.sql
+        ls -lah gb_postcode_data.*
+        # dropdb $DBNAME

data-sources/gb-postcodes/convert_codepoint.php

@@ -0,0 +1,37 @@
+#!/usr/bin/env php
+<?php
+
+echo <<< EOT
+
+ALTER TABLE gb_postcode ADD COLUMN easting bigint;
+ALTER TABLE gb_postcode ADD COLUMN northing bigint;
+
+TRUNCATE gb_postcode;
+
+COPY gb_postcode (id, postcode, easting, northing) FROM stdin;
+
+EOT;
+
+$iCounter = 0;
+while ($sLine = fgets(STDIN)) {
+    $aColumns = str_getcsv($sLine);
+
+    // insert space before the third last position
+    // https://stackoverflow.com/a/9144834
+    $postcode = $aColumns[0];
+    $postcode = preg_replace('/\s*(...)$/', ' $1', $postcode);
+
+    echo join("\t", array($iCounter, $postcode, $aColumns[2], $aColumns[3]))."\n";
+
+    $iCounter = $iCounter + 1;
+}
+
+echo <<< EOT
+\.
+
+UPDATE gb_postcode SET geometry=ST_Transform(ST_SetSRID(CONCAT('POINT(', easting, ' ', northing, ')')::geometry, 27700), 4326);
+
+ALTER TABLE gb_postcode DROP COLUMN easting;
+ALTER TABLE gb_postcode DROP COLUMN northing;
+
+EOT;

data-sources/us-tiger/README.md

@@ -0,0 +1,29 @@
+# US TIGER address data
+
+Convert [TIGER](https://www.census.gov/geo/maps-data/data/tiger.html)/Line dataset of the US Census Bureau to SQL files which can be imported by Nominatim. The created tables in the Nominatim database are separate from OpenStreetMap tables and get queried at search time separately.
+
+The dataset gets updated once per year. Downloading is prown to be slow (can take a full day) and converting them can take hours as well.
+
+Replace '2018' with the current year throughout.
+
+  1. Install the GDAL library and python bindings and the unzip tool
+
+        # Ubuntu:
+        sudo apt-get install python-gdal unzip
+        # CentOS:
+        sudo yum install gdal-python unzip
+
+  2. Get the TIGER 2018 data. You will need the EDGES files
+     (3,233 zip files, 11GB total).
+
+         wget -r ftp://ftp2.census.gov/geo/tiger/TIGER2018/EDGES/
+
+  3. Convert the data into SQL statements. Adjust the file paths in the scripts as needed
+
+        cd data-sources/us-tiger
+        ./convert.sh <input-path> <output-path>
+        
+  4. Maybe: package the created files
+  
+        tar -czf tiger2018-nominatim-preprocessed.tar.gz tiger
+        

data-sources/us-tiger/convert.sh

@@ -0,0 +1,48 @@
+#!/bin/bash
+
+INPATH=$1
+OUTPATH=$2
+
+if [[ ! -d "$INPATH" ]]; then
+    echo "input path does not exist"
+    exit 1
+fi
+
+if [[ ! -d "$OUTPATH" ]]; then
+    echo "output path does not exist"
+    exit 1
+fi
+
+INREGEX='_([0-9]{5})_edges.zip'
+WORKPATH="$OUTPATH/tmp-workdir/"
+mkdir -p "$WORKPATH"
+
+
+
+INFILES=($INPATH/*.zip)
+echo "Found ${#INFILES[*]} files."
+
+for F in ${INFILES[*]}; do
+    # echo $F
+
+    if [[ "$F" =~ $INREGEX ]]; then
+        COUNTYID=${BASH_REMATCH[1]}
+        SHAPEFILE="$WORKPATH/$(basename $F '.zip').shp"
+        SQLFILE="$OUTPATH/$COUNTYID.sql"
+
+        unzip -o -q -d "$WORKPATH" "$F"
+        if [[ ! -e "$SHAPEFILE" ]]; then
+            echo "Unzip failed. $SHAPEFILE not found."
+            exit 1
+        fi
+
+        ./tiger_address_convert.py "$SHAPEFILE" "$SQLFILE"
+
+        rm $WORKPATH/*
+    fi
+done
+
+OUTFILES=($OUTPATH/*.sql)
+echo "Wrote ${#OUTFILES[*]} files."
+
+rmdir $WORKPATH

data/words.sql

@@ -18,12 +18,13 @@ SET default_with_oids = false;
 -- Name: word_frequencies; Type: TABLE; Schema: public; Owner: -; Tablespace: 
 --
 
+DROP TABLE IF EXISTS word_frequencies;
+
 CREATE TABLE word_frequencies (
     word_token text,
     count bigint
 );
 
-
 --
 -- Data for Name: word_frequencies; Type: TABLE DATA; Schema: public; Owner: -
 --

docs/CMakeLists.txt

@@ -5,12 +5,22 @@
 
 configure_file(mkdocs.yml ../mkdocs.yml)
 file(MAKE_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/appendix)
+file(MAKE_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/data-sources)
 
 ADD_CUSTOM_TARGET(doc
    COMMAND ${CMAKE_COMMAND} -E create_symlink ${CMAKE_CURRENT_SOURCE_DIR}/admin ${CMAKE_CURRENT_BINARY_DIR}/admin
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${CMAKE_CURRENT_SOURCE_DIR}/develop ${CMAKE_CURRENT_BINARY_DIR}/develop
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${CMAKE_CURRENT_SOURCE_DIR}/api ${CMAKE_CURRENT_BINARY_DIR}/api
    COMMAND ${CMAKE_COMMAND} -E create_symlink ${CMAKE_CURRENT_SOURCE_DIR}/index.md ${CMAKE_CURRENT_BINARY_DIR}/index.md
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${CMAKE_CURRENT_SOURCE_DIR}/extra.css ${CMAKE_CURRENT_BINARY_DIR}/extra.css
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${CMAKE_CURRENT_SOURCE_DIR}/data-sources/overview.md ${CMAKE_CURRENT_BINARY_DIR}/data-sources/overview.md
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${PROJECT_SOURCE_DIR}/data-sources/us-tiger/README.md ${CMAKE_CURRENT_BINARY_DIR}/data-sources/US-Tiger.md
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${PROJECT_SOURCE_DIR}/data-sources/gb-postcodes/README.md ${CMAKE_CURRENT_BINARY_DIR}/data-sources/GB-Postcodes.md
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${PROJECT_SOURCE_DIR}/data-sources/country-grid/README.md ${CMAKE_CURRENT_BINARY_DIR}/data-sources/Country-Grid.md
+   COMMAND ${CMAKE_COMMAND} -E create_symlink ${PROJECT_SOURCE_DIR}/data-sources/country-grid/mexico.quad.png ${CMAKE_CURRENT_BINARY_DIR}/data-sources/mexico.quad.png
    COMMAND ${CMAKE_CURRENT_SOURCE_DIR}/bash2md.sh ${PROJECT_SOURCE_DIR}/vagrant/Install-on-Centos-7.sh ${CMAKE_CURRENT_BINARY_DIR}/appendix/Install-on-Centos-7.md
    COMMAND ${CMAKE_CURRENT_SOURCE_DIR}/bash2md.sh ${PROJECT_SOURCE_DIR}/vagrant/Install-on-Ubuntu-16.sh ${CMAKE_CURRENT_BINARY_DIR}/appendix/Install-on-Ubuntu-16.md
+   COMMAND ${CMAKE_CURRENT_SOURCE_DIR}/bash2md.sh ${PROJECT_SOURCE_DIR}/vagrant/Install-on-Ubuntu-18.sh ${CMAKE_CURRENT_BINARY_DIR}/appendix/Install-on-Ubuntu-18.md
    COMMAND mkdocs build -d ${CMAKE_CURRENT_BINARY_DIR}/../site-html -f ${CMAKE_CURRENT_BINARY_DIR}/../mkdocs.yml
 )
 

docs/admin/Faq.md

@@ -1,29 +1,10 @@
-# Running Your Own Instance
+# Troubleshooting Nominatim Installations
 
-### Can I import only a few countries and also keep them up to date?
+## Installation Issues
 
-You should use the extracts and updates from https://download.geofabrik.de.
-For the intial import, download the countries you need and merge them.
-See [OSM Help](https://help.openstreetmap.org/questions/48843/merging-two-or-more-geographical-areas-to-import-two-or-more-osm-files-in-nominatim)
-for examples how to do that. Use the resulting single osm file when
-running `setup.php`.
+### Can a stopped/killed import process be resumed?
 
-For updates you need to download the change files for each country
-once per day and apply them **separately** using
-
-    ./utils/update.php --import-diff <filename> --index
-    
-See [this issue](https://github.com/openstreetmap/Nominatim/issues/60#issuecomment-18679446)
-for a script that runs the updates using osmosis.
-
-### My website shows: `XML Parsing Error: XML or text declaration not at start of entity Location</code>.`
-
-Make sure there are no spaces at the beginning of your `settings/local.php` file.
-
-
-# Installation
-
-### I accidentally killed the import process after it has been running for many hours. Can it be resumed?
+"I accidentally killed the import process after it has been running for many hours. Can it be resumed?"
 
 It is possible if the import already got to the indexing stage.
 Check the last line of output that was logged before the process
@@ -41,11 +22,12 @@ then you can resume with the following command:
 If the reported rank is 26 or higher, you can also safely add `--index-noanalyse`.
 
 
-### When running the setup.php script I get a warning:
+### PHP "open_basedir restriction in effect" warnings
+
     `PHP Warning:  file_get_contents(): open_basedir restriction in effect.`
 
 You need to adjust the [open_basedir](http://www.php.net/manual/en/ini.core.php#ini.open-basedir) setting
-in your PHP configuration (php.ini file). By default this setting may look like this:
+in your PHP configuration (`php.ini file`). By default this setting may look like this:
 
     open_basedir = /srv/http/:/home/:/tmp/:/usr/share/pear/
 
@@ -54,7 +36,9 @@ dding ";" at the beginning of the line. Don't forget to enable this setting agai
 once you are done with the PHP command line operations.
 
 
-### The Apache log contains lots of PHP warnings like this:
+### PHP timzeone warnings
+
+The Apache log may contain lots of PHP warnings like this:
     `PHP Warning:  date_default_timezone_set() function.`
 
 You should set the default time zone as instructed in the warning in
@@ -71,28 +55,49 @@ Or
 echo "date.timezone = 'America/Denver'" > /etc/php.d/timezone.ini
 ```
 
-### When running the import I get a version mismatch:
-    `COPY_END for place failed: ERROR: incompatible library "/opt/Nominatim/module/nominatim.so": version mismatch`
+### nominatim.so version mismatch
+
+When running the import you may get a version mismatch:
+`COPY_END for place failed: ERROR: incompatible library "/srv/Nominatim/nominatim/build/module/nominatim.so": version mismatch`
 
 pg_config seems to use bad includes sometimes when multiple versions
 of PostgreSQL are available in the system. Make sure you remove the
-server development libraries (`postgresql-server-dev-9.1` on Ubuntu)
+server development libraries (`postgresql-server-dev-9.5` on Ubuntu)
 and recompile (`cmake .. && make`).
 
 
-### I see the error: `function transliteration(text) does not exist`
+### I see the error: "function transliteration(text) does not exist"
 
 Reinstall the nominatim functions with `setup.php --create--functions`
 and check for any errors, e.g. a missing `nominatim.so` file.
 
+### I see the error: "ERROR: mmap (remap) failed"
 
-### The website shows: `Could not get word tokens`
+This may be a simple out-of-memory error. Try reducing the memory used
+for `--osm2pgsql-cache`. Also make sure that overcommitting memory is
+allowed: `cat /proc/sys/vm/overcommit_memory` should print 0 or 1.
+
+If you are using a flatnode file, then it may also be that the underlying
+filesystem does not fully support 'mmap'. A notable candidate is virtualbox's
+vboxfs.
+
+### nominatim UPDATE failed: ERROR: buffer 179261 is not owned by resource owner Portal
+
+Several users [reported this](https://github.com/openstreetmap/Nominatim/issues/1168) during the initial import of the database. It's 
+something Postgresql internal Nominatim doesn't control. And Postgresql forums
+suggest it's threading related but definitely some kind of crash of a process.
+Users reported either rebooting the server, different hardware or just trying
+the import again worked. 
+
+### The website shows: "Could not get word tokens"
 
 The server cannot access your database. Add `&debug=1` to your URL
 to get the full error message.
 
 
-### On CentOS the website shows `could not connect to server: No such file or directory`
+### On CentOS the website shows "Could not connect to server"
+
+`could not connect to server: No such file or directory`
 
 On CentOS v7 the PostgreSQL server is started with `systemd`.
 Check if `/usr/lib/systemd/system/httpd.service` contains a line `PrivateTmp=true`.
@@ -105,22 +110,98 @@ However, you can solve this the quick and dirty way by commenting out that line
     sudo systemctl restart httpd
 
 
-### Setup.php fails with the message: `DB Error: extension not found`
+### "must be an array or an object that implements Countable" warning in /usr/share/pear/DB.php
+
+The warning started with PHP 7.2. Make sure you have at least [version 1.9.3 of PEAR DB](https://github.com/pear/DB/releases)
+installed.
+
+### Website reports "DB Error: insufficient permissions"
+
+The user the webserver, e.g. Apache, runs under needs to have access to the Nominatim database. You can find the user like [this](https://serverfault.com/questions/125865/finding-out-what-user-apache-is-running-as), for default Ubuntu operating system for example it's `www-data`.
+
+1. Repeat the `createuser` step of the installation instructions.
+
+2. Give the user permission to existing tables
+
+```
+   GRANT usage ON SCHEMA public TO "www-data";
+   GRANT SELECT ON ALL TABLES IN SCHEMA public TO "www-data";
+```
+
+### Website reports "Could not load library "nominatim.so"
+
+Example error message
+
+```
+   SELECT make_standard_name('3039 E MEADOWLARK LN') [nativecode=ERROR: could not
+   load library "/srv/nominatim/Nominatim-3.1.0/build/module/nominatim.so":
+   /srv/nominatim/Nominatim-3.1.0/build/module/nominatim.so: cannot open shared
+   object file: Permission denied
+   CONTEXT: PL/pgSQL function make_standard_name(text) line 5 at assignment]
+```
+
+The Postgresql database, i.e. user postgres, needs to have access to that file.
+
+The permission need to be read & executable by everybody, e.g.
+
+```
+   -rwxr-xr-x 1 nominatim nominatim 297984 build/module/nominatim.so
+```
+
+Try `chmod a+r nominatim.so; chmod a+x nominatim.so`.
+
+When running SELinux, make sure that the
+[context is set up correctly](../appendix/Install-on-Centos-7/#adding-selinux-security-settings).
+
+### Setup.php fails with "DB Error: extension not found"
 
 Make sure you have the Postgres extensions hstore and postgis installed.
 See the installation instruction for a full list of required packages.
 
-### When running the setup.php script I get a error:
-    `Cannot redeclare getDB() (previously declared in /your/path/Nominatim/lib/db.php:4)`
+
+### Setup.php reports "Cannot redeclare getDB()"
+
+`Cannot redeclare getDB() (previously declared in /your/path/Nominatim/lib/db.php:4)`
 
 The message is a bit misleading as PHP needs to load the file `DB.php` and
 instead re-loads Nominatim's `db.php`. To solve this make sure you
-have the [http://pear.php.net/package/DB/ Pear module 'DB'] installed.
+have the [Pear module 'DB'](http://pear.php.net/package/DB/) installed.
 
     sudo pear install DB
 
 ### I forgot to delete the flatnodes file before starting an import.
 
 That's fine. For each import the flatnodes file get overwritten.
-See https://help.openstreetmap.org/questions/52419/nominatim-flatnode-storage
+See [https://help.openstreetmap.org/questions/52419/nominatim-flatnode-storage]()
 for more information.
+
+
+## Running your own instance
+
+### Can I import multiple countries and keep them up to date?
+
+You should use the extracts and updates from https://download.geofabrik.de.
+For the initial import, download the countries you need and merge them.
+See [OSM Help](https://help.openstreetmap.org/questions/48843/merging-two-or-more-geographical-areas-to-import-two-or-more-osm-files-in-nominatim)
+for examples how to do that. Use the resulting single osm file when
+running `setup.php`.
+
+For updates you need to download the change files for each country
+once per day and apply them **separately** using
+
+    ./utils/update.php --import-diff <filename> --index
+    
+See [this issue](https://github.com/openstreetmap/Nominatim/issues/60#issuecomment-18679446)
+for a script that runs the updates using osmosis.
+
+### Can I import negative OSM ids into Nominatim?
+
+See [this question of Stackoverflow](https://help.openstreetmap.org/questions/64662/nominatim-flatnode-with-negative-id).
+
+### Missing XML or text declaration
+
+The website might show: `XML Parsing Error: XML or text declaration not at start of entity Location.`
+
+Make sure there are no spaces at the beginning of your `settings/local.php` file.
+
+

docs/admin/Import-and-Update.md

@@ -1,9 +1,11 @@
+# Importing and Updating the Database
+
 The following instructions explain how to create a Nominatim database
 from an OSM planet file and how to keep the database up to date. It
 is assumed that you have already successfully installed the Nominatim
 software itself, if not return to the [installation page](Installation.md).
 
-# Configuration setup in settings/local.php
+## Configuration setup in settings/local.php
 
 The Nominatim server can be customized via the file `settings/local.php`
 in the build directory. Note that this is a PHP file, so it must always
@@ -16,7 +18,7 @@ without any leading spaces.
 There are lots of configuration settings you can tweak. Have a look
 at `settings/default.php` for a full list. Most should have a sensible default.
 
-### Flatnode files
+#### Flatnode files
 
 If you plan to import a large dataset (e.g. Europe, North America, planet),
 you should also enable flatnode storage of node locations. With this
@@ -29,9 +31,9 @@ Add to your `settings/local.php`:
 Replace the second part with a suitable path on your system and make sure
 the directory exists. There should be at least 40GB of free space.
 
-# Downloading additional data
+## Downloading additional data
 
-## Wikipedia rankings
+### Wikipedia rankings
 
 Wikipedia can be used as an optional auxiliary data source to help indicate
 the importance of osm features. Nominatim will work without this information
@@ -49,20 +51,100 @@ size of nominatim. They also increase the install time by an hour or so.
 the initial import of the data if you want the rankings applied to the
 loaded data.
 
-## UK postcodes
+### UK postcodes
 
 Nominatim can use postcodes from an external source to improve searches that involve a UK postcode. This data can be optionally downloaded: 
 
     cd $NOMINATIM_SOURCE_DIR/data
     wget https://www.nominatim.org/data/gb_postcode_data.sql.gz
 
+## Choosing the Data to Import
 
-# Initial import of the data
+In its default setup Nominatim is configured to import the full OSM data
+set for the entire planet. Such a setup requires a powerful machine with
+at least 32GB of RAM and around 800GB of SSD hard disks. Depending on your
+use case there are various ways to reduce the amount of data imported. This
+section discusses these methods. They can also be combined.
 
-**Important:** first try the import with a small excerpt, for example from
+### Using an extract
+
+If you only need geocoding for a smaller region, then precomputed extracts
+are a good way to reduce the database size and import time.
+[Geofabrik](https://download.geofabrik.de) offers extracts for most countries.
+They even have daily updates which can be used with the update process described
+below. There are also
+[other providers for extracts](https://wiki.openstreetmap.org/wiki/Planet.osm#Downloading).
+
+Please be aware that some extracts are not cut exactly along the country
+boundaries. As a result some parts of the boundary may be missing which means
+that cannot compute the areas for some administrative areas.
+
+### Dropping Data Required for Dynamic Updates
+
+About half of the data in Nominatim's database is not really used for serving
+the API. It is only there to allow the data to be updated from the latest
+changes from OSM. For many uses these dynamic updates are not really required.
+If you don't plan to apply updates, the dynamic part of the database can be
+safely dropped using the following command:
+
+```
+./utils/setup.php --drop
+```
+
+Note that you still need to provide for sufficient disk space for the initial
+import. So this option is particularly interesting if you plan to transfer the
+database or reuse the space later.
+
+### Reverse-only Imports
+
+If you only want to use the Nominatim database for reverse lookups or
+if you plan to use the installation only for exports to a
+[photon](http://photon.komoot.de/) database, then you can set up a database
+without search indexes. Add `--reverse-only` to your setup command above.
+
+This saves about 5% of disk space.
+
+### Filtering Imported Data
+
+Nominatim normally sets up a full search database containing administrative
+boundaries, places, streets, addresses and POI data. There are also other
+import styles available which only read selected data:
+
+* **settings/import-admin.style**
+  Only import administrative boundaries and places.
+* **settings/import-street.style**
+  Like the admin style but also adds streets.
+* **settings/import-address.style**
+  Import all data necessary to compute addresses down to house number level.
+* **settings/import-full.style**
+  Default style that also includes points of interest.
+
+The style can be changed with the configuration `CONST_Import_Style`.
+
+To give you an idea of the impact of using the different style, the table
+below gives rough estimates of the final database size after import of a
+2018 planet and after using the `--drop` option. It also shows the time
+needed for the import on a machine with 32GB RAM, 4 CPUS and SSDs. Note that
+the given sizes are just an estimate meant for comparison of style requirements.
+Your planet import is likely to be larger as the OSM data grows with time.
+
+style     | Import time  |  DB size   |  after drop
+----------|--------------|------------|------------
+admin     |    5h        |  190 GB    |   20 GB
+street    |   42h        |  400 GB    |  180 GB
+address   |   59h        |  500 GB    |  260 GB
+full      |   80h        |  575 GB    |  300 GB
+
+You can also customize the styles further. For an description of the
+style format see [the developement section](../develop/Import.md).
+
+## Initial import of the data
+
+**Important:** first try the import with a small extract, for example from
 [Geofabrik](https://download.geofabrik.de).
 
-Download the data to import and load the data with the following command:
+Download the data to import and load the data with the following command
+from the build directory:
 
 ```sh
 ./utils/setup.php --osm-file <data file> --all [--osm2pgsql-cache 28000] 2>&1 | tee setup.log
@@ -97,66 +179,44 @@ you also need to enable these key phrases like this:
 Note that this command downloads the phrases from the wiki link above.
 
 
-# Installing Tiger housenumber data for the US
+## Installing Tiger housenumber data for the US
 
-Nominatim is able to use the official TIGER address set to complement the
-OSM house number data in the US. You can add TIGER data to your own Nominatim
-instance by following these steps:
+Nominatim is able to use the official [TIGER](https://www.census.gov/geo/maps-data/data/tiger.html)
+address set to complement the OSM house number data in the US. You can add
+TIGER data to your own Nominatim instance by following these steps. The
+entire US adds about 10GB to your database.
 
-  1. Install the GDAL library and python bindings and the unzip tool
-
-       * Ubuntu: `sudo apt-get install python-gdal unzip`
-       * CentOS: `sudo yum install gdal-python unzip`
-
-  2. Get preprocessed TIGER 2015 data and unpack it into the
+  1. Get preprocessed TIGER 2018 data and unpack it into the
      data directory in your Nominatim sources:
 
         cd Nominatim/data
-        wget https://nominatim.org/data/tiger2017-nominatim-preprocessed.tar.gz
-        tar xf tiger2017-nominatim-preprocessed.tar.gz
+        wget https://nominatim.org/data/tiger2018-nominatim-preprocessed.tar.gz
+        tar xf tiger2018-nominatim-preprocessed.tar.gz
 
-  3. Import the data into your Nominatim database: 
+    `data-source/us-tiger/README.md` explains how the data got preprocessed.
+
+  2. Import the data into your Nominatim database: 
 
         ./utils/setup.php --import-tiger-data
 
-  4. Enable use of the Tiger data in your `settings/local.php` by adding:
+  3. Enable use of the Tiger data in your `settings/local.php` by adding:
 
          @define('CONST_Use_US_Tiger_Data', true);
 
-  5. Apply the new settings:
+  4. Apply the new settings:
 
 ```sh
     ./utils/setup.php --create-functions --enable-diff-updates --create-partition-functions
 ```
 
-The entire US adds about 10GB to your database.
 
-You can also process the data from the original TIGER data to create the
-SQL files, Nominatim needs for the import:
-
-  1. Get the TIGER 2017 data. You will need the EDGES files
-     (3,234 zip files, 11GB total). Choose one of the two sources:
-
-         wget -r ftp://ftp2.census.gov/geo/tiger/TIGER2017/EDGES/
-         wget -r ftp://mirror1.shellbot.com/census/geo/tiger/TIGER2017/EDGES/
-
-     The first one is the original source, the second a considerably faster
-     mirror.
-
-  2. Convert the data into SQL statements: 
-
-         ./utils/imports.php --parse-tiger <tiger edge data directory>
-
-Be warned that this can take quite a long time. After this process is finished,
-the same preprocessed files as above are available in `data/tiger`.
-
-# Updates
+## Updates
 
 There are many different possibilities to update your Nominatim database.
 The following section describes how to keep it up-to-date with Pyosmium.
 For a list of other methods see the output of `./utils/update.php --help`.
 
-### Installing the newest version of Pyosmium
+#### Installing the newest version of Pyosmium
 
 It is recommended to install Pyosmium via pip. Run (as the same user who
 will later run the updates):
@@ -165,7 +225,7 @@ will later run the updates):
 pip install --user osmium
 ```
 
-Nominatim needs a tool called `pyosmium-get-updates` that comes with
+Nominatim needs a tool called `pyosmium-get-updates`, which comes with
 Pyosmium. You need to tell Nominatim where to find it. Add the
 following line to your `settings/local.php`:
 
@@ -174,7 +234,7 @@ following line to your `settings/local.php`:
 The path above is fine if you used the `--user` parameter with pip.
 Replace `user` with your user name.
 
-### Setting up the update process
+#### Setting up the update process
 
 Next the update needs to be initialised. By default Nominatim is configured
 to update using the global minutely diffs.
@@ -200,7 +260,7 @@ what you expect.
 The --init-updates command needs to be rerun whenever the replication service
 is changed.
 
-### Updating Nominatim
+#### Updating Nominatim
 
 The following command will keep your database constantly up to date:
 

docs/admin/Installation.md

@@ -1,16 +1,26 @@
+# Basic Installation
+
 This page contains generic installation instructions for Nominatim and its
 prerequisites. There are also step-by-step instructions available for
 the following operating systems:
 
+  * [Ubuntu 18.04](../appendix/Install-on-Ubuntu-18.md)
   * [Ubuntu 16.04](../appendix/Install-on-Ubuntu-16.md)
   * [CentOS 7.2](../appendix/Install-on-Centos-7.md)
 
 These OS-specific instructions can also be found in executable form
 in the `vagrant/` directory.
 
-# Prerequisites
+Users have created instructions for other frameworks. We haven't tested those
+and can't offer support.
 
-## Software
+  * [Docker](https://github.com/mediagis/nominatim-docker)
+  * [Docker on Kubernetes](https://github.com/peter-evans/nominatim-k8s)
+  * [Ansible](https://github.com/synthesio/infra-ansible-nominatim)
+
+## Prerequisites
+
+### Software
 
 For compiling:
 
@@ -42,22 +52,22 @@ For running continuous updates:
 
   * [pyosmium](http://osmcode.org/pyosmium/)
 
-## Hardware
+### Hardware
 
 A minimum of 2GB of RAM is required or installation will fail. For a full
 planet import 32GB of RAM or more strongly are recommended.
 
-For a full planet install you will need about 600GB of hard disk space (as of
-January 2017, take into account that the OSM database is growing fast). SSD disks
+For a full planet install you will need at least 700GB of hard disk space
+(take into account that the OSM database is growing fast). SSD disks
 will help considerably to speed up import and queries.
 
 On a 6-core machine with 32GB RAM and SSDs the import of a full planet takes
 a bit more than 2 days. Without SSDs 7-8 days are more realistic.
 
 
-# Setup of the server
+## Setup of the server
 
-## PostgreSQL tuning
+### PostgreSQL tuning
 
 You might want to tune your PostgreSQL installation so that the later steps
 make best use of your hardware. You should tune the following parameters in
@@ -84,13 +94,13 @@ Don't forget to reenable them after the initial import or you risk database
 corruption. Autovacuum must not be switched off because it ensures that the
 tables are frequently analysed.
 
-## Webserver setup
+### Webserver setup
 
 The `website/` directory in the build directory contains the configured
 website. Include the directory into your webbrowser to serve php files
 from there.
 
-### Configure for use with Apache
+#### Configure for use with Apache
 
 Make sure your Apache configuration contains the required permissions for the
 directory and create an alias:
@@ -109,7 +119,7 @@ build directory.
 After making changes in the apache config you need to restart apache.
 The website should now be available on http://localhost/nominatim.
 
-### Configure for use with Nginx
+#### Configure for use with Nginx
 
 Use php-fpm as a deamon for serving PHP cgi. Install php-fpm together with nginx.
 

docs/admin/Migration.md

@@ -1,13 +1,93 @@
-Database Migrations
-===================
+# Database Migrations
 
 This page describes database migrations necessary to update existing databases
 to newer versions of Nominatim.
 
 SQL statements should be executed from the postgres commandline. Execute
-`psql nominiatim` to enter command line mode.
+`psql nominatim` to enter command line mode.
 
-# 3.0.0 -> 3.1.0
+
+## 3.2.0 -> 3.3.0
+
+### New database connection string (DSN) format
+
+Previously database connection setting (`CONST_Database_DSN` in `settings/*.php`) had the format
+
+   * (simple) `pgsql://@/nominatim`
+   * (complex) `pgsql://johndoe:secret@machine1.domain.com:1234/db1`
+
+The new format is
+
+   * (simple) `pgsql:dbname=nominatim`
+   * (complex) `pgsql:dbname=db1;host=machine1.domain.com;port=1234;user=johndoe;password=secret`
+
+### Natural Earth country boundaries no longer needed as fallback
+
+```
+DROP TABLE country_naturalearthdata;
+```
+
+Finally, update all SQL functions:
+
+```sh
+./utils/setup.php --create-functions --enable-diff-updates --create-partition-functions
+```
+
+### Configurable Address Levels
+
+The new configurable address levels require a new table. Create it with the
+following command:
+
+```sh
+./utils/update.php --update-address-levels
+```
+
+## 3.1.0 -> 3.2.0
+
+### New reverse algorithm
+
+The reverse algorithm has changed and requires new indexes. Run the following
+SQL statements to create the indexes:
+
+```
+CREATE INDEX idx_placex_geometry_reverse_lookupPoint
+  ON placex USING gist (geometry)
+  WHERE (name is not null or housenumber is not null or rank_address between 26 and 27)
+    AND class not in ('railway','tunnel','bridge','man_made')
+    AND rank_address >= 26 AND indexed_status = 0 AND linked_place_id is null;
+CREATE INDEX idx_placex_geometry_reverse_lookupPolygon
+  ON placex USING gist (geometry)
+  WHERE St_GeometryType(geometry) in ('ST_Polygon', 'ST_MultiPolygon')
+    AND rank_address between 4 and 25 AND type != 'postcode'
+    AND name is not null AND indexed_status = 0 AND linked_place_id is null;
+CREATE INDEX idx_placex_geometry_reverse_placeNode
+  ON placex USING gist (geometry)
+  WHERE osm_type = 'N' AND rank_search between 5 and 25
+    AND class = 'place' AND type != 'postcode'
+    AND name is not null AND indexed_status = 0 AND linked_place_id is null;
+```
+
+You also need to grant the website user access to the `country_osm_grid` table:
+
+```
+GRANT SELECT ON table country_osm_grid to "www-user";
+```
+
+Replace the `www-user` with the user name of your website server if necessary.
+
+You can now drop the unused indexes:
+
+```
+DROP INDEX idx_placex_reverse_geometry;
+```
+
+Finally, update all SQL functions:
+
+```sh
+./utils/setup.php --create-functions --enable-diff-updates --create-partition-functions
+```
+
+## 3.0.0 -> 3.1.0
 
 ### Postcode Table
 
@@ -26,6 +106,17 @@ CREATE INDEX idx_postcode_geometry ON location_postcode USING GIST (geometry);
 CREATE UNIQUE INDEX idx_postcode_id ON location_postcode USING BTREE (place_id);
 CREATE INDEX idx_postcode_postcode ON location_postcode USING BTREE (postcode);
 GRANT SELECT ON location_postcode TO "www-data";
+drop type if exists nearfeaturecentr cascade;
+create type nearfeaturecentr as (
+  place_id BIGINT,
+  keywords int[],
+  rank_address smallint,
+  rank_search smallint,
+  distance float,
+  isguess boolean,
+  postcode TEXT,
+  centroid GEOMETRY
+);
 ```
 
 Add postcode column to `location_area` tables with SQL statement:

docs/api/Details.md

@@ -0,0 +1,152 @@
+# Place details
+
+Lookup details about a single place by id. The default output is HTML for debugging search logic and results.
+
+**The details page (including JSON output) exists for debugging only and must not be downloaded automatically**, see [Nominatim Usage Policy](https://operations.osmfoundation.org/policies/nominatim/).
+
+
+## Parameters
+
+The details API supports the following two request formats:
+
+```
+  https://nominatim.openstreetmap.org/details?osmtype=[N|W|R]&osmid=<value>&class=<value>
+```
+
+`osmtype` and `osmid` are required parameter. The type is one of node (N), way (W)
+or relation (R). The id must be a number. The `class` parameter is optional and
+allows to distinguish between entries, when the corresponding OSM object has more
+than one main tag. For example, when a place is tagged with `tourism=hotel` and
+`amenity=restaurant`, there will be two place entries in Nominatim, one for a
+restaurant, one for a hotel. You need to specify `class=tourism` or `class=amentity`
+to get exactly the one you want. If there are multiple places in the database
+but the `class` parameter is left out, then one of the places will be chosen
+at random and displayed.
+
+```
+  https://nominatim.openstreetmap.org/details?placeid=<value>
+```
+
+Placeids are assigned sequentially during Nominatim data import. The id for a place is different between Nominatim installation (servers) and changes when data gets reimported. Therefore it can't be used as permanent id and shouldn't be used in bug reports.
+
+
+Additional optional parameters are explained below.
+
+### Output format
+
+* `format=[html|json]`
+
+See [Place Output Formats](Output.md) for details on each format. (Default: html)
+
+* `json_callback=<string>`
+
+Wrap json output in a callback function (JSONP) i.e. `<string>(<json>)`.
+Only has an effect for JSON output formats.
+
+* `pretty=[0|1]`
+
+For JSON output will add indentation to make it more human-readable. (Default: 0)
+
+
+### Output details
+
+* `addressdetails=[0|1]`
+
+Include a breakdown of the address into elements. (Default for JSON: 0, for HTML: 1)
+
+* `keywords=[0|1]`
+
+Include a list of name keywords and address keywords (word ids). (Default: 0)
+
+* `linkedplaces=[0|1]`
+
+Include details of places higher in the address hierarchy. E.g. for a street this is usually the city, state, postal code, country. (Default: 1)
+
+* `hierarchy=[0|1]`
+
+Include details of places lower in the address hierarchy. E.g. for a city this usually a list of streets, suburbs, rivers. (Default for JSON: 0, for HTML: 1)
+
+* `group_hierarchy=[0|1]`
+
+For JSON output will group the places by type. (Default: 0)
+
+* `polygon_geojson=[0|1]`
+
+Include geometry of result. (Default for JSON: 0, for HTML: 1)
+
+### Language of results
+
+* `accept-language=<browser language string>`
+
+Preferred language order for showing result, overrides the value
+specified in the "Accept-Language" HTTP header.
+Either use a standard RFC2616 accept-language string or a simple
+comma-separated list of language codes.
+
+
+## Examples
+
+##### HTML
+
+[https://nominatim.openstreetmap.org/details.php?osmtype=W&osmid=38210407](https://nominatim.openstreetmap.org/details.php?osmtype=W&osmid=38210407)
+
+##### JSON
+
+[https://nominatim.openstreetmap.org/details.php?osmtype=W&osmid=38210407&format=json](https://nominatim.openstreetmap.org/details.php?osmtype=W&osmid=38210407&format=json)
+
+
+```json
+{
+  "place_id": 85993608,
+  "parent_place_id": 72765313,
+  "osm_type": "W",
+  "osm_id": 38210407,
+  "category": "place",
+  "type": "square",
+  "admin_level": "15",
+  "localname": "Pariser Platz",
+  "names": {
+    "name": "Pariser Platz",
+    "name:be": "Парыжская плошча",
+    "name:de": "Pariser Platz",
+    "name:es": "Plaza de París",
+    "name:he": "פאריזר פלאץ",
+    "name:ko": "파리저 광장",
+    "name:la": "Forum Parisinum",
+    "name:ru": "Парижская площадь",
+    "name:uk": "Паризька площа",
+    "name:zh": "巴黎廣場"
+  },
+  "addresstags": {
+    "postcode": "10117"
+  },
+  "housenumber": null,
+  "calculated_postcode": "10117",
+  "country_code": "de",
+  "indexed_date": "2018-08-18T17:02:45+00:00",
+  "importance": 0.339401620591472,
+  "calculated_importance": 0.339401620591472,
+  "extratags": {
+    "wikidata": "Q156716",
+    "wikipedia": "de:Pariser Platz"
+  },
+  "calculated_wikipedia": "de:Pariser_Platz",
+  "rank_address": 30,
+  "rank_search": 30,
+  "isarea": true,
+  "centroid": {
+    "type": "Point",
+    "coordinates": [
+      13.3786822618517,
+      52.5163654
+    ]
+  },
+  "geometry": {
+    "type": "Point",
+    "coordinates": [
+      13.3786822618517,
+      52.5163654
+    ]
+  }
+}
+```

docs/api/Faq.md

@@ -0,0 +1,61 @@
+# Frequently Asked Questions
+
+## API Results
+
+#### 1. The address of my search results contains far-away places that don't belong there.
+
+Nominatim computes the address from two sources in the OpenStreetMap data:
+from administrative boundaries and from place nodes. Boundaries are the more
+useful source. They precisely describe an area. So it is very clear for
+Nominatim if a point belongs to an area of not. Place nodes are more complicated.
+These are only points without any precise extend. So Nominatim has to take a
+guess and assume that an address belongs to the closest place nose it can find.
+In an ideal world, Nominatim would not need the place nodes but there are
+many places on earth where there are not precise boundaries available for
+all parts that make up an address. This is in particular true for the more
+local address parts, like villages and suburbs. Therefore it is not possible
+to completely dismiss place nodes. And sometimes they sneak in where they
+don't belong.
+
+As a OpenStreetMap mapper, you can improve the situation in two ways: if you
+see a place node for which already an administrative area exists, then you
+should _link_ the two by adding the node with a 'label' role to the boundary
+relation. If there is no administrative area, you can add the approximate
+extend of the place and tag it place=<something> as well.
+
+#### 2. When doing reverse search, the address details have parts that don't contain the point I was looking up.
+
+There is a common misconception how the reverse API call works in Nominatim.
+Reverse does not give you the address of the point you asked for. Reverse
+returns the closest object to the point you asked for and then returns the
+address of that object. Now, if you are close to a border, then the closest
+object may be across that border. When Nominatim then returns the address,
+contains the county/state/country across the border.
+
+#### 3. I get different counties/states/countries when I change the zoom parameter in the reverse query. How is that possible?
+
+This is basically the same problem as in the previous answer.
+The zoom level influences at which [search rank](https://wiki.openstreetmap.org/wiki/Nominatim/Development_overview#Country_to_street_level) Nominatim starts looking
+for the closest object. So the closest house number maybe on one side of the
+border while the closest street is on the other. As the address details contain
+the address of the closest object found, you might sometimes get one result,
+sometimes the other for the closest point.
+
+#### 4. Can you return the continent?
+
+Nominatim assigns each map feature one country. Those outside any administrative
+boundaries are assigned a special no-country. Continents or other super-national
+administrations (e.g. European Union, NATO, Custom unions) are not supported, 
+see also [Administrative Boundary](https://wiki.openstreetmap.org/wiki/Tag:boundary%3Dadministrative#Super-national_administrations).
+
+#### 5. Can you return the timezone?
+
+See this separate OpenStreetMap-based project [Timezone Boundary Builder](https://github.com/evansiroky/timezone-boundary-builder)
+
+#### 6. I want to download a list of streets/restaurants of a city/region
+
+The [Overpass API](https://wiki.openstreetmap.org/wiki/Overpass_API) is more
+suited for these kinds of queries.
+
+That said if you installed your own Nominatim instance you can use the
+`/utils/export.php` PHP script as basis to return such lists.

docs/api/Lookup.md

@@ -0,0 +1,147 @@
+# Address lookup
+
+The lookup API allows to query the address and other details of one or
+multiple OSM objects like node, way or relation.
+
+## Parameters
+
+The lookup API has the following format:
+
+```
+  https://nominatim.openstreetmap.org/lookup?osm_ids=[N|W|R]<value>,…,…,&<params>
+```
+
+`osm_ids` is mandatory and must contain a comma-separated list of OSM ids each
+prefixed with its type, one of node(N), way(W) or relation(R). Up to 50 ids
+can be queried at the same time.
+
+Additional optional parameters are explained below.
+
+### Output format
+
+* `format=[html|xml|json|jsonv2|geojson|geocodejson]`
+
+See [Place Output Formats](Output.md) for details on each format. (Default: xml)
+
+* `json_callback=<string>`
+
+Wrap json output in a callback function (JSONP) i.e. `<string>(<json>)`.
+Only has an effect for JSON output formats.
+
+### Output details
+
+* `addressdetails=[0|1]`
+
+Include a breakdown of the address into elements. (Default: 0)
+
+
+* `extratags=[0|1]`
+
+Include additional information in the result if available,
+e.g. wikipedia link, opening hours. (Default: 0)
+
+
+* `namedetails=[0|1]`
+
+Include a list of alternative names in the results. These may include
+language variants, references, operator and brand. (Default: 0)
+
+
+### Language of results
+
+* `accept-language=<browser language string>`
+
+Preferred language order for showing search results, overrides the value
+specified in the "Accept-Language" HTTP header.
+Either use a standard RFC2616 accept-language string or a simple
+comma-separated list of language codes.
+
+
+### Other
+
+* `email=<valid email address>`
+
+If you are making large numbers of request please include an appropriate email
+address to identify your requests. See Nominatim's [Usage Policy](https://operations.osmfoundation.org/policies/nominatim/) for more details.
+
+* `debug=[0|1]`
+
+Output assorted developer debug information. Data on internals of Nominatim's
+"Search Loop" logic, and SQL queries. The output is (rough) HTML format.
+This overrides the specified machine readable format. (Default: 0)
+
+
+## Examples
+
+##### XML
+
+[https://nominatim.openstreetmap.org/lookup?osm_ids=R146656,W104393803,N240109189](https://nominatim.openstreetmap.org/lookup?osm_ids=R146656,W104393803,N240109189)
+
+```xml
+  <lookupresults timestamp="Mon, 29 Jun 15 18:01:33 +0000" attribution="Data © OpenStreetMap contributors, ODbL 1.0. https://www.openstreetmap.org/copyright" querystring="R146656,W104393803,N240109189" polygon="false">
+    <place place_id="127761056" osm_type="relation" osm_id="146656" place_rank="16" lat="53.4791466" lon="-2.2447445" display_name="Manchester, Greater Manchester, North West England, England, United Kingdom" class="boundary" type="administrative" importance="0.704893333438333">
+      <city>Manchester</city>
+      <county>Greater Manchester</county>
+      <state_district>North West England</state_district>
+      <state>England</state>
+      <country>United Kingdom</country>
+      <country_code>gb</country_code>
+    </place>
+    <place place_id="77769745" osm_type="way" osm_id="104393803" place_rank="30" lat="52.5162024" lon="13.3777343363579" display_name="Brandenburg Gate, 1, Pariser Platz, Mitte, Berlin, 10117, Germany" class="tourism" type="attraction" importance="0.443472858361592">
+      <attraction>Brandenburg Gate</attraction>
+      <house_number>1</house_number>
+      <pedestrian>Pariser Platz</pedestrian>
+      <suburb>Mitte</suburb>
+      <city_district>Mitte</city_district>
+      <city>Berlin</city>
+      <state>Berlin</state>
+      <postcode>10117</postcode>
+      <country>Germany</country>
+      <country_code>de</country_code>
+    </place>
+    <place place_id="2570600569" osm_type="node" osm_id="240109189" place_rank="15" lat="52.5170365" lon="13.3888599" display_name="Berlin, Germany" class="place" type="city" importance="0.822149797630868">
+      <city>Berlin</city>
+      <state>Berlin</state>
+      <country>Germany</country>
+      <country_code>de</country_code>
+    </place>
+  </lookupresults>
+```
+
+##### JSON with extratags
+
+[https://nominatim.openstreetmap.org/lookup?osm_ids=W50637691&format=json](https://nominatim.openstreetmap.org/lookup?osm_ids=W50637691&format=json)
+
+```json
+[
+  {
+    "place_id": "84271358",
+    "licence": "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright",
+    "osm_type": "way",
+    "osm_id": "50637691",
+    "lat": "52.39955055",
+    "lon": "13.04806574678",
+    "display_name": "Brandenburger Tor, Brandenburger Straße, Nördliche Innenstadt, Innenstadt, Potsdam, Brandenburg, 14467, Germany",
+    "class": "historic",
+    "type": "city_gate",
+    "importance": "0.221233780277011",
+    "address": {
+      "address29": "Brandenburger Tor",
+      "pedestrian": "Brandenburger Straße",
+      "suburb": "Nördliche Innenstadt",
+      "city": "Potsdam",
+      "state": "Brandenburg",
+      "postcode": "14467",
+      "country": "Germany",
+      "country_code": "de"
+    },
+    "extratags": {
+      "image": "http://commons.wikimedia.org/wiki/File:Potsdam_brandenburger_tor.jpg",
+      "wikidata": "Q695045",
+      "wikipedia": "de:Brandenburger Tor (Potsdam)",
+      "wheelchair": "yes",
+      "description": "Kleines Brandenburger Tor in Potsdam"
+    }
+  }
+]
+```

docs/api/Output.md

@@ -0,0 +1,246 @@
+# Place Output
+
+The [\reverse](Reverse.md), [\search](Search.md) and [\lookup](Lookup.md)
+API calls produce very similar output which is explained in this section.
+There is one section for each format which is selectable via the `format`
+parameter.
+
+## Formats
+
+### JSON
+
+The JSON format returns an array of places (for search and lookup) or
+a single place (for reverse) of the following format:
+
+```
+  {
+    "place_id": "100149",
+    "licence": "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright",
+    "osm_type": "node",
+    "osm_id": "107775",
+    "boundingbox": ["51.3473219", "51.6673219", "-0.2876474", "0.0323526"],
+    "lat": "51.5073219",
+    "lon": "-0.1276474",
+    "display_name": "London, Greater London, England, SW1A 2DU, United Kingdom",
+    "class": "place",
+    "type": "city",
+    "importance": 0.9654895765402,
+    "icon": "https://nominatim.openstreetmap.org/images/mapicons/poi_place_city.p.20.png",
+    "address": {
+      "city": "London",
+      "state_district": "Greater London",
+      "state": "England",
+      "postcode": "SW1A 2DU",
+      "country": "United Kingdom",
+      "country_code": "gb"
+    },
+    "extratags": {
+      "capital": "yes",
+      "website": "http://www.london.gov.uk",
+      "wikidata": "Q84",
+      "wikipedia": "en:London",
+      "population": "8416535"
+    }
+  },
+```
+
+The possible fields are:
+
+ * `place_id` - reference to the Nominatim internal database ID (see notes below)
+ * `osm_type`, `osm_id` - reference to the OSM object
+ * `boundingbox` - area of corner coordinates
+ * `lat`, `lon` - latitude and longitude of the centroid of the object
+ * `display_name` - full comma-separated address
+ * `class`, `type` - key and value of the main OSM tag
+ * `importance` - computed importance rank
+ * `icon` - link to class icon (if available)
+ * `address` - dictionary of address details (only with `addressdetails=1`)
+ * `extratags` - dictionary with additional useful tags like website or maxspeed
+   (only with `extratags=1`)
+ * `namedetails` - dictionary with full list of available names including ref etc.
+ * `geojson`, `svg`, `geotext`, `geokml` - full geometry
+   (only with the appropriate `polygon_*` parameter)
+
+### JSONv2
+
+This is the same as the JSON format with two changes:
+
+ * `class` renamed to `category`
+ * additional field `place_rank` with the search rank of the object
+
+### GeoJSON
+
+This format follows the [RFC7946](http://geojson.org). Every feature includes
+a bounding box (`bbox`).
+
+The feature list has the following fields:
+
+ * `place_id` - reference to the Nominatim internal database ID (see notes below)
+ * `osm_type`, `osm_id` - reference to the OSM object
+ * `category`, `type` - key and value of the main OSM tag
+ * `display_name` - full comma-separated address
+ * `place_rank` - class search rank
+ * `importance` - computed importance rank
+ * `icon` - link to class icon (if available)
+ * `address` - dictionary of address details (only with `addressdetails=1`)
+ * `extratags` - dictionary with additional useful tags like website or maxspeed
+   (only with `extratags=1`)
+ * `namedetails` - dictionary with full list of available names including ref etc.
+
+Use `polygon_geojson` to output the full geometry of the object instead
+of the centroid.
+
+### GeocodeJSON
+
+The GeocodeJSON format follows the
+[GeocodeJSON spec 0.1.0](https://github.com/geocoders/geocodejson-spec).
+The following feature attributes are implemented:
+
+ * `osm_type`, `osm_id` - reference to the OSM object (unofficial extension)
+ * `type` - value of the main tag of the object (e.g. residential, restaurant, ...)
+ * `label` - full comma-separated address
+ * `name` - localised name of the place
+ * `housenumber`, `street`, `locality`, `postcode`, `city`,
+   `district`, `county`, `state`, `country` -
+   provided when it can be determined from the address
+   (see [this issue](https://github.com/openstreetmap/Nominatim/issues/1080) for
+   current limitations on the correctness of the address) and `addressdetails=1`
+   was given
+ * `admin` - list of localised names of administrative boundaries (only with `addressdetails=1`)
+
+Use `polygon_geojson` to output the full geometry of the object instead
+of the centroid.
+
+### XML
+
+The XML response returns one or more place objects in slightly different
+formats depending on the API call.
+
+#### Reverse
+
+```
+<reversegeocode timestamp="Sat, 11 Aug 18 11:53:21 +0000"
+                attribution="Data © OpenStreetMap contributors, ODbL 1.0. http://www.openstreetmap.org/copyright"
+                querystring="lat=48.400381&lon=11.745876&zoom=5&format=xml">
+  <result place_id="179509537" osm_type="relation" osm_id="2145268" ref="BY"
+          lat="48.9467562" lon="11.4038717"
+          boundingbox="47.2701114,50.5647142,8.9763497,13.8396373">
+       Bavaria, Germany
+  </result>
+  <addressparts>
+     <state>Bavaria</state>
+     <country>Germany</country>
+     <country_code>de</country_code>
+  </addressparts>
+  <extratags>
+    <tag key="place" value="state"/>
+    <tag key="wikidata" value="Q980"/>
+    <tag key="wikipedia" value="de:Bayern"/>
+    <tag key="population" value="12520000"/>
+    <tag key="name:prefix" value="Freistaat"/>
+  </extratags>
+</reversegeocode>
+```
+
+The attributes of the outer `reversegeocode` element return generic information
+about the query, including the time when the response was sent (in UTC),
+attribution to OSM and the original querystring.
+
+The place information can be found in the `result` element. The attributes of that element contain:
+
+ * `place_id` - reference to the Nominatim internal database ID (see notes below)
+ * `osm_type`, `osm_id` - reference to the OSM object
+ * `ref` - content of `ref` tag if it exists
+ * `lat`, `lon` - latitude and longitude of the centroid of the object
+ * `boundingbox` - comma-separated list of corner coordinates
+
+The full address address of the result can be found in the content of the
+`result` element as a comma-separated list.
+
+Additional information requested with `addressdetails=1`, `extratags=1` and
+`namedetails=1` can be found in extra elements.
+
+#### Search and Lookup
+
+```
+<searchresults timestamp="Sat, 11 Aug 18 11:55:35 +0000"
+               attribution="Data © OpenStreetMap contributors, ODbL 1.0. http://www.openstreetmap.org/copyright"
+               querystring="london" polygon="false" exclude_place_ids="100149"
+               more_url="https://nominatim.openstreetmap.org/search.php?q=london&addressdetails=1&extratags=1&exclude_place_ids=100149&format=xml&accept-language=en-US%2Cen%3Bq%3D0.7%2Cde%3Bq%3D0.3">
+  <place place_id="100149" osm_type="node" osm_id="107775" place_rank="15"
+         boundingbox="51.3473219,51.6673219,-0.2876474,0.0323526" lat="51.5073219" lon="-0.1276474"
+         display_name="London, Greater London, England, SW1A 2DU, United Kingdom" 
+         class="place" type="city" importance="0.9654895765402"
+         icon="https://nominatim.openstreetmap.org/images/mapicons/poi_place_city.p.20.png">
+    <extratags>
+      <tag key="capital" value="yes"/>
+      <tag key="website" value="http://www.london.gov.uk"/>
+      <tag key="wikidata" value="Q84"/>
+      <tag key="wikipedia" value="en:London"/>
+      <tag key="population" value="8416535"/>
+    </extratags>
+    <city>London</city>
+    <state_district>Greater London</state_district>
+    <state>England</state>
+    <postcode>SW1A 2DU</postcode>
+    <country>United Kingdom</country>
+    <country_code>gb</country_code>
+  </place>
+</searchresults>
+```
+
+The attributes of the outer `searchresults` or `lookupresults` element return
+generic information about the query:
+
+ * `timestamp` - UTC time when the response was sent
+ * `attribution` - OSM licensing information
+ * `querystring` - original query
+ * `polygon` - true when extra geometry information was requested
+ * `exclude_place_ids` - IDs of places that should be ignored in a follow-up request
+ * `more_url` - search call that will yield additional results for the query
+   just sent
+
+The place information can be found in the `place` elements, of which there may
+be more than one. The attributes of that element contain:
+
+ * `place_id` - reference to the Nominatim internal database ID (see notes below)
+ * `osm_type`, `osm_id` - reference to the OSM object
+ * `ref` - content of `ref` tag if it exists
+ * `lat`, `lon` - latitude and longitude of the centroid of the object
+ * `boundingbox` - comma-separated list of corner coordinates
+ * `place_rank` - class search rank
+ * `display_name` - full comma-separated address
+ * `class`, `type` - key and value of the main OSM tag
+ * `importance` - computed importance rank
+ * `icon` - link to class icon (if available)
+
+When `addressdetails=1` is requested, the localised address parts appear
+as subelements with the type of the address part.
+
+Additional information requested with `extratags=1` and `namedetails=1` can
+be found in extra elements as sub-element of each place.
+
+
+## Notes on field values
+
+### place_id is not a persistent id
+
+The `place_id` is created when a Nominatim database gets installed. A
+single place will have a different value on another server or even when
+the same data gets re-imported. It's thus not useful to treat it as
+permanent for later use.
+
+The combination `osm_type`+`osm_id` is slighly better but remember in
+OpenStreetMap mappers can delete, split, recreate places (and those
+get a new `osm_id`), there is no link between those old and new id.
+Places can also change their meaning without changing their `osm_id`,
+e.g. when a restaurant is retagged as supermarket. For a more in-depth
+discussion see [Permanent ID](https://wiki.openstreetmap.org/wiki/Permanent_ID).
+
+Nominatim merges some places (e.g. center node of a city with the boundary
+relation) so `osm_type`+`osm_id`+`class_name` would be more unique.
+
+### boundingbox
+
+Comma separated list of min latitude, max latitude, min longitude, max longitude.
+The whole planet would be `-90,90,-180,180`.

docs/api/Overview.md

@@ -0,0 +1,14 @@
+### Nominatim API
+
+Nominatim indexes named (or numbered) features within the OpenStreetMap (OSM) dataset and a subset of other unnamed features (pubs, hotels, churches, etc).
+
+Its API has the following endpoints for querying the data:
+
+ * __[/search](Search.md)__ - search OSM objects by name or type
+ * __[/reverse](Reverse.md)__ - search OSM object by their location
+ * __[/lookup](Lookup.md)__ - look up address details for OSM objects by their ID
+ * __/status__ - query the status of the server
+ * __/deletable__ - list objects that have been deleted in OSM but are held
+                    back in Nominatim in case the deletion was accidental
+ * __/polygons__ - list of broken polygons detected by Nominatim
+ * __[/details](Details.md)__ - show internal details for an object (for debugging only)

docs/api/Reverse.md

@@ -0,0 +1,272 @@
+# Reverse Geocoding
+
+Reverse geocoding generates an address from a latitude and longitude or from
+an OSM object.
+
+## Parameters
+
+The main format of the reverse API is
+
+```
+https://nominatim.openstreetmap.org/reverse?<query>
+```
+
+There are two ways how the requested location can be specified:
+
+* `lat=<value>` `lon=<value>`
+
+    A geographic location to generate an address for. The coordiantes must be
+    in WGS84 format.
+
+* `osm_type=[N|W|R]` `osm_id=<value>`
+
+    A specific OSM node(N), way(W) or relation(R) to return an address for.
+
+In both cases exactly one object is returned. The two input paramters cannot
+be used at the same time. Both accept the additional optional parameters listed
+below.
+
+### Output format
+
+* `format=[xml|json|jsonv2|geojson|geocodejson]`
+
+See [Place Output Formats](Output.md) for details on each format. (Default: html)
+
+* `json_callback=<string>`
+
+Wrap json output in a callback function ([JSONP](https://en.wikipedia.org/wiki/JSONP)) i.e. `<string>(<json>)`.
+Only has an effect for JSON output formats.
+
+### Output details
+
+* `addressdetails=[0|1]`
+
+Include a breakdown of the address into elements. (Default: 1)
+
+
+* `extratags=[0|1]`
+
+Include additional information in the result if available,
+e.g. wikipedia link, opening hours. (Default: 0)
+
+
+* `namedetails=[0|1]`
+
+Include a list of alternative names in the results. These may include
+language variants, references, operator and brand. (Default: 0)
+
+
+### Language of results
+
+* `accept-language=<browser language string>`
+
+Preferred language order for showing search results, overrides the value
+specified in the "Accept-Language" HTTP header.
+Either use a standard RFC2616 accept-language string or a simple
+comma-separated list of language codes.
+
+### Result limitation
+
+* `zoom=[0-18]`
+
+Level of detail required for the address. Default: 18. This is a number that corresponds
+roughly to the zoom level used in map frameworks like Leaflet.js, Openlayers etc.
+In terms of address details the zoom levels are as follows:
+
+ zoom | address detail
+ -----|---------------
+  3   | country
+  5   | state
+  8   | county
+  10  | city
+  14  | suburb
+  16  | major streets
+  17  | major and minor streets
+  18  | building
+
+
+### Polygon output
+
+* `polygon_geojson=1`
+* `polygon_kml=1`
+* `polygon_svg=1`
+* `polygon_text=1`
+
+Output geometry of results as a GeoJSON, KML, SVG or WKT. Only one of these
+options can be used at a time. (Default: 0)
+
+* `polygon_threshold=0.0`
+
+Simplify the output geometry before returning. The parameter is the
+tolerance in degrees with which the geometry may differ from the original
+geometry. Topology is preserved in the result. (Default: 0.0)
+
+### Other
+
+* `email=<valid email address>`
+
+If you are making large numbers of request please include an appropriate email
+address to identify your requests. See Nominatim's [Usage Policy](https://operations.osmfoundation.org/policies/nominatim/) for more details.
+
+
+* `debug=[0|1]`
+
+Output assorted developer debug information. Data on internals of Nominatim's
+"Search Loop" logic, and SQL queries. The output is (rough) HTML format.
+This overrides the specified machine readable format. (Default: 0)
+
+
+## Examples
+
+* [https://nominatim.openstreetmap.org/reverse?format=xml&lat=52.5487429714954&lon=-1.81602098644987&zoom=18&addressdetails=1](https://nominatim.openstreetmap.org/reverse?format=xml&lat=52.5487429714954&lon=-1.81602098644987&zoom=18&addressdetails=1)
+
+```xml
+  <reversegeocode timestamp="Fri, 06 Nov 09 16:33:54 +0000" querystring="...">
+    <result place_id="1620612" osm_type="node" osm_id="452010817">
+      135, Pilkington Avenue, Wylde Green, City of Birmingham, West Midlands (county), B72, United Kingdom
+    </result>
+    <addressparts>
+      <house_number>135</house_number>
+      <road>Pilkington Avenue</road>
+      <village>Wylde Green</village>
+      <town>Sutton Coldfield</town>
+      <city>City of Birmingham</city>
+      <county>West Midlands (county)</county>
+      <postcode>B72</postcode>
+      <country>United Kingdom</country>
+      <country_code>gb</country_code>
+    </addressparts>   
+  </reversegeocode>
+```
+
+##### Example with `format=jsonv2`
+
+* [https://nominatim.openstreetmap.org/reverse?format=jsonv2&lat=-34.44076&lon=-58.70521](https://nominatim.openstreetmap.org/reverse?format=jsonv2&lat=-34.44076&lon=-58.70521)
+
+```json
+{
+  "place_id":"134140761",
+  "licence":"Data © OpenStreetMap contributors, ODbL 1.0. http:\/\/www.openstreetmap.org\/copyright",
+  "osm_type":"way",
+  "osm_id":"280940520",
+"lat":"-34.4391708",
+  "lon":"-58.7064573",
+  "place_rank":"26",
+  "category":"highway",
+  "type":"motorway",
+  "importance":"0.1",
+  "addresstype":"road",
+  "display_name":"Autopista Pedro Eugenio Aramburu, El Triángulo, Partido de Malvinas Argentinas, Buenos Aires, 1.619, Argentina",
+  "name":"Autopista Pedro Eugenio Aramburu",
+  "address":{
+    "road":"Autopista Pedro Eugenio Aramburu",
+    "village":"El Triángulo",
+    "state_district":"Partido de Malvinas Argentinas",
+    "state":"Buenos Aires",
+    "postcode":"1.619",
+    "country":"Argentina",
+    "country_code":"ar"
+  },
+  "boundingbox":["-34.44159","-34.4370994","-58.7086067","-58.7044712"]
+}
+```
+
+##### Example with `format=geojson`
+
+* [https://nominatim.openstreetmap.org/reverse?format=geojson&lat=44.50155&lon=11.33989](https://nominatim.openstreetmap.org/reverse?format=geojson&lat=44.50155&lon=11.33989)
+
+```json
+{
+  "type": "FeatureCollection",
+  "licence": "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright",
+  "features": [
+    {
+      "type": "Feature",
+      "properties": {
+        "place_id": "18512203",
+        "osm_type": "node",
+        "osm_id": "1704756187",
+        "place_rank": "30",
+        "category": "place",
+        "type": "house",
+        "importance": "0",
+        "addresstype": "place",
+        "name": null,
+        "display_name": "71, Via Guglielmo Marconi, Saragozza-Porto, Bologna, BO, Emilia-Romagna, 40122, Italy",
+        "address": {
+          "house_number": "71",
+          "road": "Via Guglielmo Marconi",
+          "suburb": "Saragozza-Porto",
+          "city": "Bologna",
+          "county": "BO",
+          "state": "Emilia-Romagna",
+          "postcode": "40122",
+          "country": "Italy",
+          "country_code": "it"
+        }
+      },
+      "bbox": [
+        11.3397676,
+        44.5014307,
+        11.3399676,
+        44.5016307
+      ],
+      "geometry": {
+        "type": "Point",
+        "coordinates": [
+          11.3398676,
+          44.5015307
+        ]
+      }
+    }
+  ]
+}
+```
+
+##### Example with `format=geocodejson`
+
+[https://nominatim.openstreetmap.org/reverse?format=geocodejson&lat=60.2299&lon=11.1663](https://nominatim.openstreetmap.org/reverse?format=geocodejson&lat=60.2299&lon=11.1663)
+
+```json
+{
+  "type": "FeatureCollection",
+  "geocoding": {
+    "version": "0.1.0",
+    "attribution": "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright",
+    "licence": "ODbL",
+    "query": "60.229917843587,11.16630979382"
+  },
+  "features": {
+    "type": "Feature",
+    "properties": {
+      "geocoding": {
+        "place_id": "42700574",
+        "osm_type": "node",
+        "osm_id": "3110596255",
+        "type": "house",
+        "accuracy": 0,
+        "label": "1, Løvenbergvegen, Mogreina, Ullensaker, Akershus, 2054, Norway",
+        "name": null,
+        "housenumber": "1",
+        "street": "Løvenbergvegen",
+        "postcode": "2054",
+        "county": "Akershus",
+        "country": "Norway",
+        "admin": {
+          "level7": "Ullensaker",
+          "level4": "Akershus",
+          "level2": "Norway"
+        }
+      }
+    },
+    "geometry": {
+      "type": "Point",
+      "coordinates": [
+        11.1658572,
+        60.2301296
+      ]
+    }
+  }
+}
+```
+

docs/api/Search.md

@@ -0,0 +1,349 @@
+# Search queries
+
+The search API allows to look up a location from a textual description.
+Nominatim supports structured as well as free-form search queries.
+
+The search query may also contain
+[special phrases](https://wiki.openstreetmap.org/wiki/Nominatim/Special_Phrases)
+which are translated into specific OpenStreetMap (OSM) tags (e.g. Pub => `amenity=pub`).
+Note that this only limits the items to be found, it's not suited to return complete
+lists of OSM objects of a specific type. For those use [Overpass API](https://overpass-api.de/).
+
+## Parameters
+
+The search API has the following two formats:
+
+```
+   https://nominatim.openstreetmap.org/search/<query>?<params>
+```
+
+This format only accepts a free-form query string where the
+parts of the query are separated by slashes.
+
+```
+   https://nominatim.openstreetmap.org/search?<params>
+```
+
+In this form, the query may be given through two different sets of parameters:
+
+* `q=<query>`
+
+    Free-form query string to search for.
+    Free-form queries are processed first left-to-right and then right-to-left if that fails. So you may search for
+    [pilkington avenue, birmingham](//nominatim.openstreetmap.org/search?q=pilkington+avenue,birmingham) as well as for
+    [birmingham, pilkington avenue](//nominatim.openstreetmap.org/search?q=birmingham,+pilkington+avenue).
+    Commas are optional, but improve performance by reducing the complexity of the search.
+
+
+* `street=<housenumber> <streetname>`
+* `city=<city>`
+* `county=<county>`
+* `state=<state>`
+* `country=<country>`
+* `postalcode=<postalcode>`
+
+    Alternative query string format split into several parameters for structured requests.
+    Structured requests are faster but are less robust against alternative
+    OSM tagging schemas. **Do not combine with** `q=<query>` **parameter**.
+
+All three query forms accept the additional paramters listed below.
+
+### Output format
+
+* `format=[html|xml|json|jsonv2|geojson|geocodejson]`
+
+See [Place Output Formats](Output.md) for details on each format. (Default: html)
+
+* `json_callback=<string>`
+
+Wrap json output in a callback function ([JSONP](https://en.wikipedia.org/wiki/JSONP)) i.e. `<string>(<json>)`.
+Only has an effect for JSON output formats.
+
+### Output details
+
+* `addressdetails=[0|1]`
+
+Include a breakdown of the address into elements. (Default: 0)
+
+
+* `extratags=[0|1]`
+
+Include additional information in the result if available,
+e.g. wikipedia link, opening hours. (Default: 0)
+
+
+* `namedetails=[0|1]`
+
+Include a list of alternative names in the results. These may include
+language variants, references, operator and brand. (Default: 0)
+
+
+### Language of results
+
+* `accept-language=<browser language string>`
+
+Preferred language order for showing search results, overrides the value
+specified in the ["Accept-Language" HTTP header](https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Accept-Language).
+Either use a standard RFC2616 accept-language string or a simple
+comma-separated list of language codes.
+
+### Result limitation
+
+* `countrycodes=<countrycode>[,<countrycode>][,<countrycode>]...`
+
+Limit search results to one or more countries. `<countrycode>` must be the
+ISO 3166-1alpha2 code, e.g. `gb` for the United Kingdom, `de` for Germany.
+
+
+* `exclude_place_ids=<place_id,[place_id],[place_id]`
+
+If you do not want certain OSM objects to appear in the search
+result, give a comma separated list of the `place_id`s you want to skip.
+This can be used to broaden search results. For example, if a previous
+query only returned a few results, then including those here would cause
+the search to return other, less accurate, matches (if possible).
+
+
+* `limit=<integer>`
+
+Limit the number of returned results. (Default: 10, Maximum: 50)
+
+
+* `viewbox=<x1>,<y1>,<x2>,<y2>`
+
+The preferred area to find search results. Any two corner points of the box
+are accepted in any order as long as they span a real box.
+
+
+* `bounded=[0|1]`
+
+When a viewbox is given, restrict the result to items contained with that
+viewbox (see above). When `viewbox` and `bounded=1` are given, an amenity
+only search is allowed. In this case, give the special keyword for the
+amenity in square brackets, e.g. `[pub]`. (Default: 0)
+
+
+### Polygon output
+
+* `polygon_geojson=1`
+* `polygon_kml=1`
+* `polygon_svg=1`
+* `polygon_text=1`
+
+Output geometry of results as a GeoJSON, KML, SVG or WKT. Only one of these
+options can be used at a time. (Default: 0)
+
+* `polygon_threshold=0.0`
+
+Simplify the output geometry before returning. The parameter is the
+tolerance in degrees with which the geometry may differ from the original
+geometry. Topology is preserved in the result. (Default: 0.0)
+
+### Other
+
+* `email=<valid email address>`
+
+If you are making large numbers of request please include an appropriate email
+address to identify your requests. See Nominatim's [Usage Policy](https://operations.osmfoundation.org/policies/nominatim/) for more details.
+
+* `dedupe=[0|1]`
+
+Sometimes you have several objects in OSM identifying the same place or
+object in reality. The simplest case is a street being split in many
+different OSM ways due to different characteristics. Nominatim will
+attempt to detect such duplicates and only return one match unless
+this parameter is set to 0. (Default: 1)
+
+
+
+* `debug=[0|1]`
+
+Output assorted developer debug information. Data on internals of Nominatim's
+"Search Loop" logic, and SQL queries. The output is (rough) HTML format.
+This overrides the specified machine readable format. (Default: 0)
+
+
+
+## Examples
+
+
+##### XML with polygon points
+
+* [https://nominatim.openstreetmap.org/search?q=135+pilkington+avenue,+birmingham&format=xml&polygon=1&addressdetails=1](https://nominatim.openstreetmap.org/search?q=135+pilkington+avenue,+birmingham&format=xml&polygon=1&addressdetails=1)
+* [https://nominatim.openstreetmap.org/search/gb/birmingham/pilkington%20avenue/135?format=xml&polygon=1&addressdetails=1](https://nominatim.openstreetmap.org/search/gb/birmingham/pilkington%20avenue/135?format=xml&polygon=1&addressdetails=1)
+* [https://nominatim.openstreetmap.org/search/135%20pilkington%20avenue,%20birmingham?format=xml&polygon=1&addressdetails=1](https://nominatim.openstreetmap.org/search/135%20pilkington%20avenue,%20birmingham?format=xml&polygon=1&addressdetails=1)
+
+```xml
+  <searchresults timestamp="Sat, 07 Nov 09 14:42:10 +0000" querystring="135 pilkington, avenue birmingham" polygon="true">
+    <place 
+      place_id="1620612" osm_type="node" osm_id="452010817" 
+      boundingbox="52.548641204834,52.5488433837891,-1.81612110137939,-1.81592094898224" 
+      polygonpoints="[['-1.81592098644987','52.5487429714954'],['-1.81592290792183','52.5487234624632'],...]" 
+      lat="52.5487429714954" lon="-1.81602098644987" 
+      display_name="135, Pilkington Avenue, Wylde Green, City of Birmingham, West Midlands (county), B72, United Kingdom" 
+      class="place" type="house">
+      <house_number>135</house_number>
+      <road>Pilkington Avenue</road>
+      <village>Wylde Green</village>
+      <town>Sutton Coldfield</town>
+      <city>City of Birmingham</city>
+      <county>West Midlands (county)</county>
+      <postcode>B72</postcode>
+      <country>United Kingdom</country>
+      <country_code>gb</country_code>
+    </place>
+  </searchresults>
+```
+
+##### JSON with SVG polygon
+
+[https://nominatim.openstreetmap.org/search/Unter%20den%20Linden%201%20Berlin?format=json&addressdetails=1&limit=1&polygon_svg=1](https://nominatim.openstreetmap.org/search/Unter%20den%20Linden%201%20Berlin?format=json&addressdetails=1&limit=1&polygon_svg=1)
+
+```json
+    {
+        "address": {
+            "city": "Berlin",
+            "city_district": "Mitte",
+            "construction": "Unter den Linden",
+            "continent": "European Union",
+            "country": "Deutschland",
+            "country_code": "de",
+            "house_number": "1",
+            "neighbourhood": "Scheunenviertel",
+            "postcode": "10117",
+            "public_building": "Kommandantenhaus",
+            "state": "Berlin",
+            "suburb": "Mitte"
+        },
+        "boundingbox": [
+            "52.5170783996582",
+            "52.5173187255859",
+            "13.3975105285645",
+            "13.3981599807739"
+        ],
+        "class": "amenity",
+        "display_name": "Kommandantenhaus, 1, Unter den Linden, Scheunenviertel, Mitte, Berlin, 10117, Deutschland, European Union",
+        "importance": 0.73606775332943,
+        "lat": "52.51719785",
+        "licence": "Data \u00a9 OpenStreetMap contributors, ODbL 1.0. https://www.openstreetmap.org/copyright",
+        "lon": "13.3978352028938",
+        "osm_id": "15976890",
+        "osm_type": "way",
+        "place_id": "30848715",
+        "svg": "M 13.397511 -52.517283599999999 L 13.397829400000001 -52.517299800000004 13.398131599999999 -52.517315099999998 13.398159400000001 -52.517112099999999 13.3975388 -52.517080700000001 Z",
+        "type": "public_building"
+    }
+```
+
+##### JSON with address details
+
+[https://nominatim.openstreetmap.org/?format=json&addressdetails=1&q=bakery+in+berlin+wedding&format=json&limit=1](https://nominatim.openstreetmap.org/?format=json&addressdetails=1&q=bakery+in+berlin+wedding&format=json&limit=1)
+
+```json
+    {
+        "address": {
+            "bakery": "B\u00e4cker Kamps",
+            "city_district": "Mitte",
+            "continent": "European Union",
+            "country": "Deutschland",
+            "country_code": "de",
+            "footway": "Bahnsteig U6",
+            "neighbourhood": "Sprengelkiez",
+            "postcode": "13353",
+            "state": "Berlin",
+            "suburb": "Wedding"
+        },
+        "boundingbox": [
+            "52.5460929870605",
+            "52.5460968017578",
+            "13.3591794967651",
+            "13.3591804504395"
+        ],
+        "class": "shop",
+        "display_name": "B\u00e4cker Kamps, Bahnsteig U6, Sprengelkiez, Wedding, Mitte, Berlin, 13353, Deutschland, European Union",
+        "icon": "https://nominatim.openstreetmap.org/images/mapicons/shopping_bakery.p.20.png",
+        "importance": 0.201,
+        "lat": "52.5460941",
+        "licence": "Data \u00a9 OpenStreetMap contributors, ODbL 1.0. https://www.openstreetmap.org/copyright",
+        "lon": "13.35918",
+        "osm_id": "317179427",
+        "osm_type": "node",
+        "place_id": "1453068",
+        "type": "bakery"
+    }
+```
+
+##### GeoJSON
+
+[https://nominatim.openstreetmap.org/search?q=17+Strada+Pictor+Alexandru+Romano%2C+Bukarest&format=geojson](https://nominatim.openstreetmap.org/search?q=17+Strada+Pictor+Alexandru+Romano%2C+Bukarest&format=geojson)
+
+```json
+{
+  "type": "FeatureCollection",
+  "licence": "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright",
+  "features": [
+    {
+      "type": "Feature",
+      "properties": {
+        "place_id": "35811445",
+        "osm_type": "node",
+        "osm_id": "2846295644",
+        "display_name": "17, Strada Pictor Alexandru Romano, Bukarest, Bucharest, Sector 2, Bucharest, 023964, Romania",
+        "place_rank": "30",
+        "category": "place",
+        "type": "house",
+        "importance": 0.62025
+      },
+      "bbox": [
+        26.1156689,
+        44.4354754,
+        26.1157689,
+        44.4355754
+      ],
+      "geometry": {
+        "type": "Point",
+        "coordinates": [
+          26.1157189,
+          44.4355254
+        ]
+      }
+    }
+  ]
+}
+```
+
+##### GeocodeJSON
+
+[https://nominatim.openstreetmap.org/search?q=%CE%91%CE%B3%CE%AF%CE%B1+%CE%A4%CF%81%CE%B9%CE%AC%CE%B4%CE%B1%2C+%CE%91%CE%B4%CF%89%CE%BD%CE%B9%CE%B4%CE%BF%CF%82%2C+Athens%2C+Greece&format=geocodejson](https://nominatim.openstreetmap.org/search?q=%CE%91%CE%B3%CE%AF%CE%B1+%CE%A4%CF%81%CE%B9%CE%AC%CE%B4%CE%B1%2C+%CE%91%CE%B4%CF%89%CE%BD%CE%B9%CE%B4%CE%BF%CF%82%2C+Athens%2C+Greece&format=geocodejson)
+
+```json
+{
+  "type": "FeatureCollection",
+  "geocoding": {
+    "version": "0.1.0",
+    "attribution": "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright",
+    "licence": "ODbL",
+    "query": "Αγία Τριάδα, Αδωνιδος, Athens, Greece"
+  },
+  "features": [
+    {
+      "type": "Feature",
+      "properties": {
+        "geocoding": {
+          "type": "place_of_worship",
+          "label": "Αγία Τριάδα, Αδωνιδος, Άγιος Νικόλαος, 5º Δημοτικό Διαμέρισμα Αθηνών, Athens, Municipality of Athens, Regional Unit of Central Athens, Region of Attica, Attica, 11472, Greece",
+          "name": "Αγία Τριάδα",
+          "admin": null
+        }
+      },
+      "geometry": {
+        "type": "Point",
+        "coordinates": [
+          23.72949633941,
+          38.0051697
+        ]
+      }
+    }
+  ]
+}
+```

docs/data-sources/overview.md

@@ -0,0 +1,4 @@
+# Additional Data Sources
+
+This guide explains how data sources other than OpenStreetMap mentioned in
+the install instructions got obtained and converted.

docs/develop/Import.md

@@ -0,0 +1,170 @@
+# OSM Data Import
+
+OSM data is initially imported using osm2pgsql. Nominatim uses its own data
+output style 'gazetteer', which differs from the output style created for
+map rendering.
+
+## Database Layout
+
+The gazetteer style produces a single table `place` with the following rows:
+
+ * `osm_type` - kind of OSM object (**N** - node, **W** - way, **R** - relation)
+ * `osm_id` - original OSM ID
+ * `class` - key of principal tag defining the object type
+ * `type` - value of principal tag defining the object type
+ * `name` - collection of tags that contain a name or reference
+ * `admin_level` - numerical value of the tagged administrative level
+ * `address` - collection of tags defining the address of an object
+ * `extratags` - collection of additional interesting tags that are not
+                 directly relevant for searching
+ * `geometry` - geometry of the object (in WGS84)
+
+A single OSM object may appear multiple times in this table when it is tagged
+with multiple tags that may constitute a principal tag. Take for example a
+motorway bridge. In OSM, this would be a way which is tagged with
+`highway=motorway` and `bridge=yes`. This way would appear in the `place` table
+once with `class` of `highway` and once with a `class` of `bridge`. Thus the
+*uique key* for `place` is (`osm_type`, `osm_id`, `class`).
+
+## Configuring the Import
+
+How tags are interpreted and assigned to the different `place` columns can be
+configured via the import style configuration file (`CONST_Import_style`). This
+is a JSON file which contains a list of rules which are matched against every
+tag of every object and then assign the tag its specific role.
+
+### Configuration Rules
+
+A single rule looks like this:
+
+```json
+{
+    "keys" : ["key1", "key2", ...],
+    "values" : {
+        "value1" : "prop",
+        "value2" : "prop1,prop2"
+    }
+}
+```
+
+A rule first defines a list of keys to apply the rule to. This is always a list
+of strings. The string may have four forms. An empty string matches against
+any key. A string that ends in an asterisk `*` is a prefix match and accordingly
+matches against any key that starts with the given string (minus the `*`). A
+suffix match can be defined similarly with a string that starts with a `*`. Any
+other string constitutes an exact match.
+
+The second part of the rules defines a list of values and the properties that
+apply to a successful match. Value strings may be either empty, which again
+means that thy match against any value, or describe an exact match. Prefix
+or suffix matching of values is not possible.
+
+For a rule to match, it has to find a valid combination of keys and values. The
+resulting property is that of the matched values.
+
+The rules in a configuration file are processed sequentially and the first
+match for each tag wins.
+
+A rule where key and value are the empty string is special. This defines the
+fallback when none of the rules matches. The fallback is always used as a last
+resort when nothing else matches, no matter where the rule appears in the file.
+Defining multiple fallback rules is not allowed. What happens in this case,
+is undefined.
+
+### Tag Properties
+
+One or more of the following properties may be given for each tag:
+
+* `main`
+
+    A principal tag. A new row will be added for the object with key and value
+    as `class` and `type`.
+
+* `with_name`
+
+    When the tag is a principal tag (`main` property set): only really add a new
+    row, if there is any name tag found (a reference tag is not sufficient, see
+    below).
+
+* `with_name_key`
+
+    When the tag is a principal tag (`main` property set): only really add a new
+    row, if there is also a name tag that matches the key of the principal tag.
+    For example, if the main tag is `bridge=yes`, then it will only be added as
+    an extra row, if there is a tag `bridge:name[:XXX]` for the same object.
+    If this property is set, all other names that are not domain-specific are
+    ignored.
+
+* `fallback`
+
+    When the tag is a principal tag (`main` property set): only really add a new
+    row, when no other principal tags for this object have been found. Only one
+    fallback tag can win for an object.
+
+* `operator`
+
+    When the tag is a principal tag (`main` property set): also include the
+    `operator` tag in the list of names. This is a special construct for an
+    out-dated tagging practise in OSM. Fuel stations and chain restaurants
+    in particular used to have the name of the chain tagged as `operator`.
+    These days the chain can be more commonly found in the `brand` tag but
+    there is still enough old data around to warrant this special case.
+
+* `name`
+
+    Add tag to the list of names.
+
+* `ref`
+
+    Add tag to the list of names as a reference. At the moment this only means
+    that the object is not considered to be named for `with_name`.
+
+* `address`
+
+    At tag to the list of address tags. If the tag starts with `addr:` or
+    `is_in:`, then this prefix is cut off before adding it to the list.
+
+* `postcode`
+
+    At the value as a postcode to the address tags. If multiple tags are
+    candidate for postcodes, one wins out and the others are dropped.
+
+* `country`
+
+    At the value as a country code to the address tags. The value must be a
+    two letter country code, otherwise it is ignored. If there are multiple
+    tags that match, then one wins out and the others are dropped.
+
+* `house`
+
+    If no principle tags can be found for the object, still add the object with
+    `class`=`place` and `type`=`house`. Use this for address nodes that have no
+    other function.
+
+* `interpolation`
+
+    Add this object as an address interpolation (appears as `class`=`place` and
+    `type`=`houses` in the database).
+
+* `extra`
+
+    Add tag to the list of extra tags.
+
+* `skip`
+
+    Skip the tag completely. Useful when a custom default fallback is defined
+    or to define exceptions to rules.
+
+A rule can define as many of these properties for one match as it likes. For
+example, if the property is `"main,extra"` then the tag will open a new row
+but also have the tag appear in the list of extra tags.
+
+There are a number of pre-defined styles in the `settings/` directory. It is
+advisable to start from one of these styles when defining your own.
+
+### Changing the Style of Existing Databases
+
+There is normally no issue changing the style of a database that is already
+imported and now kept up-to-date with change files. Just be aware that any
+change in the style applies to updates only. If you want to change the data
+that is already in the database, then a reimport is necessary.

docs/develop/Ranking.md

@@ -0,0 +1,89 @@
+# Place Ranking in Nominatim
+
+Nominatim uses two metrics to rank a place: search rank and address rank.
+Both can be assigned a value between 0 and 30. They serve slightly
+different purposes, which are explained in this chapter.
+
+## Search rank
+
+The search rank describes the extent and importance of a place. It is used
+when ranking search result. Simply put, if there are two results for a
+search query which are otherwise equal, then the result with the _lower_
+search rank will be appear higher in the result list.
+
+Search ranks are not so important these days because many well-known
+places use the Wikipedia importance ranking instead.
+
+## Address rank
+
+The address rank describes where a place shows up in an address hierarchy.
+Usually only administrative boundaries and place nodes and areas are
+eligible to be part of an address. All other objects have an address rank
+of 0.
+
+Note that the search rank of a place place a role in the address computation
+as well. When collecting the places that should make up the address parts
+then only places are taken into account that have a lower address rank than
+the search rank of the base object.
+
+## Rank configuration
+
+Search and address ranks are assigned to a place when it is first imported
+into the database. There are a few hard-coded rules for the assignment:
+
+ * postcodes follow special rules according to their length
+ * boundaries that are not areas and railway=rail are dropped completely
+ * the following are always search rank 30 and address rank 0:
+    * highway nodes
+    * landuse that is not an area
+
+Other than that, the ranks can be freely assigned via the json file
+defined with `CONST_Address_Level_Config` according to their type and
+the country they are in.
+
+The address level configuration must consist of an array of configuration
+entries, each containing a tag definition and an optional country array:
+
+```
+[ {
+    "tags" : {
+      "place" : {
+        "county" : 12,
+        "city" : 16,
+      },
+      "landuse" : {
+        "residential" : 22,
+        "" : 30
+      }
+    }
+  },
+  {
+    "countries" : [ "ca", "us" ],
+    "tags" : {
+      "boundary" : {
+        "administrative8" : 18,
+        "administrative9" : 20
+      },
+      "landuse" : {
+        "residential" : [22, 0]
+      }
+    }
+  }
+]
+```
+
+The `countries` field contains a list of countries (as ISO 3166-1 alpha 2 code)
+for which the definition applies. When the field is omitted, then the
+definition is used as a fallback, when nothing more specific for a given
+country exists.
+
+`tags` contains the ranks for key/value pairs. The ranks can be either a
+single number, in which case they are to search and address rank, or a tuple
+of search and address rank (in that order). The value may be left empty.
+Then the rank is used when no more specific value is found for the given
+key.
+
+Countries and key/value combination may appear in multiple defintions. Just
+make sure that each combination of counrty/key/value appears only once per
+file. Otherwise the import will fail with a UNIQUE INDEX constraint violation
+on import.

docs/develop/overview.md

@@ -0,0 +1,24 @@
+# Basic Architecture
+
+Nominatim provides geocoding based on OpenStreetMap data. It uses a Postgresql
+database as a backend for storing the data.
+
+There are three basic parts to Nominatim's architecture: the data import,
+the address computation and the search frontend.
+
+The __data import__ stage reads the raw OSM data and extracts all information
+that is useful for geocoding. This part is done by osm2pgsql, the same tool
+that can also be used to import a rendering database. It uses the special
+gazetteer output plugin in `osm2pgsql/output-gazetter.[ch]pp`. The result of
+the import can be found in the database table `place`.
+
+The __address computation__ or __indexing__ stage takes the data from `place`
+and adds additional information needed for geocoding. It ranks the places by
+importance, links objects that belong together and computes addresses and
+the search index. Most of this work is done in Pl/pqSQL via database triggers
+and can be found in the file `sql/functions.sql`.
+
+The __search frontend__ implements the actual API. It takes queries for
+search and reverse geocoding queries from the user, looks up the data and
+returns the results in the requested format. This part is written in PHP
+and can be found in the `lib/` and `website/` directories.

docs/extra.css

@@ -0,0 +1,3 @@
+.toctree-l3 {
+    display: none!important
+}

docs/index.md

@@ -1 +1,8 @@
 Nominatim (from the Latin, 'by name') is a tool to search OSM data by name and address and to generate synthetic addresses of OSM points (reverse geocoding).
+
+This guide comes in three parts:
+
+ * __[API reference](api/Overview.md)__ for users of Nominatim
+ * __[Administration Guide](admin/Installation.md)__ for those who want
+   to install their own Nominatim server
+ * __[Developer's Guide](develop/overview.md)__ for developers of the software

docs/mkdocs.yml

@@ -5,14 +5,35 @@ site_url: http://nominatim.org
 repo_url: https://github.com/openstreetmap/Nominatim
 pages:
     - 'Introduction' : 'index.md'
+    - 'API Reference':
+        - 'Overview': 'api/Overview.md'
+        - 'Search': 'api/Search.md'
+        - 'Reverse': 'api/Reverse.md'
+        - 'Address Lookup': 'api/Lookup.md'
+        - 'Details' : 'api/Details.md'
+        - 'Place Output Formats': 'api/Output.md'
+        - 'FAQ': 'api/Faq.md'
     - 'Administration Guide':
         - 'Basic Installation': 'admin/Installation.md'
         - 'Importing and Updating' : 'admin/Import-and-Update.md'
         - 'Migration from older Versions' : 'admin/Migration.md'
         - 'Troubleshooting' : 'admin/Faq.md'
+    - 'Developers Guide':
+        - 'Overview' : 'develop/overview.md'
+        - 'OSM Data Import' : 'develop/Import.md'
+        - 'Place Ranking' : 'develop/Ranking.md'
+    - 'External Data Sources':
+        - 'Overview' : 'data-sources/overview.md'
+        - 'US Census (Tiger)': 'data-sources/US-Tiger.md'
+        - 'GB Postcodes': 'data-sources/GB-Postcodes.md'
+        - 'Country Grid': 'data-sources/Country-Grid.md'
     - 'Appendix':
         - 'Installation on CentOS 7' : 'appendix/Install-on-Centos-7.md'
         - 'Installation on Ubuntu 16' : 'appendix/Install-on-Ubuntu-16.md'
+        - 'Installation on Ubuntu 18' : 'appendix/Install-on-Ubuntu-18.md'
 markdown_extensions:
     - codehilite:
         use_pygments: False
+    - toc:
+        permalink: 
+extra_css: [extra.css]

lib/AddressDetails.php

@@ -0,0 +1,122 @@
+<?php
+
+namespace Nominatim;
+
+require_once(CONST_BasePath.'/lib/ClassTypes.php');
+
+/**
+ * Detailed list of address parts for a single result
+ */
+class AddressDetails
+{
+    private $aAddressLines;
+
+    public function __construct(&$oDB, $iPlaceID, $sHousenumber, $mLangPref)
+    {
+        if (is_array($mLangPref)) {
+            $mLangPref = 'ARRAY['.join(',', array_map('getDBQuoted', $mLangPref)).']';
+        }
+
+        if (!isset($sHousenumber)) {
+            $sHousenumber = -1;
+        }
+
+        $sSQL = 'SELECT *,';
+        $sSQL .= ' get_name_by_language(name,'.$mLangPref.') as localname';
+        $sSQL .= ' FROM get_addressdata('.$iPlaceID.','.$sHousenumber.')';
+        $sSQL .= ' ORDER BY rank_address DESC, isaddress DESC';
+
+        $this->aAddressLines = $oDB->getAll($sSQL);
+    }
+
+    private static function isAddress($aLine)
+    {
+        return $aLine['isaddress'] || $aLine['type'] == 'country_code';
+    }
+
+    public function getAddressDetails($bAll = false)
+    {
+        if ($bAll) {
+            return $this->aAddressLines;
+        }
+
+        return array_filter($this->aAddressLines, array(__CLASS__, 'isAddress'));
+    }
+
+    public function getLocaleAddress()
+    {
+        $aParts = array();
+        $sPrevResult = '';
+
+        foreach ($this->aAddressLines as $aLine) {
+            if ($aLine['isaddress'] && $sPrevResult != $aLine['localname']) {
+                $sPrevResult = $aLine['localname'];
+                $aParts[] = $sPrevResult;
+            }
+        }
+
+        return join(', ', $aParts);
+    }
+
+    public function getAddressNames()
+    {
+        $aAddress = array();
+        $aFallback = array();
+
+        foreach ($this->aAddressLines as $aLine) {
+            if (!self::isAddress($aLine)) {
+                continue;
+            }
+
+            $bFallback = false;
+            $aTypeLabel = ClassTypes\getInfo($aLine);
+
+            if ($aTypeLabel === false) {
+                $aTypeLabel = ClassTypes\getFallbackInfo($aLine);
+                $bFallback = true;
+            }
+
+            $sName = false;
+            if (isset($aLine['localname']) && $aLine['localname']) {
+                $sName = $aLine['localname'];
+            } elseif (isset($aLine['housenumber']) && $aLine['housenumber']) {
+                $sName = $aLine['housenumber'];
+            }
+
+            if ($sName) {
+                $sTypeLabel = strtolower(isset($aTypeLabel['simplelabel']) ? $aTypeLabel['simplelabel'] : $aTypeLabel['label']);
+                $sTypeLabel = str_replace(' ', '_', $sTypeLabel);
+                if (!isset($aAddress[$sTypeLabel])
+                    || isset($aFallback[$sTypeLabel])
+                    || $aLine['class'] == 'place'
+                ) {
+                    $aAddress[$sTypeLabel] = $sName;
+                    if ($bFallback) {
+                        $aFallback[$sTypeLabel] = $bFallback;
+                    }
+                }
+            }
+        }
+        return $aAddress;
+    }
+
+    public function getAdminLevels()
+    {
+        $aAddress = array();
+        foreach (array_reverse($this->aAddressLines) as $aLine) {
+            if (self::isAddress($aLine)
+                && isset($aLine['admin_level'])
+                && $aLine['admin_level'] < 15
+                && !isset($aAddress['level'.$aLine['admin_level']])
+            ) {
+                $aAddress['level'.$aLine['admin_level']] = $aLine['localname'];
+            }
+        }
+        return $aAddress;
+    }
+
+    public function debugInfo()
+    {
+        return $this->aAddressLines;
+    }
+}

lib/ClassTypes.php

@@ -0,0 +1,374 @@
+<?php
+
+namespace Nominatim\ClassTypes;
+
+function getInfo($aPlace)
+{
+    $aClassType = getList();
+
+    if (isset($aPlace['admin_level'])) {
+        $sName = $aPlace['class'].':'.$aPlace['type'].':'.$aPlace['admin_level'];
+        if (isset($aClassType[$sName])) {
+            return $aClassType[$sName];
+        }
+    }
+
+    $sName = $aPlace['class'].':'.$aPlace['type'];
+    if (isset($aClassType[$sName])) {
+        return $aClassType[$sName];
+    }
+
+    return false;
+}
+
+function getFallbackInfo($aPlace)
+{
+    $aClassType = getList();
+
+    $sFallback = 'boundary:administrative:'.((int)($aPlace['rank_address']/2));
+    if (isset($aClassType[$sFallback])) {
+        return $aClassType[$sFallback];
+    }
+
+    return array('simplelabel' => 'address'.$aPlace['rank_address']);
+}
+
+function getProperty($aPlace, $sProp, $mDefault = false)
+{
+    $aClassType = getList();
+
+    if (isset($aPlace['admin_level'])) {
+        $sName = $aPlace['class'].':'.$aPlace['type'].':'.$aPlace['admin_level'];
+        if (isset($aClassType[$sName]) && isset($aClassType[$sName][$sProp])) {
+            return $aClassType[$sName][$sProp];
+        }
+    }
+
+    $sName = $aPlace['class'].':'.$aPlace['type'];
+    if (isset($aClassType[$sName]) && isset($aClassType[$sName][$sProp])) {
+        return $aClassType[$sName][$sProp];
+    }
+
+    return $mDefault;
+}
+
+function getListWithImportance()
+{
+    static $aOrders = null;
+    if ($aOrders === null) {
+        $aOrders = getList();
+        $i = 1;
+        foreach ($aOrders as $sID => $a) {
+            $aOrders[$sID]['importance'] = $i++;
+        }
+    }
+
+    return $aOrders;
+}
+
+function getList()
+{
+    return array(
+            'boundary:administrative:1' => array('label' => 'Continent', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'boundary:administrative:2' => array('label' => 'Country', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'place:country' => array('label' => 'Country', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defzoom' => 6, 'defdiameter' => 15),
+            'boundary:administrative:3' => array('label' => 'State', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'boundary:administrative:4' => array('label' => 'State', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'place:state' => array('label' => 'State', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defzoom' => 8, 'defdiameter' => 5.12),
+            'boundary:administrative:5' => array('label' => 'State District', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'boundary:administrative:6' => array('label' => 'County', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'boundary:administrative:7' => array('label' => 'County', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'place:county' => array('label' => 'County', 'frequency' => 108, 'icon' => 'poi_boundary_administrative', 'defzoom' => 10, 'defdiameter' => 1.28),
+            'boundary:administrative:8' => array('label' => 'City', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'place:city' => array('label' => 'City', 'frequency' => 66, 'icon' => 'poi_place_city', 'defzoom' => 12, 'defdiameter' => 0.32),
+            'boundary:administrative:9' => array('label' => 'City District', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'boundary:administrative:10' => array('label' => 'Suburb', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'boundary:administrative:11' => array('label' => 'Neighbourhood', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'place:region' => array('label' => 'Region', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defzoom' => 8, 'defdiameter' => 0.04),
+            'place:island' => array('label' => 'Island', 'frequency' => 288, 'defzoom' => 11, 'defdiameter' => 0.64),
+            'boundary:administrative' => array('label' => 'Administrative', 'frequency' => 413, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'boundary:postal_code' => array('label' => 'Postcode', 'frequency' => 413, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
+            'place:town' => array('label' => 'Town', 'frequency' => 1497, 'icon' => 'poi_place_town', 'defzoom' => 14, 'defdiameter' => 0.08),
+            'place:village' => array('label' => 'Village', 'frequency' => 11230, 'icon' => 'poi_place_village', 'defzoom' => 15, 'defdiameter' => 0.04),
+            'place:hamlet' => array('label' => 'Hamlet', 'frequency' => 7075, 'icon' => 'poi_place_village', 'defzoom' => 15, 'defdiameter' => 0.04),
+            'place:suburb' => array('label' => 'Suburb', 'frequency' => 2528, 'icon' => 'poi_place_village', 'defdiameter' => 0.04),
+            'place:locality' => array('label' => 'Locality', 'frequency' => 4113, 'icon' => 'poi_place_village', 'defdiameter' => 0.02),
+            'landuse:farm' => array('label' => 'Farm', 'frequency' => 1201, 'defdiameter' => 0.02),
+            'place:farm' => array('label' => 'Farm', 'frequency' => 1162, 'defdiameter' => 0.02),
+
+            'highway:motorway_junction' => array('label' => 'Motorway Junction', 'frequency' => 1126, 'simplelabel' => 'Junction'),
+            'highway:motorway' => array('label' => 'Motorway', 'frequency' => 4627, 'simplelabel' => 'Road'),
+            'highway:trunk' => array('label' => 'Trunk', 'frequency' => 23084, 'simplelabel' => 'Road'),
+            'highway:primary' => array('label' => 'Primary', 'frequency' => 32138, 'simplelabel' => 'Road'),
+            'highway:secondary' => array('label' => 'Secondary', 'frequency' => 25807, 'simplelabel' => 'Road'),
+            'highway:tertiary' => array('label' => 'Tertiary', 'frequency' => 29829, 'simplelabel' => 'Road'),
+            'highway:residential' => array('label' => 'Residential', 'frequency' => 361498, 'simplelabel' => 'Road'),
+            'highway:unclassified' => array('label' => 'Unclassified', 'frequency' => 66441, 'simplelabel' => 'Road'),
+            'highway:living_street' => array('label' => 'Living Street', 'frequency' => 710, 'simplelabel' => 'Road'),
+            'highway:service' => array('label' => 'Service', 'frequency' => 9963, 'simplelabel' => 'Road'),
+            'highway:track' => array('label' => 'Track', 'frequency' => 2565, 'simplelabel' => 'Road'),
+            'highway:road' => array('label' => 'Road', 'frequency' => 591, 'simplelabel' => 'Road'),
+            'highway:byway' => array('label' => 'Byway', 'frequency' => 346, 'simplelabel' => 'Road'),
+            'highway:bridleway' => array('label' => 'Bridleway', 'frequency' => 1556),
+            'highway:cycleway' => array('label' => 'Cycleway', 'frequency' => 2419),
+            'highway:pedestrian' => array('label' => 'Pedestrian', 'frequency' => 2757),
+            'highway:footway' => array('label' => 'Footway', 'frequency' => 15008),
+            'highway:steps' => array('label' => 'Steps', 'frequency' => 444, 'simplelabel' => 'Footway'),
+            'highway:motorway_link' => array('label' => 'Motorway Link', 'frequency' => 795, 'simplelabel' => 'Road'),
+            'highway:trunk_link' => array('label' => 'Trunk Link', 'frequency' => 1258, 'simplelabel' => 'Road'),
+            'highway:primary_link' => array('label' => 'Primary Link', 'frequency' => 313, 'simplelabel' => 'Road'),
+
+            'landuse:industrial' => array('label' => 'Industrial', 'frequency' => 1062),
+            'landuse:residential' => array('label' => 'Residential', 'frequency' => 886),
+            'landuse:retail' => array('label' => 'Retail', 'frequency' => 754),
+            'landuse:commercial' => array('label' => 'Commercial', 'frequency' => 657),
+
+            'place:airport' => array('label' => 'Airport', 'frequency' => 36, 'icon' => 'transport_airport2', 'defdiameter' => 0.03),
+            'aeroway:aerodrome' => array('label' => 'Aerodrome', 'frequency' => 36, 'icon' => 'transport_airport2', 'defdiameter' => 0.03),
+            'aeroway' => array('label' => 'Aeroway', 'frequency' => 36, 'icon' => 'transport_airport2', 'defdiameter' => 0.03),
+            'railway:station' => array('label' => 'Station', 'frequency' => 3431, 'icon' => 'transport_train_station2', 'defdiameter' => 0.01),
+            'amenity:place_of_worship' => array('label' => 'Place Of Worship', 'frequency' => 9049, 'icon' => 'place_of_worship_unknown3'),
+            'amenity:pub' => array('label' => 'Pub', 'frequency' => 18969, 'icon' => 'food_pub'),
+            'amenity:bar' => array('label' => 'Bar', 'frequency' => 164, 'icon' => 'food_bar'),
+            'amenity:university' => array('label' => 'University', 'frequency' => 607, 'icon' => 'education_university'),
+            'tourism:museum' => array('label' => 'Museum', 'frequency' => 543, 'icon' => 'tourist_museum'),
+            'amenity:arts_centre' => array('label' => 'Arts Centre', 'frequency' => 136, 'icon' => 'tourist_art_gallery2'),
+            'tourism:zoo' => array('label' => 'Zoo', 'frequency' => 47, 'icon' => 'tourist_zoo'),
+            'tourism:theme_park' => array('label' => 'Theme Park', 'frequency' => 24, 'icon' => 'poi_point_of_interest'),
+            'tourism:attraction' => array('label' => 'Attraction', 'frequency' => 1463, 'icon' => 'poi_point_of_interest'),
+            'leisure:golf_course' => array('label' => 'Golf Course', 'frequency' => 712, 'icon' => 'sport_golf'),
+            'historic:castle' => array('label' => 'Castle', 'frequency' => 316, 'icon' => 'tourist_castle'),
+            'amenity:hospital' => array('label' => 'Hospital', 'frequency' => 879, 'icon' => 'health_hospital'),
+            'amenity:school' => array('label' => 'School', 'frequency' => 8192, 'icon' => 'education_school'),
+            'amenity:theatre' => array('label' => 'Theatre', 'frequency' => 371, 'icon' => 'tourist_theatre'),
+            'amenity:public_building' => array('label' => 'Public Building', 'frequency' => 985),
+            'amenity:library' => array('label' => 'Library', 'frequency' => 794, 'icon' => 'amenity_library'),
+            'amenity:townhall' => array('label' => 'Townhall', 'frequency' => 242),
+            'amenity:community_centre' => array('label' => 'Community Centre', 'frequency' => 157),
+            'amenity:fire_station' => array('label' => 'Fire Station', 'frequency' => 221, 'icon' => 'amenity_firestation3'),
+            'amenity:police' => array('label' => 'Police', 'frequency' => 334, 'icon' => 'amenity_police2'),
+            'amenity:bank' => array('label' => 'Bank', 'frequency' => 1248, 'icon' => 'money_bank2'),
+            'amenity:post_office' => array('label' => 'Post Office', 'frequency' => 859, 'icon' => 'amenity_post_office'),
+            'leisure:park' => array('label' => 'Park', 'frequency' => 2378),
+            'amenity:park' => array('label' => 'Park', 'frequency' => 53),
+            'landuse:park' => array('label' => 'Park', 'frequency' => 50),
+            'landuse:recreation_ground' => array('label' => 'Recreation Ground', 'frequency' => 517),
+            'tourism:hotel' => array('label' => 'Hotel', 'frequency' => 2150, 'icon' => 'accommodation_hotel2'),
+            'tourism:motel' => array('label' => 'Motel', 'frequency' => 43),
+            'amenity:cinema' => array('label' => 'Cinema', 'frequency' => 277, 'icon' => 'tourist_cinema'),
+            'tourism:artwork' => array('label' => 'Artwork', 'frequency' => 171, 'icon' => 'tourist_art_gallery2'),
+            'historic:archaeological_site' => array('label' => 'Archaeological Site', 'frequency' => 407, 'icon' => 'tourist_archaeological2'),
+            'amenity:doctors' => array('label' => 'Doctors', 'frequency' => 581, 'icon' => 'health_doctors'),
+            'leisure:sports_centre' => array('label' => 'Sports Centre', 'frequency' => 767, 'icon' => 'sport_leisure_centre'),
+            'leisure:swimming_pool' => array('label' => 'Swimming Pool', 'frequency' => 24, 'icon' => 'sport_swimming_outdoor'),
+            'shop:supermarket' => array('label' => 'Supermarket', 'frequency' => 2673, 'icon' => 'shopping_supermarket'),
+            'shop:convenience' => array('label' => 'Convenience', 'frequency' => 1469, 'icon' => 'shopping_convenience'),
+            'amenity:restaurant' => array('label' => 'Restaurant', 'frequency' => 3179, 'icon' => 'food_restaurant'),
+            'amenity:fast_food' => array('label' => 'Fast Food', 'frequency' => 2289, 'icon' => 'food_fastfood'),
+            'amenity:cafe' => array('label' => 'Cafe', 'frequency' => 1780, 'icon' => 'food_cafe'),
+            'tourism:guest_house' => array('label' => 'Guest House', 'frequency' => 223, 'icon' => 'accommodation_bed_and_breakfast'),
+            'amenity:pharmacy' => array('label' => 'Pharmacy', 'frequency' => 733, 'icon' => 'health_pharmacy_dispensing'),
+            'amenity:fuel' => array('label' => 'Fuel', 'frequency' => 1308, 'icon' => 'transport_fuel'),
+            'natural:peak' => array('label' => 'Peak', 'frequency' => 3212, 'icon' => 'poi_peak'),
+            'waterway:waterfall' => array('label' => 'Waterfall', 'frequency' => 24),
+            'natural:wood' => array('label' => 'Wood', 'frequency' => 1845, 'icon' => 'landuse_coniferous_and_deciduous'),
+            'natural:water' => array('label' => 'Water', 'frequency' => 1790),
+            'landuse:forest' => array('label' => 'Forest', 'frequency' => 467),
+            'landuse:cemetery' => array('label' => 'Cemetery', 'frequency' => 463),
+            'landuse:allotments' => array('label' => 'Allotments', 'frequency' => 408),
+            'landuse:farmyard' => array('label' => 'Farmyard', 'frequency' => 397),
+            'railway:rail' => array('label' => 'Rail', 'frequency' => 4894),
+            'waterway:canal' => array('label' => 'Canal', 'frequency' => 1723),
+            'waterway:river' => array('label' => 'River', 'frequency' => 4089),
+            'waterway:stream' => array('label' => 'Stream', 'frequency' => 2684),
+            'shop:bicycle' => array('label' => 'Bicycle', 'frequency' => 349, 'icon' => 'shopping_bicycle'),
+            'shop:clothes' => array('label' => 'Clothes', 'frequency' => 315, 'icon' => 'shopping_clothes'),
+            'shop:hairdresser' => array('label' => 'Hairdresser', 'frequency' => 312, 'icon' => 'shopping_hairdresser'),
+            'shop:doityourself' => array('label' => 'Doityourself', 'frequency' => 247, 'icon' => 'shopping_diy'),
+            'shop:estate_agent' => array('label' => 'Estate Agent', 'frequency' => 162, 'icon' => 'shopping_estateagent2'),
+            'shop:car' => array('label' => 'Car', 'frequency' => 159, 'icon' => 'shopping_car'),
+            'shop:garden_centre' => array('label' => 'Garden Centre', 'frequency' => 143, 'icon' => 'shopping_garden_centre'),
+            'shop:car_repair' => array('label' => 'Car Repair', 'frequency' => 141, 'icon' => 'shopping_car_repair'),
+            'shop:newsagent' => array('label' => 'Newsagent', 'frequency' => 132),
+            'shop:bakery' => array('label' => 'Bakery', 'frequency' => 129, 'icon' => 'shopping_bakery'),
+            'shop:furniture' => array('label' => 'Furniture', 'frequency' => 124),
+            'shop:butcher' => array('label' => 'Butcher', 'frequency' => 105, 'icon' => 'shopping_butcher'),
+            'shop:apparel' => array('label' => 'Apparel', 'frequency' => 98, 'icon' => 'shopping_clothes'),
+            'shop:electronics' => array('label' => 'Electronics', 'frequency' => 96),
+            'shop:department_store' => array('label' => 'Department Store', 'frequency' => 86),
+            'shop:books' => array('label' => 'Books', 'frequency' => 85),
+            'shop:yes' => array('label' => 'Shop', 'frequency' => 68),
+            'shop:outdoor' => array('label' => 'Outdoor', 'frequency' => 67),
+            'shop:mall' => array('label' => 'Mall', 'frequency' => 63),
+            'shop:florist' => array('label' => 'Florist', 'frequency' => 61),
+            'shop:charity' => array('label' => 'Charity', 'frequency' => 60),
+            'shop:hardware' => array('label' => 'Hardware', 'frequency' => 59),
+            'shop:laundry' => array('label' => 'Laundry', 'frequency' => 51, 'icon' => 'shopping_laundrette'),
+            'shop:shoes' => array('label' => 'Shoes', 'frequency' => 49),
+            'shop:beverages' => array('label' => 'Beverages', 'frequency' => 48, 'icon' => 'shopping_alcohol'),
+            'shop:dry_cleaning' => array('label' => 'Dry Cleaning', 'frequency' => 46),
+            'shop:carpet' => array('label' => 'Carpet', 'frequency' => 45),
+            'shop:computer' => array('label' => 'Computer', 'frequency' => 44),
+            'shop:alcohol' => array('label' => 'Alcohol', 'frequency' => 44, 'icon' => 'shopping_alcohol'),
+            'shop:optician' => array('label' => 'Optician', 'frequency' => 55, 'icon' => 'health_opticians'),
+            'shop:chemist' => array('label' => 'Chemist', 'frequency' => 42, 'icon' => 'health_pharmacy'),
+            'shop:gallery' => array('label' => 'Gallery', 'frequency' => 38, 'icon' => 'tourist_art_gallery2'),
+            'shop:mobile_phone' => array('label' => 'Mobile Phone', 'frequency' => 37),
+            'shop:sports' => array('label' => 'Sports', 'frequency' => 37),
+            'shop:jewelry' => array('label' => 'Jewelry', 'frequency' => 32, 'icon' => 'shopping_jewelry'),
+            'shop:pet' => array('label' => 'Pet', 'frequency' => 29),
+            'shop:beauty' => array('label' => 'Beauty', 'frequency' => 28),
+            'shop:stationery' => array('label' => 'Stationery', 'frequency' => 25),
+            'shop:shopping_centre' => array('label' => 'Shopping Centre', 'frequency' => 25),
+            'shop:general' => array('label' => 'General', 'frequency' => 25),
+            'shop:electrical' => array('label' => 'Electrical', 'frequency' => 25),
+            'shop:toys' => array('label' => 'Toys', 'frequency' => 23),
+            'shop:jeweller' => array('label' => 'Jeweller', 'frequency' => 23),
+            'shop:betting' => array('label' => 'Betting', 'frequency' => 23),
+            'shop:household' => array('label' => 'Household', 'frequency' => 21),
+            'shop:travel_agency' => array('label' => 'Travel Agency', 'frequency' => 21),
+            'shop:hifi' => array('label' => 'Hifi', 'frequency' => 21),
+            'amenity:shop' => array('label' => 'Shop', 'frequency' => 61),
+            'tourism:information' => array('label' => 'Information', 'frequency' => 224, 'icon' => 'amenity_information'),
+
+            'place:house' => array('label' => 'House', 'frequency' => 2086, 'defzoom' => 18),
+            'place:house_name' => array('label' => 'House', 'frequency' => 2086, 'defzoom' => 18),
+            'place:house_number' => array('label' => 'House Number', 'frequency' => 2086, 'defzoom' => 18),
+            'place:country_code' => array('label' => 'Country Code', 'frequency' => 2086, 'defzoom' => 18),
+
+            //
+
+            'leisure:pitch' => array('label' => 'Pitch', 'frequency' => 762),
+            'highway:unsurfaced' => array('label' => 'Unsurfaced', 'frequency' => 492),
+            'historic:ruins' => array('label' => 'Ruins', 'frequency' => 483, 'icon' => 'tourist_ruin'),
+            'amenity:college' => array('label' => 'College', 'frequency' => 473, 'icon' => 'education_school'),
+            'historic:monument' => array('label' => 'Monument', 'frequency' => 470, 'icon' => 'tourist_monument'),
+            'railway:subway' => array('label' => 'Subway', 'frequency' => 385),
+            'historic:memorial' => array('label' => 'Memorial', 'frequency' => 382, 'icon' => 'tourist_monument'),
+            'leisure:nature_reserve' => array('label' => 'Nature Reserve', 'frequency' => 342),
+            'leisure:common' => array('label' => 'Common', 'frequency' => 322),
+            'waterway:lock_gate' => array('label' => 'Lock Gate', 'frequency' => 321),
+            'natural:fell' => array('label' => 'Fell', 'frequency' => 308),
+            'amenity:nightclub' => array('label' => 'Nightclub', 'frequency' => 292),
+            'highway:path' => array('label' => 'Path', 'frequency' => 287),
+            'leisure:garden' => array('label' => 'Garden', 'frequency' => 285),
+            'landuse:reservoir' => array('label' => 'Reservoir', 'frequency' => 276),
+            'leisure:playground' => array('label' => 'Playground', 'frequency' => 264),
+            'leisure:stadium' => array('label' => 'Stadium', 'frequency' => 212),
+            'historic:mine' => array('label' => 'Mine', 'frequency' => 193, 'icon' => 'poi_mine'),
+            'natural:cliff' => array('label' => 'Cliff', 'frequency' => 193),
+            'tourism:caravan_site' => array('label' => 'Caravan Site', 'frequency' => 183, 'icon' => 'accommodation_caravan_park'),
+            'amenity:bus_station' => array('label' => 'Bus Station', 'frequency' => 181, 'icon' => 'transport_bus_station'),
+            'amenity:kindergarten' => array('label' => 'Kindergarten', 'frequency' => 179),
+            'highway:construction' => array('label' => 'Construction', 'frequency' => 176),
+            'amenity:atm' => array('label' => 'Atm', 'frequency' => 172, 'icon' => 'money_atm2'),
+            'amenity:emergency_phone' => array('label' => 'Emergency Phone', 'frequency' => 164),
+            'waterway:lock' => array('label' => 'Lock', 'frequency' => 146),
+            'waterway:riverbank' => array('label' => 'Riverbank', 'frequency' => 143),
+            'natural:coastline' => array('label' => 'Coastline', 'frequency' => 142),
+            'tourism:viewpoint' => array('label' => 'Viewpoint', 'frequency' => 140, 'icon' => 'tourist_view_point'),
+            'tourism:hostel' => array('label' => 'Hostel', 'frequency' => 140),
+            'tourism:bed_and_breakfast' => array('label' => 'Bed And Breakfast', 'frequency' => 140, 'icon' => 'accommodation_bed_and_breakfast'),
+            'railway:halt' => array('label' => 'Halt', 'frequency' => 135),
+            'railway:platform' => array('label' => 'Platform', 'frequency' => 134),
+            'railway:tram' => array('label' => 'Tram', 'frequency' => 130, 'icon' => 'transport_tram_stop'),
+            'amenity:courthouse' => array('label' => 'Courthouse', 'frequency' => 129, 'icon' => 'amenity_court'),
+            'amenity:recycling' => array('label' => 'Recycling', 'frequency' => 126, 'icon' => 'amenity_recycling'),
+            'amenity:dentist' => array('label' => 'Dentist', 'frequency' => 124, 'icon' => 'health_dentist'),
+            'natural:beach' => array('label' => 'Beach', 'frequency' => 121, 'icon' => 'tourist_beach'),
+            'place:moor' => array('label' => 'Moor', 'frequency' => 118),
+            'amenity:grave_yard' => array('label' => 'Grave Yard', 'frequency' => 110),
+            'waterway:drain' => array('label' => 'Drain', 'frequency' => 108),
+            'landuse:grass' => array('label' => 'Grass', 'frequency' => 106),
+            'landuse:village_green' => array('label' => 'Village Green', 'frequency' => 106),
+            'natural:bay' => array('label' => 'Bay', 'frequency' => 102),
+            'railway:tram_stop' => array('label' => 'Tram Stop', 'frequency' => 101, 'icon' => 'transport_tram_stop'),
+            'leisure:marina' => array('label' => 'Marina', 'frequency' => 98),
+            'highway:stile' => array('label' => 'Stile', 'frequency' => 97),
+            'natural:moor' => array('label' => 'Moor', 'frequency' => 95),
+            'railway:light_rail' => array('label' => 'Light Rail', 'frequency' => 91),
+            'railway:narrow_gauge' => array('label' => 'Narrow Gauge', 'frequency' => 90),
+            'natural:land' => array('label' => 'Land', 'frequency' => 86),
+            'amenity:village_hall' => array('label' => 'Village Hall', 'frequency' => 82),
+            'waterway:dock' => array('label' => 'Dock', 'frequency' => 80),
+            'amenity:veterinary' => array('label' => 'Veterinary', 'frequency' => 79),
+            'landuse:brownfield' => array('label' => 'Brownfield', 'frequency' => 77),
+            'leisure:track' => array('label' => 'Track', 'frequency' => 76),
+            'railway:historic_station' => array('label' => 'Historic Station', 'frequency' => 74),
+            'landuse:construction' => array('label' => 'Construction', 'frequency' => 72),
+            'amenity:prison' => array('label' => 'Prison', 'frequency' => 71, 'icon' => 'amenity_prison'),
+            'landuse:quarry' => array('label' => 'Quarry', 'frequency' => 71),
+            'amenity:telephone' => array('label' => 'Telephone', 'frequency' => 70),
+            'highway:traffic_signals' => array('label' => 'Traffic Signals', 'frequency' => 66),
+            'natural:heath' => array('label' => 'Heath', 'frequency' => 62),
+            'historic:house' => array('label' => 'House', 'frequency' => 61),
+            'amenity:social_club' => array('label' => 'Social Club', 'frequency' => 61),
+            'landuse:military' => array('label' => 'Military', 'frequency' => 61),
+            'amenity:health_centre' => array('label' => 'Health Centre', 'frequency' => 59),
+            'historic:building' => array('label' => 'Building', 'frequency' => 58),
+            'amenity:clinic' => array('label' => 'Clinic', 'frequency' => 57),
+            'highway:services' => array('label' => 'Services', 'frequency' => 56),
+            'amenity:ferry_terminal' => array('label' => 'Ferry Terminal', 'frequency' => 55),
+            'natural:marsh' => array('label' => 'Marsh', 'frequency' => 55),
+            'natural:hill' => array('label' => 'Hill', 'frequency' => 54),
+            'highway:raceway' => array('label' => 'Raceway', 'frequency' => 53),
+            'amenity:taxi' => array('label' => 'Taxi', 'frequency' => 47),
+            'amenity:take_away' => array('label' => 'Take Away', 'frequency' => 45),
+            'amenity:car_rental' => array('label' => 'Car Rental', 'frequency' => 44),
+            'place:islet' => array('label' => 'Islet', 'frequency' => 44),
+            'amenity:nursery' => array('label' => 'Nursery', 'frequency' => 44),
+            'amenity:nursing_home' => array('label' => 'Nursing Home', 'frequency' => 43),
+            'amenity:toilets' => array('label' => 'Toilets', 'frequency' => 38),
+            'amenity:hall' => array('label' => 'Hall', 'frequency' => 38),
+            'waterway:boatyard' => array('label' => 'Boatyard', 'frequency' => 36),
+            'highway:mini_roundabout' => array('label' => 'Mini Roundabout', 'frequency' => 35),
+            'historic:manor' => array('label' => 'Manor', 'frequency' => 35),
+            'tourism:chalet' => array('label' => 'Chalet', 'frequency' => 34),
+            'amenity:bicycle_parking' => array('label' => 'Bicycle Parking', 'frequency' => 34),
+            'amenity:hotel' => array('label' => 'Hotel', 'frequency' => 34),
+            'waterway:weir' => array('label' => 'Weir', 'frequency' => 33),
+            'natural:wetland' => array('label' => 'Wetland', 'frequency' => 33),
+            'natural:cave_entrance' => array('label' => 'Cave Entrance', 'frequency' => 32),
+            'amenity:crematorium' => array('label' => 'Crematorium', 'frequency' => 31),
+            'tourism:picnic_site' => array('label' => 'Picnic Site', 'frequency' => 31),
+            'landuse:wood' => array('label' => 'Wood', 'frequency' => 30),
+            'landuse:basin' => array('label' => 'Basin', 'frequency' => 30),
+            'natural:tree' => array('label' => 'Tree', 'frequency' => 30),
+            'leisure:slipway' => array('label' => 'Slipway', 'frequency' => 29),
+            'landuse:meadow' => array('label' => 'Meadow', 'frequency' => 29),
+            'landuse:piste' => array('label' => 'Piste', 'frequency' => 28),
+            'amenity:care_home' => array('label' => 'Care Home', 'frequency' => 28),
+            'amenity:club' => array('label' => 'Club', 'frequency' => 28),
+            'amenity:medical_centre' => array('label' => 'Medical Centre', 'frequency' => 27),
+            'historic:roman_road' => array('label' => 'Roman Road', 'frequency' => 27),
+            'historic:fort' => array('label' => 'Fort', 'frequency' => 26),
+            'railway:subway_entrance' => array('label' => 'Subway Entrance', 'frequency' => 26),
+            'historic:yes' => array('label' => 'Historic', 'frequency' => 25),
+            'highway:gate' => array('label' => 'Gate', 'frequency' => 25),
+            'leisure:fishing' => array('label' => 'Fishing', 'frequency' => 24),
+            'historic:museum' => array('label' => 'Museum', 'frequency' => 24),
+            'amenity:car_wash' => array('label' => 'Car Wash', 'frequency' => 24),
+            'railway:level_crossing' => array('label' => 'Level Crossing', 'frequency' => 23),
+            'leisure:bird_hide' => array('label' => 'Bird Hide', 'frequency' => 23),
+            'natural:headland' => array('label' => 'Headland', 'frequency' => 21),
+            'tourism:apartments' => array('label' => 'Apartments', 'frequency' => 21),
+            'amenity:shopping' => array('label' => 'Shopping', 'frequency' => 21),
+            'natural:scrub' => array('label' => 'Scrub', 'frequency' => 20),
+            'natural:fen' => array('label' => 'Fen', 'frequency' => 20),
+            'building:yes' => array('label' => 'Building', 'frequency' => 200),
+            'mountain_pass:yes' => array('label' => 'Mountain Pass', 'frequency' => 200),
+
+            'amenity:parking' => array('label' => 'Parking', 'frequency' => 3157),
+            'highway:bus_stop' => array('label' => 'Bus Stop', 'frequency' => 35777, 'icon' => 'transport_bus_stop2'),
+            'place:postcode' => array('label' => 'Postcode', 'frequency' => 27267),
+            'amenity:post_box' => array('label' => 'Post Box', 'frequency' => 9613),
+
+            'place:houses' => array('label' => 'Houses', 'frequency' => 85),
+            'railway:preserved' => array('label' => 'Preserved', 'frequency' => 227),
+            'waterway:derelict_canal' => array('label' => 'Derelict Canal', 'frequency' => 21),
+            'amenity:dead_pub' => array('label' => 'Dead Pub', 'frequency' => 20),
+            'railway:disused_station' => array('label' => 'Disused Station', 'frequency' => 114),
+            'railway:abandoned' => array('label' => 'Abandoned', 'frequency' => 641),
+            'railway:disused' => array('label' => 'Disused', 'frequency' => 72),
+           );
+}

lib/DB.php

@@ -0,0 +1,298 @@
+<?php
+
+namespace Nominatim;
+
+require_once(CONST_BasePath.'/lib/DatabaseError.php');
+
+/**
+ * Uses PDO to access the database specified in the CONST_Database_DSN
+ * setting.
+ */
+class DB
+{
+    protected $connection;
+
+    public function __construct($sDSN = CONST_Database_DSN)
+    {
+        $this->sDSN = $sDSN;
+    }
+
+    public function connect($bNew = false, $bPersistent = true)
+    {
+        if (isset($this->connection) && !$bNew) {
+            return true;
+        }
+        $aConnOptions = array(
+                         \PDO::ATTR_ERRMODE            => \PDO::ERRMODE_EXCEPTION,
+                         \PDO::ATTR_DEFAULT_FETCH_MODE => \PDO::FETCH_ASSOC,
+                         \PDO::ATTR_PERSISTENT         => $bPersistent
+        );
+
+        // https://secure.php.net/manual/en/ref.pdo-pgsql.connection.php
+        try {
+            $conn = new \PDO($this->sDSN, null, null, $aConnOptions);
+        } catch (\PDOException $e) {
+            $sMsg = 'Failed to establish database connection:' . $e->getMessage();
+            throw new \Nominatim\DatabaseError($sMsg, 500, null, $e->getMessage());
+        }
+
+        $conn->exec("SET DateStyle TO 'sql,european'");
+        $conn->exec("SET client_encoding TO 'utf-8'");
+        $iMaxExecution = ini_get('max_execution_time');
+        if ($iMaxExecution > 0) $conn->setAttribute(\PDO::ATTR_TIMEOUT, $iMaxExecution); // seconds
+
+        $this->connection = $conn;
+        return true;
+    }
+
+    // returns the number of rows that were modified or deleted by the SQL
+    // statement. If no rows were affected returns 0.
+    public function exec($sSQL, $aInputVars = null, $sErrMessage = 'Database query failed')
+    {
+        $val = null;
+        try {
+            if (isset($aInputVars)) {
+                $stmt = $this->connection->prepare($sSQL);
+                $stmt->execute($aInputVars);
+            } else {
+                $val = $this->connection->exec($sSQL);
+            }
+        } catch (\PDOException $e) {
+            throw new \Nominatim\DatabaseError($sErrMessage, 500, null, $e, $sSQL);
+        }
+        return $val;
+    }
+
+    /**
+     * Executes query. Returns first row as array.
+     * Returns false if no result found.
+     *
+     * @param string  $sSQL
+     *
+     * @return array[]
+     */
+    public function getRow($sSQL, $aInputVars = null, $sErrMessage = 'Database query failed')
+    {
+        try {
+            $stmt = $this->getQueryStatement($sSQL, $aInputVars, $sErrMessage);
+            $row = $stmt->fetch();
+        } catch (\PDOException $e) {
+            throw new \Nominatim\DatabaseError($sErrMessage, 500, null, $e, $sSQL);
+        }
+        return $row;
+    }
+
+    /**
+     * Executes query. Returns first value of first result.
+     * Returns false if no results found.
+     *
+     * @param string  $sSQL
+     *
+     * @return array[]
+     */
+    public function getOne($sSQL, $aInputVars = null, $sErrMessage = 'Database query failed')
+    {
+        try {
+            $stmt = $this->getQueryStatement($sSQL, $aInputVars, $sErrMessage);
+            $row = $stmt->fetch(\PDO::FETCH_NUM);
+            if ($row === false) return false;
+        } catch (\PDOException $e) {
+            throw new \Nominatim\DatabaseError($sErrMessage, 500, null, $e, $sSQL);
+        }
+        return $row[0];
+    }
+
+    /**
+     * Executes query. Returns array of results (arrays).
+     * Returns empty array if no results found.
+     *
+     * @param string  $sSQL
+     *
+     * @return array[]
+     */
+    public function getAll($sSQL, $aInputVars = null, $sErrMessage = 'Database query failed')
+    {
+        try {
+            $stmt = $this->getQueryStatement($sSQL, $aInputVars, $sErrMessage);
+            $rows = $stmt->fetchAll();
+        } catch (\PDOException $e) {
+            throw new \Nominatim\DatabaseError($sErrMessage, 500, null, $e, $sSQL);
+        }
+        return $rows;
+    }
+
+    /**
+     * Executes query. Returns array of the first value of each result.
+     * Returns empty array if no results found.
+     *
+     * @param string  $sSQL
+     *
+     * @return array[]
+     */
+    public function getCol($sSQL, $aInputVars = null, $sErrMessage = 'Database query failed')
+    {
+        $aVals = array();
+        try {
+            $stmt = $this->getQueryStatement($sSQL, $aInputVars, $sErrMessage);
+
+            while ($val = $stmt->fetchColumn(0)) { // returns first column or false
+                $aVals[] = $val;
+            }
+        } catch (\PDOException $e) {
+            throw new \Nominatim\DatabaseError($sErrMessage, 500, null, $e, $sSQL);
+        }
+        return $aVals;
+    }
+
+    /**
+     * Executes query. Returns associate array mapping first value to second value of each result.
+     * Returns empty array if no results found.
+     *
+     * @param string  $sSQL
+     *
+     * @return array[]
+     */
+    public function getAssoc($sSQL, $aInputVars = null, $sErrMessage = 'Database query failed')
+    {
+        try {
+            $stmt = $this->getQueryStatement($sSQL, $aInputVars, $sErrMessage);
+
+            $aList = array();
+            while ($aRow = $stmt->fetch(\PDO::FETCH_NUM)) {
+                $aList[$aRow[0]] = $aRow[1];
+            }
+        } catch (\PDOException $e) {
+            throw new \Nominatim\DatabaseError($sErrMessage, 500, null, $e, $sSQL);
+        }
+        return $aList;
+    }
+
+    /**
+     * Executes query. Returns a PDO statement to iterate over.
+     *
+     * @param string  $sSQL
+     *
+     * @return PDOStatement
+     */
+    public function getQueryStatement($sSQL, $aInputVars = null, $sErrMessage = 'Database query failed')
+    {
+        try {
+            if (isset($aInputVars)) {
+                $stmt = $this->connection->prepare($sSQL);
+                $stmt->execute($aInputVars);
+            } else {
+                $stmt = $this->connection->query($sSQL);
+            }
+        } catch (\PDOException $e) {
+            throw new \Nominatim\DatabaseError($sErrMessage, 500, null, $e, $sSQL);
+        }
+        return $stmt;
+    }
+
+    /**
+     * St. John's Way => 'St. John\'s Way'
+     *
+     * @param string  $sVal  Text to be quoted.
+     *
+     * @return string
+     */
+    public function getDBQuoted($sVal)
+    {
+        return $this->connection->quote($sVal);
+    }
+
+    /**
+     * Like getDBQuoted, but takes an array.
+     *
+     * @param array  $aVals  List of text to be quoted.
+     *
+     * @return array[]
+     */
+    public function getDBQuotedList($aVals)
+    {
+        return array_map(function ($sVal) {
+            return $this->getDBQuoted($sVal);
+        }, $aVals);
+    }
+
+    /**
+     * [1,2,'b'] => 'ARRAY[1,2,'b']''
+     *
+     * @param array  $aVals  List of text to be quoted.
+     *
+     * @return string
+     */
+    public function getArraySQL($a)
+    {
+        return 'ARRAY['.join(',', $a).']';
+    }
+
+    /**
+     * Check if a table exists in the database. Returns true if it does.
+     *
+     * @param string  $sTableName
+     *
+     * @return boolean
+     */
+    public function tableExists($sTableName)
+    {
+        $sSQL = 'SELECT count(*) FROM pg_tables WHERE tablename = :tablename';
+        return ($this->getOne($sSQL, array(':tablename' => $sTableName)) == 1);
+    }
+
+    /**
+     * Since the DSN includes the database name, checks if the connection works.
+     *
+     * @return boolean
+     */
+    public function databaseExists()
+    {
+        $bExists = true;
+        try {
+            $this->connect(true);
+        } catch (\Nominatim\DatabaseError $e) {
+            $bExists = false;
+        }
+        return $bExists;
+    }
+
+    /**
+     * e.g. 9.6, 10, 11.2
+     *
+     * @return float
+     */
+    public function getPostgresVersion()
+    {
+        $sVersionString = $this->getOne('SHOW server_version_num');
+        preg_match('#([0-9]?[0-9])([0-9][0-9])[0-9][0-9]#', $sVersionString, $aMatches);
+        return (float) ($aMatches[1].'.'.$aMatches[2]);
+    }
+
+    /**
+     * e.g. 2, 2.2
+     *
+     * @return float
+     */
+    public function getPostgisVersion()
+    {
+        $sVersionString = $this->getOne('select postgis_lib_version()');
+        preg_match('#^([0-9]+)[.]([0-9]+)[.]#', $sVersionString, $aMatches);
+        return (float) ($aMatches[1].'.'.$aMatches[2]);
+    }
+
+    public static function parseDSN($sDSN)
+    {
+        // https://secure.php.net/manual/en/ref.pdo-pgsql.connection.php
+        $aInfo = array();
+        if (preg_match('/^pgsql:(.+)/', $sDSN, $aMatches)) {
+            foreach (explode(';', $aMatches[1]) as $sKeyVal) {
+                list($sKey, $sVal) = explode('=', $sKeyVal, 2);
+                if ($sKey == 'host') $sKey = 'hostspec';
+                if ($sKey == 'dbname') $sKey = 'database';
+                if ($sKey == 'user') $sKey = 'username';
+                $aInfo[$sKey] = $sVal;
+            }
+        }
+        return $aInfo;
+    }
+}

lib/DatabaseError.php

@@ -0,0 +1,34 @@
+<?php
+
+namespace Nominatim;
+
+class DatabaseError extends \Exception
+{
+
+    public function __construct($message, $code = 500, Exception $previous = null, $oPDOErr, $sSql = null)
+    {
+        parent::__construct($message, $code, $previous);
+        // https://secure.php.net/manual/en/class.pdoexception.php
+        $this->oPDOErr = $oPDOErr;
+        $this->sSql = $sSql;
+    }
+
+    public function __toString()
+    {
+        return __CLASS__ . ": [{$this->code}]: {$this->message}\n";
+    }
+
+    public function getSqlError()
+    {
+        return $this->oPDOErr->getMessage();
+    }
+
+    public function getSqlDebugDump()
+    {
+        if (CONST_Debug) {
+            return var_export($this->oPDOErr, true);
+        } else {
+            return $this->sSql;
+        }
+    }
+}

lib/DebugHtml.php

@@ -0,0 +1,180 @@
+<?php
+
+namespace Nominatim;
+
+class Debug
+{
+    public static function newFunction($sHeading)
+    {
+        echo "<pre><h2>Debug output for $sHeading</h2></pre>\n";
+    }
+
+    public static function newSection($sHeading)
+    {
+        echo "<hr><pre><h3>$sHeading</h3></pre>\n";
+    }
+
+    public static function printVar($sHeading, $mVar)
+    {
+        echo '<pre><b>'.$sHeading. ':</b>  ';
+        Debug::outputVar($mVar, str_repeat(' ', strlen($sHeading) + 3));
+        echo "</pre>\n";
+    }
+
+    public static function fmtArrayVals($aArr)
+    {
+        return array('__debug_format' => 'array_vals', 'data' => $aArr);
+    }
+
+    public static function printDebugArray($sHeading, $oVar)
+    {
+
+        if ($oVar === null) {
+            Debug::printVar($sHeading, 'null');
+        } else {
+            Debug::printVar($sHeading, $oVar->debugInfo());
+        }
+    }
+
+    public static function printDebugTable($sHeading, $aVar)
+    {
+        echo '<b>'.$sHeading.":</b>\n";
+        echo "<table border='1'>\n";
+        if (!empty($aVar)) {
+            echo "  <tr>\n";
+            $aKeys = array();
+            $aInfo = reset($aVar);
+            if (!is_array($aInfo)) {
+                $aInfo = $aInfo->debugInfo();
+            }
+            foreach ($aInfo as $sKey => $mVal) {
+                echo '    <th><small>'.$sKey.'</small></th>'."\n";
+                $aKeys[] = $sKey;
+            }
+            echo "  </tr>\n";
+            foreach ($aVar as $oRow) {
+                $aInfo = $oRow;
+                if (!is_array($oRow)) {
+                    $aInfo = $oRow->debugInfo();
+                }
+                echo "  <tr>\n";
+                foreach ($aKeys as $sKey) {
+                    echo '    <td><pre>';
+                    if (isset($aInfo[$sKey])) {
+                        Debug::outputVar($aInfo[$sKey], '');
+                    }
+                    echo '</pre></td>'."\n";
+                }
+                echo "  </tr>\n";
+            }
+        }
+        echo "</table>\n";
+    }
+
+    public static function printGroupedSearch($aSearches, $aWordsIDs)
+    {
+        echo '<table border="1">';
+        echo '<tr><th>rank</th><th>Name Tokens</th><th>Name Not</th>';
+        echo '<th>Address Tokens</th><th>Address Not</th>';
+        echo '<th>country</th><th>operator</th>';
+        echo '<th>class</th><th>type</th><th>postcode</th><th>housenumber</th></tr>';
+        foreach ($aSearches as $iRank => $aRankedSet) {
+            foreach ($aRankedSet as $aRow) {
+                $aRow->dumpAsHtmlTableRow($aWordsIDs);
+            }
+        }
+        echo '</table>';
+    }
+
+    public static function printGroupTable($sHeading, $aVar)
+    {
+        echo '<b>'.$sHeading.":</b>\n";
+        echo "<table border='1'>\n";
+        if (!empty($aVar)) {
+            echo "  <tr>\n";
+            echo '    <th><small>Group</small></th>'."\n";
+            $aKeys = array();
+            $aInfo = reset($aVar)[0];
+            if (!is_array($aInfo)) {
+                $aInfo = $aInfo->debugInfo();
+            }
+            foreach ($aInfo as $sKey => $mVal) {
+                echo '    <th><small>'.$sKey.'</small></th>'."\n";
+                $aKeys[] = $sKey;
+            }
+            echo "  </tr>\n";
+            foreach ($aVar as $sGrpKey => $aGroup) {
+                foreach ($aGroup as $oRow) {
+                    $aInfo = $oRow;
+                    if (!is_array($oRow)) {
+                        $aInfo = $oRow->debugInfo();
+                    }
+                    echo "  <tr>\n";
+                    echo '    <td><pre>'.$sGrpKey.'</pre></td>'."\n";
+                    foreach ($aKeys as $sKey) {
+                        echo '    <td><pre>';
+                        if (!empty($aInfo[$sKey])) {
+                            Debug::outputVar($aInfo[$sKey], '');
+                        }
+                        echo '</pre></td>'."\n";
+                    }
+                    echo "  </tr>\n";
+                }
+            }
+        }
+        echo "</table>\n";
+    }
+
+    public static function printSQL($sSQL)
+    {
+        echo '<p><tt><font color="#aaa">'.$sSQL.'</font></tt></p>'."\n";
+    }
+
+    private static function outputVar($mVar, $sPreNL)
+    {
+        if (is_array($mVar) && !isset($mVar['__debug_format'])) {
+            $sPre = '';
+            foreach ($mVar as $mKey => $aValue) {
+                echo $sPre;
+                $iKeyLen = Debug::outputSimpleVar($mKey);
+                echo ' => ';
+                Debug::outputVar(
+                    $aValue,
+                    $sPreNL.str_repeat(' ', $iKeyLen + 4)
+                );
+                $sPre = "\n".$sPreNL;
+            }
+        } elseif (is_array($mVar) && isset($mVar['__debug_format'])) {
+            if (!empty($mVar['data'])) {
+                $sPre = '';
+                foreach ($mVar['data'] as $mValue) {
+                    echo $sPre;
+                    Debug::outputSimpleVar($mValue);
+                    $sPre = ', ';
+                }
+            }
+        } elseif (is_object($mVar) && method_exists($mVar, 'debugInfo')) {
+            Debug::outputVar($mVar->debugInfo(), $sPreNL);
+        } elseif (is_a($mVar, 'stdClass')) {
+            Debug::outputVar(json_decode(json_encode($mVar), true), $sPreNL);
+        } else {
+            Debug::outputSimpleVar($mVar);
+        }
+    }
+
+    private static function outputSimpleVar($mVar)
+    {
+        if (is_bool($mVar)) {
+            echo '<i>'.($mVar ? 'True' : 'False').'</i>';
+            return $mVar ? 4 : 5;
+        }
+
+        if (is_string($mVar)) {
+            echo "'$mVar'";
+            return strlen($mVar) + 2;
+        }
+
+        echo (string)$mVar;
+        return strlen((string)$mVar);
+    }
+}

lib/DebugNone.php

@@ -0,0 +1,11 @@
+<?php
+
+namespace Nominatim;
+
+class Debug
+{
+    public static function __callStatic($name, $arguments)
+    {
+        // nothing
+    }
+}

lib/Geocode.php

@@ -7,6 +7,7 @@ require_once(CONST_BasePath.'/lib/Phrase.php');
 require_once(CONST_BasePath.'/lib/ReverseGeocode.php');
 require_once(CONST_BasePath.'/lib/SearchDescription.php');
 require_once(CONST_BasePath.'/lib/SearchContext.php');
+require_once(CONST_BasePath.'/lib/TokenList.php');
 
 class Geocode
 {
@@ -16,8 +17,6 @@ class Geocode
 
     protected $aLangPrefOrder = array();
 
-    protected $bIncludeAddressDetails = false;
-
     protected $aExcludePlaceIDs = array();
     protected $bReverseInPlan = false;
 
@@ -86,7 +85,6 @@ class Geocode
             $aParams['exclude_place_ids'] = implode(',', $this->aExcludePlaceIDs);
         }
 
-        if ($this->bIncludeAddressDetails) $aParams['addressdetails'] = '1';
         if ($this->bBoundedSearch) $aParams['bounded'] = '1';
 
         if ($this->aCountryCodes) {
@@ -151,6 +149,10 @@ class Geocode
 
     private function viewboxImportanceFactor($fX, $fY)
     {
+        if (!$this->aViewBox) {
+            return 1;
+        }
+
         $fWidth = ($this->aViewBox[2] - $this->aViewBox[0])/2;
         $fHeight = ($this->aViewBox[3] - $this->aViewBox[1])/2;
 
@@ -182,9 +184,6 @@ class Geocode
 
     public function loadParamArray($oParams, $sForceGeometryType = null)
     {
-        $this->bIncludeAddressDetails
-         = $oParams->getBool('addressdetails', $this->bIncludeAddressDetails);
-
         $this->bBoundedSearch = $oParams->getBool('bounded', $this->bBoundedSearch);
 
         $this->setLimit($oParams->getInt('limit', $this->iFinalLimit));
@@ -225,14 +224,14 @@ class Geocode
         $aViewbox = $oParams->getStringList('viewboxlbrt');
         if ($aViewbox) {
             if (count($aViewbox) != 4) {
-                userError("Bad parmater 'viewboxlbrt'. Expected 4 coordinates.");
+                userError("Bad parameter 'viewboxlbrt'. Expected 4 coordinates.");
             }
             $this->setViewbox($aViewbox);
         } else {
             $aViewbox = $oParams->getStringList('viewbox');
             if ($aViewbox) {
                 if (count($aViewbox) != 4) {
-                    userError("Bad parmater 'viewbox'. Expected 4 coordinates.");
+                    userError("Bad parameter 'viewbox'. Expected 4 coordinates.");
                 }
                 $this->setViewBox($aViewbox);
             } else {
@@ -246,8 +245,8 @@ class Geocode
         }
 
         $this->oPlaceLookup->loadParamArray($oParams, $sForceGeometryType);
-        $this->oPlaceLookup->setIncludeAddressDetails(false);
         $this->oPlaceLookup->setIncludePolygonAsPoints($oParams->getBool('polygon'));
+        $this->oPlaceLookup->setIncludeAddressDetails($oParams->getBool('addressdetails', false));
     }
 
     public function setQueryFromParams($oParams)
@@ -303,7 +302,7 @@ class Geocode
         $this->loadStructuredAddressElement($sPostalCode, 'postalcode', 5, 11, array(5, 11));
         $this->loadStructuredAddressElement($sCountry, 'country', 4, 4, false);
 
-        if (sizeof($this->aStructuredQuery) > 0) {
+        if (!empty($this->aStructuredQuery)) {
             $this->sQuery = join(', ', $this->aStructuredQuery);
             if ($this->iMaxAddressRank < 30) {
                 $this->sAllowedTypesSQLList = '(\'place\',\'boundary\')';
@@ -317,7 +316,7 @@ class Geocode
 
         $aParams = $this->aStructuredQuery;
 
-        if (sizeof($aParams) == 1) return false;
+        if (count($aParams) == 1) return false;
 
         $aOrderToFallback = array('postalcode', 'street', 'city', 'county', 'state');
 
@@ -332,10 +331,10 @@ class Geocode
         return false;
     }
 
-    public function getGroupedSearches($aSearches, $aPhrases, $aValidTokens, $bIsStructured)
+    public function getGroupedSearches($aSearches, $aPhrases, $oValidTokens, $bIsStructured)
     {
         /*
-             Calculate all searches using aValidTokens i.e.
+             Calculate all searches using oValidTokens i.e.
              'Wodsworth Road, Sheffield' =>
 
              Phrase Wordset
@@ -365,38 +364,37 @@ class Geocode
                         //var_dump($oCurrentSearch);
                         //echo "</i>";
 
-                        // If the token is valid
-                        if (isset($aValidTokens[' '.$sToken])) {
-                            foreach ($aValidTokens[' '.$sToken] as $aSearchTerm) {
-                                $aNewSearches = $oCurrentSearch->extendWithFullTerm(
-                                    $aSearchTerm,
-                                    isset($aValidTokens[$sToken])
-                                      && strpos($sToken, ' ') === false,
-                                    $sPhraseType,
-                                    $iToken == 0 && $iPhrase == 0,
-                                    $iPhrase == 0,
-                                    $iToken + 1 == sizeof($aWordset)
-                                      && $iPhrase + 1 == sizeof($aPhrases)
-                                );
+                        // Tokens with full name matches.
+                        foreach ($oValidTokens->get(' '.$sToken) as $oSearchTerm) {
+                            $aNewSearches = $oCurrentSearch->extendWithFullTerm(
+                                $oSearchTerm,
+                                $oValidTokens->contains($sToken)
+                                  && strpos($sToken, ' ') === false,
+                                $sPhraseType,
+                                $iToken == 0 && $iPhrase == 0,
+                                $iPhrase == 0,
+                                $iToken + 1 == count($aWordset)
+                                  && $iPhrase + 1 == count($aPhrases)
+                            );
 
-                                foreach ($aNewSearches as $oSearch) {
-                                    if ($oSearch->getRank() < $this->iMaxRank) {
-                                        $aNewWordsetSearches[] = $oSearch;
-                                    }
+                            foreach ($aNewSearches as $oSearch) {
+                                if ($oSearch->getRank() < $this->iMaxRank) {
+                                    $aNewWordsetSearches[] = $oSearch;
                                 }
                             }
                         }
                         // Look for partial matches.
                         // Note that there is no point in adding country terms here
                         // because country is omitted in the address.
-                        if (isset($aValidTokens[$sToken]) && $sPhraseType != 'country') {
+                        if ($sPhraseType != 'country') {
                             // Allow searching for a word - but at extra cost
-                            foreach ($aValidTokens[$sToken] as $aSearchTerm) {
+                            foreach ($oValidTokens->get($sToken) as $oSearchTerm) {
                                 $aNewSearches = $oCurrentSearch->extendWithPartialTerm(
-                                    $aSearchTerm,
+                                    $sToken,
+                                    $oSearchTerm,
                                     $bIsStructured,
                                     $iPhrase,
-                                    isset($aValidTokens[' '.$sToken]) ? $aValidTokens[' '.$sToken] : array()
+                                    $oValidTokens->get(' '.$sToken)
                                 );
 
                                 foreach ($aNewSearches as $oSearch) {
@@ -411,7 +409,7 @@ class Geocode
                     usort($aNewWordsetSearches, array('Nominatim\SearchDescription', 'bySearchRank'));
                     $aWordsetSearches = array_slice($aNewWordsetSearches, 0, 50);
                 }
-                //var_Dump('<hr>',sizeof($aWordsetSearches)); exit;
+                //var_Dump('<hr>',count($aWordsetSearches)); exit;
 
                 $aNewPhraseSearches = array_merge($aNewPhraseSearches, $aNewWordsetSearches);
                 usort($aNewPhraseSearches, array('Nominatim\SearchDescription', 'bySearchRank'));
@@ -442,12 +440,10 @@ class Geocode
             $iSearchCount = 0;
             $aSearches = array();
             foreach ($aGroupedSearches as $iScore => $aNewSearches) {
-                $iSearchCount += sizeof($aNewSearches);
+                $iSearchCount += count($aNewSearches);
                 $aSearches = array_merge($aSearches, $aNewSearches);
                 if ($iSearchCount > 50) break;
             }
-
-            //if (CONST_Debug) _debugDumpGroupedSearches($aGroupedSearches, $aValidTokens);
         }
 
         // Revisit searches, drop bad searches and give penalty to unlikely combinations.
@@ -502,8 +498,11 @@ class Geocode
 
     public function lookup()
     {
+        Debug::newFunction('Geocode::lookup');
         if (!$this->sQuery && !$this->aStructuredQuery) return array();
 
+        Debug::printDebugArray('Geocode', $this);
+
         $oCtx = new SearchContext();
 
         if ($this->aRoutePoints) {
@@ -523,9 +522,13 @@ class Geocode
             $oCtx->setCountryList($this->aCountryCodes);
         }
 
+        Debug::newSection('Query Preprocessing');
+
         $sNormQuery = $this->normTerm($this->sQuery);
-        $sLanguagePrefArraySQL = getArraySQL(
-            array_map('getDBQuoted', $this->aLangPrefOrder)
+        Debug::printVar('Normalized query', $sNormQuery);
+
+        $sLanguagePrefArraySQL = $this->oDB->getArraySQL(
+            $this->oDB->getDBQuotedList($this->aLangPrefOrder)
         );
 
         $sQuery = $this->sQuery;
@@ -535,15 +538,14 @@ class Geocode
 
         // Conflicts between US state abreviations and various words for 'the' in different languages
         if (isset($this->aLangPrefOrder['name:en'])) {
-            $sQuery = preg_replace('/(^|,)\s*il\s*(,|$)/', '\1illinois\2', $sQuery);
-            $sQuery = preg_replace('/(^|,)\s*al\s*(,|$)/', '\1alabama\2', $sQuery);
-            $sQuery = preg_replace('/(^|,)\s*la\s*(,|$)/', '\1louisiana\2', $sQuery);
+            $sQuery = preg_replace('/(^|,)\s*il\s*(,|$)/i', '\1illinois\2', $sQuery);
+            $sQuery = preg_replace('/(^|,)\s*al\s*(,|$)/i', '\1alabama\2', $sQuery);
+            $sQuery = preg_replace('/(^|,)\s*la\s*(,|$)/i', '\1louisiana\2', $sQuery);
         }
 
         // Do we have anything that looks like a lat/lon pair?
         $sQuery = $oCtx->setNearPointFromQuery($sQuery);
 
-        $aResults = array();
         if ($sQuery || $this->aStructuredQuery) {
             // Start with a single blank search
             $aSearches = array(new SearchDescription($oCtx));
@@ -560,6 +562,10 @@ class Geocode
                     $aSpecialTermsRaw,
                     PREG_SET_ORDER
                 );
+                if (!empty($aSpecialTermsRaw)) {
+                    Debug::printVar('Special terms', $aSpecialTermsRaw);
+                }
+
                 foreach ($aSpecialTermsRaw as $aSpecialTerm) {
                     $sQuery = str_replace($aSpecialTerm[0], ' ', $sQuery);
                     if (!$sSpecialTerm) {
@@ -575,15 +581,17 @@ class Geocode
 
             if ($sSpecialTerm && !$aSearches[0]->hasOperator()) {
                 $sSpecialTerm = pg_escape_string($sSpecialTerm);
-                $sToken = chksql(
-                    $this->oDB->getOne("SELECT make_standard_name('$sSpecialTerm')"),
+                $sToken = $this->oDB->getOne(
+                    'SELECT make_standard_name(:term)',
+                    array(':term' => $sSpecialTerm),
                     'Cannot decode query. Wrong encoding?'
                 );
                 $sSQL = 'SELECT class, type FROM word ';
                 $sSQL .= '   WHERE word_token in (\' '.$sToken.'\')';
                 $sSQL .= '   AND class is not null AND class not in (\'place\')';
-                if (CONST_Debug) var_Dump($sSQL);
-                $aSearchWords = chksql($this->oDB->getAll($sSQL));
+
+                Debug::printSQL($sSQL);
+                $aSearchWords = $this->oDB->getAll($sSQL);
                 $aNewSearches = array();
                 foreach ($aSearches as $oSearch) {
                     foreach ($aSearchWords as $aSearchTerm) {
@@ -609,15 +617,21 @@ class Geocode
                 $bStructuredPhrases = false;
             }
 
+            Debug::printDebugArray('Search context', $oCtx);
+            Debug::printDebugArray('Base search', empty($aSearches) ? null : $aSearches[0]);
+            Debug::printVar('Final query phrases', $aInPhrases);
+
             // Convert each phrase to standard form
             // Create a list of standard words
             // Get all 'sets' of words
             // Generate a complete list of all
+            Debug::newSection('Tokenization');
             $aTokens = array();
             $aPhrases = array();
             foreach ($aInPhrases as $iPhrase => $sPhrase) {
-                $sPhrase = chksql(
-                    $this->oDB->getOne('SELECT make_standard_name('.getDBQuoted($sPhrase).')'),
+                $sPhrase = $this->oDB->getOne(
+                    'SELECT make_standard_name(:phrase)',
+                    array(':phrase' => $sPhrase),
                     'Cannot normalize query string (is it a UTF-8 string?)'
                 );
                 if (trim($sPhrase)) {
@@ -627,74 +641,54 @@ class Geocode
                 }
             }
 
-            if (sizeof($aTokens)) {
-                // Check which tokens we have, get the ID numbers
+            Debug::printDebugTable('Phrases', $aPhrases);
+            Debug::printVar('Tokens', $aTokens);
+
+            $oValidTokens = new TokenList();
+
+            if (!empty($aTokens)) {
                 $sSQL = 'SELECT word_id, word_token, word, class, type, country_code, operator, search_name_count';
                 $sSQL .= ' FROM word ';
-                $sSQL .= ' WHERE word_token in ('.join(',', array_map('getDBQuoted', $aTokens)).')';
+                $sSQL .= ' WHERE word_token in ('.join(',', $this->oDB->getDBQuotedList($aTokens)).')';
 
-                if (CONST_Debug) var_Dump($sSQL);
+                Debug::printSQL($sSQL);
 
-                $aValidTokens = array();
-                $aDatabaseWords = chksql(
-                    $this->oDB->getAll($sSQL),
-                    'Could not get word tokens.'
+                $oValidTokens->addTokensFromDB(
+                    $this->oDB,
+                    $aTokens,
+                    $this->aCountryCodes,
+                    $sNormQuery,
+                    $this->oNormalizer
                 );
-                $aWordFrequencyScores = array();
-                foreach ($aDatabaseWords as $aToken) {
-                    // Filter country tokens that do not match restricted countries.
-                    if ($this->aCountryCodes
-                        && $aToken['country_code']
-                        && !in_array($aToken['country_code'], $this->aCountryCodes)
-                    ) {
-                        continue;
-                    }
 
-                    // Special terms need to appear in their normalized form.
-                    if ($aToken['word'] && $aToken['class']) {
-                        $sNormWord = $this->normTerm($aToken['word']);
-                        if (strpos($sNormQuery, $sNormWord) === false) {
-                            continue;
-                        }
-                    }
-
-                    if (isset($aValidTokens[$aToken['word_token']])) {
-                        $aValidTokens[$aToken['word_token']][] = $aToken;
-                    } else {
-                        $aValidTokens[$aToken['word_token']] = array($aToken);
-                    }
-                    $aWordFrequencyScores[$aToken['word_id']] = $aToken['search_name_count'] + 1;
-                }
-                if (CONST_Debug) var_Dump($aPhrases, $aValidTokens);
-
-                // US ZIP+4 codes - if there is no token, merge in the 5-digit ZIP code
+                // Try more interpretations for Tokens that could not be matched.
                 foreach ($aTokens as $sToken) {
-                    if (!isset($aValidTokens[$sToken]) && preg_match('/^([0-9]{5}) [0-9]{4}$/', $sToken, $aData)) {
-                        if (isset($aValidTokens[$aData[1]])) {
-                            foreach ($aValidTokens[$aData[1]] as $aToken) {
-                                if (!$aToken['class']) {
-                                    if (isset($aValidTokens[$sToken])) {
-                                        $aValidTokens[$sToken][] = $aToken;
-                                    } else {
-                                        $aValidTokens[$sToken] = array($aToken);
-                                    }
-                                }
-                            }
+                    if ($sToken[0] == ' ' && !$oValidTokens->contains($sToken)) {
+                        if (preg_match('/^ ([0-9]{5}) [0-9]{4}$/', $sToken, $aData)) {
+                            // US ZIP+4 codes - merge in the 5-digit ZIP code
+                            $oValidTokens->addToken(
+                                $sToken,
+                                new Token\Postcode(null, $aData[1], 'us')
+                            );
+                        } elseif (preg_match('/^ [0-9]+$/', $sToken)) {
+                            // Unknown single word token with a number.
+                            // Assume it is a house number.
+                            $oValidTokens->addToken(
+                                $sToken,
+                                new Token\HouseNumber(null, trim($sToken))
+                            );
                         }
                     }
                 }
 
-                foreach ($aTokens as $sToken) {
-                    // Unknown single word token with a number - assume it is a house number
-                    if (!isset($aValidTokens[' '.$sToken]) && strpos($sToken, ' ') === false && preg_match('/^[0-9]+$/', $sToken)) {
-                        $aValidTokens[' '.$sToken] = array(array('class' => 'place', 'type' => 'house', 'word_token' => ' '.$sToken));
-                    }
-                }
-
                 // Any words that have failed completely?
                 // TODO: suggestions
 
-                $aGroupedSearches = $this->getGroupedSearches($aSearches, $aPhrases, $aValidTokens, $bStructuredPhrases);
+                Debug::printGroupTable('Valid Tokens', $oValidTokens->debugInfo());
+
+                Debug::newSection('Search candidates');
+
+                $aGroupedSearches = $this->getGroupedSearches($aSearches, $aPhrases, $oValidTokens, $bStructuredPhrases);
 
                 if ($this->bReverseInPlan) {
                     // Reverse phrase array and also reverse the order of the wordsets in
@@ -702,10 +696,10 @@ class Geocode
                     // because order in the address doesn't matter.
                     $aPhrases = array_reverse($aPhrases);
                     $aPhrases[0]->invertWordSets();
-                    if (sizeof($aPhrases) > 1) {
-                        $aPhrases[sizeof($aPhrases)-1]->invertWordSets();
+                    if (count($aPhrases) > 1) {
+                        $aPhrases[count($aPhrases)-1]->invertWordSets();
                     }
-                    $aReverseGroupedSearches = $this->getGroupedSearches($aSearches, $aPhrases, $aValidTokens, false);
+                    $aReverseGroupedSearches = $this->getGroupedSearches($aSearches, $aPhrases, $oValidTokens, false);
 
                     foreach ($aGroupedSearches as $aSearches) {
                         foreach ($aSearches as $aSearch) {
@@ -738,40 +732,71 @@ class Geocode
                     $sHash = serialize($aSearch);
                     if (isset($aSearchHash[$sHash])) {
                         unset($aGroupedSearches[$iGroup][$iSearch]);
-                        if (sizeof($aGroupedSearches[$iGroup]) == 0) unset($aGroupedSearches[$iGroup]);
+                        if (empty($aGroupedSearches[$iGroup])) unset($aGroupedSearches[$iGroup]);
                     } else {
                         $aSearchHash[$sHash] = 1;
                     }
                 }
             }
 
-            if (CONST_Debug) _debugDumpGroupedSearches($aGroupedSearches, $aValidTokens);
+            Debug::printGroupedSearch(
+                $aGroupedSearches,
+                $oValidTokens->debugTokenByWordIdList()
+            );
 
             // Start the search process
             $iGroupLoop = 0;
             $iQueryLoop = 0;
+            $aNextResults = array();
             foreach ($aGroupedSearches as $iGroupedRank => $aSearches) {
                 $iGroupLoop++;
+                $aResults = $aNextResults;
                 foreach ($aSearches as $oSearch) {
                     $iQueryLoop++;
 
-                    if (CONST_Debug) {
-                        echo "<hr><b>Search Loop, group $iGroupLoop, loop $iQueryLoop</b>";
-                        _debugDumpGroupedSearches(array($iGroupedRank => array($oSearch)), $aValidTokens);
-                    }
+                    Debug::newSection("Search Loop, group $iGroupLoop, loop $iQueryLoop");
+                    Debug::printGroupedSearch(
+                        array($iGroupedRank => array($oSearch)),
+                        $oValidTokens->debugTokenByWordIdList()
+                    );
 
-                    $aResults += $oSearch->query(
+                    $aNewResults = $oSearch->query(
                         $this->oDB,
-                        $aWordFrequencyScores,
                         $this->iMinAddressRank,
                         $this->iMaxAddressRank,
                         $this->iLimit
                     );
 
+                    // The same result may appear in different rounds, only
+                    // use the one with minimal rank.
+                    foreach ($aNewResults as $iPlace => $oRes) {
+                        if (!isset($aResults[$iPlace])
+                            || $aResults[$iPlace]->iResultRank > $oRes->iResultRank) {
+                            $aResults[$iPlace] = $oRes;
+                        }
+                    }
+
                     if ($iQueryLoop > 20) break;
                 }
 
-                if (sizeof($aResults) && ($this->iMinAddressRank != 0 || $this->iMaxAddressRank != 30)) {
+                if (!empty($aResults)) {
+                    $aSplitResults = Result::splitResults($aResults);
+                    Debug::printVar('Split results', $aSplitResults);
+                    if ($iGroupLoop <= 4 && empty($aSplitResults['tail'])
+                        && reset($aSplitResults['head'])->iResultRank > 0) {
+                        // Haven't found an exact match for the query yet.
+                        // Therefore add result from the next group level.
+                        $aNextResults = $aSplitResults['head'];
+                        foreach ($aNextResults as $oRes) {
+                            $oRes->iResultRank--;
+                        }
+                        $aResults = array();
+                    } else {
+                        $aResults = $aSplitResults['head'];
+                    }
+                }
+
+                if (!empty($aResults) && ($this->iMinAddressRank != 0 || $this->iMaxAddressRank != 30)) {
                     // Need to verify passes rank limits before dropping out of the loop (yuk!)
                     // reduces the number of place ids, like a filter
                     // rank_address is 30 for interpolated housenumbers
@@ -806,8 +831,8 @@ class Geocode
                     $aFilteredIDs = array();
                     if ($aFilterSql) {
                         $sSQL = join(' UNION ', $aFilterSql);
-                        if (CONST_Debug) var_dump($sSQL);
-                        $aFilteredIDs = chksql($this->oDB->getCol($sSQL));
+                        Debug::printSQL($sSQL);
+                        $aFilteredIDs = $this->oDB->getCol($sSQL);
                     }
 
                     $tempIDs = array();
@@ -824,7 +849,7 @@ class Geocode
                     $aResults = $tempIDs;
                 }
 
-                if (sizeof($aResults)) break;
+                if (!empty($aResults)) break;
                 if ($iGroupLoop > 4) break;
                 if ($iQueryLoop > 30) break;
             }
@@ -835,7 +860,7 @@ class Geocode
 
             $oLookup = $oReverse->lookupPoint($oCtx->sqlNear, false);
 
-            if (CONST_Debug) var_dump('Reverse search', $aLookup);
+            Debug::printVar('Reverse search', $oLookup);
 
             if ($oLookup) {
                 $aResults = array($oLookup->iId => $oLookup);
@@ -843,7 +868,7 @@ class Geocode
         }
 
         // No results? Done
-        if (!sizeof($aResults)) {
+        if (empty($aResults)) {
             if ($this->bFallback) {
                 if ($this->fallbackStructuredQuery()) {
                     return $this->lookup();
@@ -864,20 +889,17 @@ class Geocode
 
         $aSearchResults = $this->oPlaceLookup->lookup($aResults);
 
-        $aClassType = getClassTypesWithImportance();
+        $aClassType = ClassTypes\getListWithImportance();
         $aRecheckWords = preg_split('/\b[\s,\\-]*/u', $sQuery);
         foreach ($aRecheckWords as $i => $sWord) {
             if (!preg_match('/[\pL\pN]/', $sWord)) unset($aRecheckWords[$i]);
         }
 
-        if (CONST_Debug) {
-            echo '<i>Recheck words:<\i>';
-            var_dump($aRecheckWords);
-        }
+        Debug::printVar('Recheck words', $aRecheckWords);
 
         foreach ($aSearchResults as $iIdx => $aResult) {
             // Default
-            $fDiameter = getResultDiameter($aResult);
+            $fDiameter = ClassTypes\getProperty($aResult, 'defdiameter', 0.0001);
 
             $aOutlineResult = $this->oPlaceLookup->getOutlines($aResult['place_id'], $aResult['lon'], $aResult['lat'], $fDiameter/2);
             if ($aOutlineResult) {
@@ -891,27 +913,15 @@ class Geocode
             }
 
             // Is there an icon set for this type of result?
-            if (isset($aClassType[$aResult['class'].':'.$aResult['type']]['icon'])
-                && $aClassType[$aResult['class'].':'.$aResult['type']]['icon']
-            ) {
-                $aResult['icon'] = CONST_Website_BaseURL.'images/mapicons/'.$aClassType[$aResult['class'].':'.$aResult['type']]['icon'].'.p.20.png';
-            }
+            $aClassInfo = ClassTypes\getInfo($aResult);
 
-            if (isset($aClassType[$aResult['class'].':'.$aResult['type'].':'.$aResult['admin_level']]['label'])
-                && $aClassType[$aResult['class'].':'.$aResult['type'].':'.$aResult['admin_level']]['label']
-            ) {
-                $aResult['label'] = $aClassType[$aResult['class'].':'.$aResult['type'].':'.$aResult['admin_level']]['label'];
-            } elseif (isset($aClassType[$aResult['class'].':'.$aResult['type']]['label'])
-                && $aClassType[$aResult['class'].':'.$aResult['type']]['label']
-            ) {
-                $aResult['label'] = $aClassType[$aResult['class'].':'.$aResult['type']]['label'];
-            }
-            // if tag '&addressdetails=1' is set in query
-            if ($this->bIncludeAddressDetails) {
-                // getAddressDetails() is defined in lib.php and uses the SQL function get_addressdata in functions.sql
-                $aResult['address'] = getAddressDetails($this->oDB, $sLanguagePrefArraySQL, $aResult['place_id'], $aResult['country_code'], $aResults[$aResult['place_id']]->iHouseNumber);
-                if ($aResult['extra_place'] == 'city' && !isset($aResult['address']['city'])) {
-                    $aResult['address'] = array_merge(array('city' => array_values($aResult['address'])[0]), $aResult['address']);
+            if ($aClassInfo) {
+                if (isset($aClassInfo['icon'])) {
+                    $aResult['icon'] = CONST_Website_BaseURL.'images/mapicons/'.$aClassInfo['icon'].'.p.20.png';
+                }
+
+                if (isset($aClassInfo['label'])) {
+                    $aResult['label'] = $aClassInfo['label'];
                 }
             }
 
@@ -921,12 +931,12 @@ class Geocode
                 $aResult['importance'] = 0.001;
                 $aResult['foundorder'] = $aResult['addressimportance'];
             } else {
-                // Adjust importance for the number of exact string matches in the result
+                $aResult['importance'] = max(0.001, $aResult['importance']);
                 $aResult['importance'] *= $this->viewboxImportanceFactor(
                     $aResult['lon'],
                     $aResult['lat']
                 );
-                $aResult['importance'] = max(0.001, $aResult['importance']);
+                // Adjust importance for the number of exact string matches in the result
                 $iCountWords = 0;
                 $sAddress = $aResult['langaddress'];
                 foreach ($aRecheckWords as $i => $sWord) {
@@ -952,18 +962,16 @@ class Geocode
                     $aResult['foundorder'] += 0.01;
                 }
             }
-            if (CONST_Debug) var_dump($aResult);
             $aSearchResults[$iIdx] = $aResult;
         }
         uasort($aSearchResults, 'byImportance');
+        Debug::printVar('Pre-filter results', $aSearchResults);
 
         $aOSMIDDone = array();
         $aClassTypeNameDone = array();
         $aToFilter = $aSearchResults;
         $aSearchResults = array();
 
-        if (CONST_Debug) var_dump($aToFilter);
-
         $bFirst = true;
         foreach ($aToFilter as $aResult) {
             $this->aExcludePlaceIDs[$aResult['place_id']] = $aResult['place_id'];
@@ -982,10 +990,32 @@ class Geocode
             }
 
             // Absolute limit on number of results
-            if (sizeof($aSearchResults) >= $this->iFinalLimit) break;
+            if (count($aSearchResults) >= $this->iFinalLimit) break;
         }
 
-        if (CONST_Debug) var_dump($aSearchResults);
+        Debug::printVar('Post-filter results', $aSearchResults);
         return $aSearchResults;
     } // end lookup()
+
+    public function debugInfo()
+    {
+        return array(
+                'Query' => $this->sQuery,
+                'Structured query' => $this->aStructuredQuery,
+                'Name keys' => Debug::fmtArrayVals($this->aLangPrefOrder),
+                'Excluded place IDs' => Debug::fmtArrayVals($this->aExcludePlaceIDs),
+                'Try reversed query'=> $this->bReverseInPlan,
+                'Limit (for searches)' => $this->iLimit,
+                'Limit (for results)'=> $this->iFinalLimit,
+                'Country codes' => Debug::fmtArrayVals($this->aCountryCodes),
+                'Bounded search' => $this->bBoundedSearch,
+                'Viewbox' => Debug::fmtArrayVals($this->aViewBox),
+                'Route points' => Debug::fmtArrayVals($this->aRoutePoints),
+                'Route width' => $this->aRouteWidth,
+                'Max rank' => $this->iMaxRank,
+                'Min address rank' => $this->iMinAddressRank,
+                'Max address rank' => $this->iMaxAddressRank,
+                'Address rank list' => Debug::fmtArrayVals($this->aAddressRankList)
+               );
+    }
 } // end class

lib/ParameterParser.php

@@ -23,7 +23,7 @@ class ParameterParser
 
     public function getInt($sName, $bDefault = false)
     {
-        if (!isset($this->aParams[$sName]) || strlen($this->aParams[$sName]) == 0) {
+        if (!isset($this->aParams[$sName])) {
             return $bDefault;
         }
 
@@ -36,7 +36,7 @@ class ParameterParser
 
     public function getFloat($sName, $bDefault = false)
     {
-        if (!isset($this->aParams[$sName]) || strlen($this->aParams[$sName]) == 0) {
+        if (!isset($this->aParams[$sName])) {
             return $bDefault;
         }
 
@@ -74,7 +74,8 @@ class ParameterParser
         $sValue = $this->getString($sName);
 
         if ($sValue) {
-            return explode(',', $sValue);
+            // removes all NULL, FALSE and Empty Strings but leaves 0 (zero) values
+            return array_values(array_filter(explode(',', $sValue), 'strlen'));
         }
 
         return $aDefault;
@@ -90,7 +91,7 @@ class ParameterParser
         $sLangString = $this->getString('accept-language', $sFallback);
 
         if ($sLangString) {
-            if (preg_match_all('/(([a-z]{1,8})(-[a-z]{1,8})?)\s*(;\s*q\s*=\s*(1|0\.[0-9]+))?/i', $sLangString, $aLanguagesParse, PREG_SET_ORDER)) {
+            if (preg_match_all('/(([a-z]{1,8})([-_][a-z]{1,8})?)\s*(;\s*q\s*=\s*(1|0\.[0-9]+))?/i', $sLangString, $aLanguagesParse, PREG_SET_ORDER)) {
                 foreach ($aLanguagesParse as $iLang => $aLanguage) {
                     $aLanguages[$aLanguage[1]] = isset($aLanguage[5])?(float)$aLanguage[5]:1 - ($iLang/100);
                     if (!isset($aLanguages[$aLanguage[2]])) $aLanguages[$aLanguage[2]] = $aLanguages[$aLanguage[1]]/10;
@@ -98,7 +99,7 @@ class ParameterParser
                 arsort($aLanguages);
             }
         }
-        if (!sizeof($aLanguages) && CONST_Default_Language) {
+        if (empty($aLanguages) && CONST_Default_Language) {
             $aLanguages[CONST_Default_Language] = 1;
         }
 

lib/Phrase.php

@@ -83,7 +83,7 @@ class Phrase
         $aResult = array(array(join(' ', $aWords)));
         $sFirstToken = '';
         if ($iDepth < Phrase::MAX_DEPTH) {
-            while (sizeof($aWords) > 1) {
+            while (count($aWords) > 1) {
                 $sWord = array_shift($aWords);
                 $sFirstToken .= ($sFirstToken?' ':'').$sWord;
                 $aRest = $this->createWordSets($aWords, $iDepth + 1);
@@ -101,7 +101,7 @@ class Phrase
         $aResult = array(array(join(' ', $aWords)));
         $sFirstToken = '';
         if ($iDepth < Phrase::MAX_DEPTH) {
-            while (sizeof($aWords) > 1) {
+            while (count($aWords) > 1) {
                 $sWord = array_pop($aWords);
                 $sFirstToken = $sWord.($sFirstToken?' ':'').$sFirstToken;
                 $aRest = $this->createInverseWordSets($aWords, $iDepth + 1);
@@ -113,4 +113,14 @@ class Phrase
 
         return $aResult;
     }
+
+    public function debugInfo()
+    {
+        return array(
+                'Type' => $this->sPhraseType,
+                'Phrase' => $this->sPhrase,
+                'Words' => $this->aWords,
+                'WordSets' => $this->aWordSets
+               );
+    }
 }

lib/PlaceLookup.php

@@ -2,6 +2,7 @@
 
 namespace Nominatim;
 
+require_once(CONST_BasePath.'/lib/AddressDetails.php');
 require_once(CONST_BasePath.'/lib/Result.php');
 
 class PlaceLookup
@@ -42,29 +43,37 @@ class PlaceLookup
         $this->bIncludePolygonAsPoints = $b;
     }
 
+    public function setIncludeAddressDetails($b)
+    {
+        $this->bAddressDetails = $b;
+    }
+
     public function loadParamArray($oParams, $sGeomType = null)
     {
         $aLangs = $oParams->getPreferredLanguages();
         $this->aLangPrefOrderSql =
-            'ARRAY['.join(',', array_map('getDBQuoted', $aLangs)).']';
+            'ARRAY['.join(',', $this->oDB->getDBQuotedList($aLangs)).']';
 
-        $this->bAddressDetails = $oParams->getBool('addressdetails', true);
         $this->bExtraTags = $oParams->getBool('extratags', false);
         $this->bNameDetails = $oParams->getBool('namedetails', false);
 
         $this->bDeDupe = $oParams->getBool('dedupe', $this->bDeDupe);
 
-        if ($sGeomType === null || $sGeomType == 'text') {
-            $this->bIncludePolygonAsText = $oParams->getBool('polygon_text');
-        }
         if ($sGeomType === null || $sGeomType == 'geojson') {
             $this->bIncludePolygonAsGeoJSON = $oParams->getBool('polygon_geojson');
+            $this->bIncludePolygonAsPoints = false;
         }
-        if ($sGeomType === null || $sGeomType == 'kml') {
-            $this->bIncludePolygonAsKML = $oParams->getBool('polygon_kml');
-        }
-        if ($sGeomType === null || $sGeomType == 'svg') {
-            $this->bIncludePolygonAsSVG = $oParams->getBool('polygon_svg');
+
+        if ($oParams->getString('format', '') !== 'geojson') {
+            if ($sGeomType === null || $sGeomType == 'text') {
+                $this->bIncludePolygonAsText = $oParams->getBool('polygon_text');
+            }
+            if ($sGeomType === null || $sGeomType == 'kml') {
+                $this->bIncludePolygonAsKML = $oParams->getBool('polygon_kml');
+            }
+            if ($sGeomType === null || $sGeomType == 'svg') {
+                $this->bIncludePolygonAsSVG = $oParams->getBool('polygon_svg');
+            }
         }
         $this->fPolygonSimplificationThreshold
             = $oParams->getFloat('polygon_threshold', 0.0);
@@ -123,13 +132,9 @@ class PlaceLookup
 
     public function setLanguagePreference($aLangPrefOrder)
     {
-        $this->aLangPrefOrderSql =
-            'ARRAY['.join(',', array_map('getDBQuoted', $aLangPrefOrder)).']';
-    }
-
-    public function setIncludeAddressDetails($bAddressDetails = true)
-    {
-        $this->bAddressDetails = $bAddressDetails;
+        $this->aLangPrefOrderSql = $this->oDB->getArraySQL(
+            $this->oDB->getDBQuotedList($aLangPrefOrder)
+        );
     }
 
     private function addressImportanceSql($sGeometry, $sPlaceId)
@@ -150,13 +155,16 @@ class PlaceLookup
 
     private function langAddressSql($sHousenumber)
     {
+        if ($this->bAddressDetails)
+            return ''; // langaddress will be computed from address details
+
         return 'get_address_by_language(place_id,'.$sHousenumber.','.$this->aLangPrefOrderSql.') AS langaddress,';
     }
 
     public function lookupOSMID($sType, $iID)
     {
-        $sSQL = "select place_id from placex where osm_type = '".$sType."' and osm_id = ".$iID;
-        $iPlaceID = chksql($this->oDB->getOne($sSQL));
+        $sSQL = 'select place_id from placex where osm_type = :type and osm_id = :id';
+        $iPlaceID = $this->oDB->getOne($sSQL, array(':type' => $sType, ':id' => $iID));
 
         if (!$iPlaceID) {
             return null;
@@ -164,19 +172,21 @@ class PlaceLookup
 
         $aResults = $this->lookup(array($iPlaceID => new Result($iPlaceID)));
 
-        return sizeof($aResults) ? reset($aResults) : null;
+        return empty($aResults) ? null : reset($aResults);
     }
 
     public function lookup($aResults, $iMinRank = 0, $iMaxRank = 30)
     {
-        if (!sizeof($aResults)) {
+        Debug::newFunction('Place lookup');
+
+        if (empty($aResults)) {
             return array();
         }
         $aSubSelects = array();
 
         $sPlaceIDs = Result::joinIdsByTable($aResults, Result::TABLE_PLACEX);
-        if (CONST_Debug) var_dump('PLACEX', $sPlaceIDs);
         if ($sPlaceIDs) {
+            Debug::printVar('Ids from placex', $sPlaceIDs);
             $sSQL  = 'SELECT ';
             $sSQL .= '    osm_type,';
             $sSQL .= '    osm_id,';
@@ -233,7 +243,7 @@ class PlaceLookup
             $sSQL .= '     country_code, ';
             $sSQL .= '     importance, ';
             if (!$this->bDeDupe) $sSQL .= 'place_id,';
-            $sSQL .= '     langaddress, ';
+            if (!$this->bAddressDetails) $sSQL .= 'langaddress, ';
             $sSQL .= '     placename, ';
             $sSQL .= '     ref, ';
             if ($this->bExtraTags) $sSQL .= 'extratags, ';
@@ -246,23 +256,24 @@ class PlaceLookup
         // postcode table
         $sPlaceIDs = Result::joinIdsByTable($aResults, Result::TABLE_POSTCODE);
         if ($sPlaceIDs) {
+            Debug::printVar('Ids from location_postcode', $sPlaceIDs);
             $sSQL = 'SELECT';
             $sSQL .= "  'P' as osm_type,";
             $sSQL .= '  (SELECT osm_id from placex p WHERE p.place_id = lp.parent_place_id) as osm_id,';
             $sSQL .= "  'place' as class, 'postcode' as type,";
-            $sSQL .= '  null as admin_level, rank_search, rank_address,';
+            $sSQL .= '  null::smallint as admin_level, rank_search, rank_address,';
             $sSQL .= '  place_id, parent_place_id,';
-            $sSQL .= '  null as housenumber,';
+            $sSQL .= '  -1 as housenumber,';
             $sSQL .= '  country_code,';
             $sSQL .= $this->langAddressSql('-1');
             $sSQL .= '  postcode as placename,';
             $sSQL .= '  postcode as ref,';
-            if ($this->bExtraTags) $sSQL .= 'null AS extra,';
-            if ($this->bNameDetails) $sSQL .= 'null AS names,';
+            if ($this->bExtraTags) $sSQL .= 'null::text AS extra,';
+            if ($this->bNameDetails) $sSQL .= 'null::text AS names,';
             $sSQL .= '  ST_x(geometry) AS lon, ST_y(geometry) AS lat,';
             $sSQL .= '  (0.75-(rank_search::float/40)) AS importance, ';
             $sSQL .= $this->addressImportanceSql('geometry', 'lp.parent_place_id');
-            $sSQL .= '  null AS extra_place ';
+            $sSQL .= '  null::text AS extra_place ';
             $sSQL .= 'FROM location_postcode lp';
             $sSQL .= " WHERE place_id in ($sPlaceIDs) ";
             $sSQL .= "   AND lp.rank_address between $iMinRank and $iMaxRank";
@@ -276,6 +287,7 @@ class PlaceLookup
             if (CONST_Use_US_Tiger_Data) {
                 $sPlaceIDs = Result::joinIdsByTable($aResults, Result::TABLE_TIGER);
                 if ($sPlaceIDs) {
+                    Debug::printVar('Ids from Tiger table', $sPlaceIDs);
                     $sHousenumbers = Result::sqlHouseNumberTable($aResults, Result::TABLE_TIGER);
                     // Tiger search only if a housenumber was searched and if it was found
                     // (realized through a join)
@@ -284,7 +296,7 @@ class PlaceLookup
                     $sSQL .= '     (SELECT osm_id from placex p WHERE p.place_id=blub.parent_place_id) as osm_id, ';
                     $sSQL .= "     'place' AS class, ";
                     $sSQL .= "     'house' AS type, ";
-                    $sSQL .= '     null AS admin_level, ';
+                    $sSQL .= '     null::smallint AS admin_level, ';
                     $sSQL .= '     30 AS rank_search, ';
                     $sSQL .= '     30 AS rank_address, ';
                     $sSQL .= '     place_id, ';
@@ -292,15 +304,15 @@ class PlaceLookup
                     $sSQL .= '     housenumber_for_place as housenumber,';
                     $sSQL .= "     'us' AS country_code, ";
                     $sSQL .= $this->langAddressSql('housenumber_for_place');
-                    $sSQL .= '     null AS placename, ';
-                    $sSQL .= '     null AS ref, ';
-                    if ($this->bExtraTags) $sSQL .= 'null AS extra,';
-                    if ($this->bNameDetails) $sSQL .= 'null AS names,';
+                    $sSQL .= '     null::text AS placename, ';
+                    $sSQL .= '     null::text AS ref, ';
+                    if ($this->bExtraTags) $sSQL .= 'null::text AS extra,';
+                    if ($this->bNameDetails) $sSQL .= 'null::text AS names,';
                     $sSQL .= '     st_x(centroid) AS lon, ';
                     $sSQL .= '     st_y(centroid) AS lat,';
                     $sSQL .= '     -1.15 AS importance, ';
                     $sSQL .= $this->addressImportanceSql('centroid', 'blub.parent_place_id');
-                    $sSQL .= '     null AS extra_place ';
+                    $sSQL .= '     null::text AS extra_place ';
                     $sSQL .= ' FROM (';
                     $sSQL .= '     SELECT place_id, ';    // interpolate the Tiger housenumbers here
                     $sSQL .= '         ST_LineInterpolatePoint(linegeo, (housenumber_for_place-startnumber::float)/(endnumber-startnumber)::float) AS centroid, ';
@@ -321,6 +333,7 @@ class PlaceLookup
             // osmline - interpolated housenumbers
             $sPlaceIDs = Result::joinIdsByTable($aResults, Result::TABLE_OSMLINE);
             if ($sPlaceIDs) {
+                Debug::printVar('Ids from interpolation', $sPlaceIDs);
                 $sHousenumbers = Result::sqlHouseNumberTable($aResults, Result::TABLE_OSMLINE);
                 // interpolation line search only if a housenumber was searched
                 // (realized through a join)
@@ -329,7 +342,7 @@ class PlaceLookup
                 $sSQL .= '  osm_id, ';
                 $sSQL .= "  'place' AS class, ";
                 $sSQL .= "  'house' AS type, ";
-                $sSQL .= '  15 AS admin_level, ';
+                $sSQL .= '  null::smallint AS admin_level, ';
                 $sSQL .= '  30 AS rank_search, ';
                 $sSQL .= '  30 AS rank_address, ';
                 $sSQL .= '  place_id, ';
@@ -337,16 +350,16 @@ class PlaceLookup
                 $sSQL .= '  housenumber_for_place as housenumber,';
                 $sSQL .= '  country_code, ';
                 $sSQL .= $this->langAddressSql('housenumber_for_place');
-                $sSQL .= '  null AS placename, ';
-                $sSQL .= '  null AS ref, ';
-                if ($this->bExtraTags) $sSQL .= 'null AS extra, ';
-                if ($this->bNameDetails) $sSQL .= 'null AS names, ';
+                $sSQL .= '  null::text AS placename, ';
+                $sSQL .= '  null::text AS ref, ';
+                if ($this->bExtraTags) $sSQL .= 'null::text AS extra, ';
+                if ($this->bNameDetails) $sSQL .= 'null::text AS names, ';
                 $sSQL .= '  st_x(centroid) AS lon, ';
                 $sSQL .= '  st_y(centroid) AS lat, ';
                 // slightly smaller than the importance for normal houses
                 $sSQL .= '  -0.1 AS importance, ';
                 $sSQL .= $this->addressImportanceSql('centroid', 'blub.parent_place_id');
-                $sSQL .= '  null AS extra_place ';
+                $sSQL .= '  null::text AS extra_place ';
                 $sSQL .= '  FROM (';
                 $sSQL .= '     SELECT ';
                 $sSQL .= '         osm_id, ';
@@ -378,7 +391,7 @@ class PlaceLookup
                     $sSQL .= '     place_id AS osm_id, ';
                     $sSQL .= "     'place' AS class,";
                     $sSQL .= "     'house' AS type, ";
-                    $sSQL .= '     null AS admin_level, ';
+                    $sSQL .= '     null::smallint AS admin_level, ';
                     $sSQL .= '     30 AS rank_search,';
                     $sSQL .= '     30 AS rank_address, ';
                     $sSQL .= '     place_id,';
@@ -386,10 +399,10 @@ class PlaceLookup
                     $sSQL .= '     housenumber,';
                     $sSQL .= "     'us' AS country_code, ";
                     $sSQL .= $this->langAddressSql('-1');
-                    $sSQL .= '     null AS placename, ';
-                    $sSQL .= '     null AS ref, ';
-                    if ($this->bExtraTags) $sSQL .= 'null AS extra, ';
-                    if ($this->bNameDetails) $sSQL .= 'null AS names, ';
+                    $sSQL .= '     null::text AS placename, ';
+                    $sSQL .= '     null::text AS ref, ';
+                    if ($this->bExtraTags) $sSQL .= 'null::text AS extra, ';
+                    if ($this->bNameDetails) $sSQL .= 'null::text AS names, ';
                     $sSQL .= '     ST_X(centroid) AS lon, ';
                     $sSQL .= '     ST_Y(centroid) AS lat, ';
                     $sSQL .= '     -1.10 AS importance, ';
@@ -397,7 +410,7 @@ class PlaceLookup
                         'centroid',
                         'location_property_aux.parent_place_id'
                     );
-                    $sSQL .= '     null AS extra_place ';
+                    $sSQL .= '     null::text AS extra_place ';
                     $sSQL .= '  FROM location_property_aux ';
                     $sSQL .= "  WHERE place_id in ($sPlaceIDs) ";
 
@@ -406,25 +419,25 @@ class PlaceLookup
             }
         }
 
-        if (CONST_Debug) var_dump($aSubSelects);
-
-        if (!sizeof($aSubSelects)) {
+        if (empty($aSubSelects)) {
             return array();
         }
 
-        $aPlaces = chksql(
-            $this->oDB->getAll(join(' UNION ', $aSubSelects)),
-            'Could not lookup place'
-        );
+        $sSQL = join(' UNION ', $aSubSelects);
+        Debug::printSQL($sSQL);
+        $aPlaces = $this->oDB->getAll($sSQL, null, 'Could not lookup place');
 
-        $aClassType = getClassTypes();
         foreach ($aPlaces as &$aPlace) {
+            $aPlace['importance'] = (float) $aPlace['importance'];
             if ($this->bAddressDetails) {
                 // to get addressdetails for tiger data, the housenumber is needed
-                $aPlace['aAddress'] = $this->getAddressNames(
+                $aPlace['address'] = new AddressDetails(
+                    $this->oDB,
                     $aPlace['place_id'],
-                    $aPlace['housenumber']
+                    $aPlace['housenumber'],
+                    $this->aLangPrefOrderSql
                 );
+                $aPlace['langaddress'] = $aPlace['address']->getLocaleAddress();
             }
 
             if ($this->bExtraTags) {
@@ -443,77 +456,18 @@ class PlaceLookup
                 }
             }
 
-            $sAddressType = '';
-            $sClassType = $aPlace['class'].':'.$aPlace['type'].':'.$aPlace['admin_level'];
-            if (isset($aClassType[$sClassType]) && isset($aClassType[$sClassType]['simplelabel'])) {
-                $sAddressType = $aClassType[$aClassType]['simplelabel'];
-            } else {
-                $sClassType = $aPlace['class'].':'.$aPlace['type'];
-                if (isset($aClassType[$sClassType]) && isset($aClassType[$sClassType]['simplelabel']))
-                    $sAddressType = $aClassType[$sClassType]['simplelabel'];
-                else $sAddressType = $aPlace['class'];
-            }
-
-            $aPlace['addresstype'] = $sAddressType;
+            $aPlace['addresstype'] = ClassTypes\getProperty(
+                $aPlace,
+                'simplelabel',
+                $aPlace['class']
+            );
         }
 
-        if (CONST_Debug) var_dump($aPlaces);
+        Debug::printVar('Places', $aPlaces);
 
         return $aPlaces;
     }
 
-    private function getAddressDetails($iPlaceID, $bAll, $sHousenumber)
-    {
-        $sSQL = 'SELECT *,';
-        $sSQL .= '  get_name_by_language(name,'.$this->aLangPrefOrderSql.') as localname';
-        $sSQL .= ' FROM get_addressdata('.$iPlaceID.','.$sHousenumber.')';
-        if (!$bAll) {
-            $sSQL .= " WHERE isaddress OR type = 'country_code'";
-        }
-        $sSQL .= ' ORDER BY rank_address desc,isaddress DESC';
-
-        return chksql($this->oDB->getAll($sSQL));
-    }
-
-    public function getAddressNames($iPlaceID, $sHousenumber = null)
-    {
-        $aAddressLines = $this->getAddressDetails(
-            $iPlaceID,
-            false,
-            $sHousenumber === null ? -1 : $sHousenumber
-        );
-
-        $aAddress = array();
-        $aFallback = array();
-        $aClassType = getClassTypes();
-        foreach ($aAddressLines as $aLine) {
-            $bFallback = false;
-            $aTypeLabel = false;
-            if (isset($aClassType[$aLine['class'].':'.$aLine['type'].':'.$aLine['admin_level']])) {
-                $aTypeLabel = $aClassType[$aLine['class'].':'.$aLine['type'].':'.$aLine['admin_level']];
-            } elseif (isset($aClassType[$aLine['class'].':'.$aLine['type']])) {
-                $aTypeLabel = $aClassType[$aLine['class'].':'.$aLine['type']];
-            } elseif (isset($aClassType['boundary:administrative:'.((int)($aLine['rank_address']/2))])) {
-                $aTypeLabel = $aClassType['boundary:administrative:'.((int)($aLine['rank_address']/2))];
-                $bFallback = true;
-            } else {
-                $aTypeLabel = array('simplelabel' => 'address'.$aLine['rank_address']);
-                $bFallback = true;
-            }
-            if ($aTypeLabel && ((isset($aLine['localname']) && $aLine['localname']) || (isset($aLine['housenumber']) && $aLine['housenumber']))) {
-                $sTypeLabel = strtolower(isset($aTypeLabel['simplelabel'])?$aTypeLabel['simplelabel']:$aTypeLabel['label']);
-                $sTypeLabel = str_replace(' ', '_', $sTypeLabel);
-                if (!isset($aAddress[$sTypeLabel]) || (isset($aFallback[$sTypeLabel]) && $aFallback[$sTypeLabel]) || $aLine['class'] == 'place') {
-                    $aAddress[$sTypeLabel] = $aLine['localname']?$aLine['localname']:$aLine['housenumber'];
-                }
-                $aFallback[$sTypeLabel] = $bFallback;
-            }
-        }
-        return $aAddress;
-    }
-
-
-
     /* returns an array which will contain the keys
      *   aBoundingBox
      * and may also contain one or more of the keys
@@ -526,7 +480,7 @@ class PlaceLookup
      */
 
 
-    public function getOutlines($iPlaceID, $fLon = null, $fLat = null, $fRadius = null)
+    public function getOutlines($iPlaceID, $fLon = null, $fLat = null, $fRadius = null, $fLonReverse = null, $fLatReverse = null)
     {
 
         $aOutlineResult = array();
@@ -534,24 +488,36 @@ class PlaceLookup
 
         if (CONST_Search_AreaPolygons) {
             // Get the bounding box and outline polygon
-            $sSQL  = 'select place_id,0 as numfeatures,st_area(geometry) as area,';
-            $sSQL .= 'ST_Y(centroid) as centrelat,ST_X(centroid) as centrelon,';
-            $sSQL .= 'ST_YMin(geometry) as minlat,ST_YMax(geometry) as maxlat,';
-            $sSQL .= 'ST_XMin(geometry) as minlon,ST_XMax(geometry) as maxlon';
+            $sSQL = 'select place_id,0 as numfeatures,st_area(geometry) as area,';
+            if ($fLonReverse != null && $fLatReverse != null) {
+                $sSQL .= ' ST_Y(closest_point) as centrelat,';
+                $sSQL .= ' ST_X(closest_point) as centrelon,';
+            } else {
+                $sSQL .= ' ST_Y(centroid) as centrelat, ST_X(centroid) as centrelon,';
+            }
+            $sSQL .= ' ST_YMin(geometry) as minlat,ST_YMax(geometry) as maxlat,';
+            $sSQL .= ' ST_XMin(geometry) as minlon,ST_XMax(geometry) as maxlon';
             if ($this->bIncludePolygonAsGeoJSON) $sSQL .= ',ST_AsGeoJSON(geometry) as asgeojson';
             if ($this->bIncludePolygonAsKML) $sSQL .= ',ST_AsKML(geometry) as askml';
             if ($this->bIncludePolygonAsSVG) $sSQL .= ',ST_AsSVG(geometry) as assvg';
             if ($this->bIncludePolygonAsText || $this->bIncludePolygonAsPoints) $sSQL .= ',ST_AsText(geometry) as astext';
-            $sFrom = ' from placex where place_id = '.$iPlaceID;
+            if ($fLonReverse != null && $fLatReverse != null) {
+                $sFrom = ' from (SELECT * , CASE WHEN (class = \'highway\') AND (ST_GeometryType(geometry) = \'ST_LineString\') THEN ';
+                $sFrom .=' ST_ClosestPoint(geometry, ST_SetSRID(ST_Point('.$fLatReverse.','.$fLonReverse.'),4326))';
+                $sFrom .=' ELSE centroid END AS closest_point';
+                $sFrom .= ' from placex where place_id = '.$iPlaceID.') as plx';
+            } else {
+                $sFrom = ' from placex where place_id = '.$iPlaceID;
+            }
             if ($this->fPolygonSimplificationThreshold > 0) {
                 $sSQL .= ' from (select place_id,centroid,ST_SimplifyPreserveTopology(geometry,'.$this->fPolygonSimplificationThreshold.') as geometry'.$sFrom.') as plx';
             } else {
                 $sSQL .= $sFrom;
             }
 
-            $aPointPolygon = chksql($this->oDB->getRow($sSQL), 'Could not get outline');
+            $aPointPolygon = $this->oDB->getRow($sSQL, null, 'Could not get outline');
 
-            if ($aPointPolygon['place_id']) {
+            if ($aPointPolygon && $aPointPolygon['place_id']) {
                 if ($aPointPolygon['centrelon'] !== null && $aPointPolygon['centrelat'] !== null) {
                     $aOutlineResult['lat'] = $aPointPolygon['centrelat'];
                     $aOutlineResult['lon'] = $aPointPolygon['centrelon'];

lib/Result.php

@@ -27,6 +27,17 @@ class Result
     /// Subranking within the results (the higher the worse).
     public $iResultRank = 0;
 
+    public function debugInfo()
+    {
+        return array(
+                'Table' => $this->iTable,
+                'ID' => $this->iId,
+                'House number' => $this->iHouseNumber,
+                'Exact Matches' => $this->iExactMatches,
+                'Result rank' => $this->iResultRank
+               );
+    }
+
 
     public function __construct($sId, $iTable = Result::TABLE_PLACEX)
     {
@@ -57,4 +68,32 @@ class Result
 
         return $sHousenumbers;
     }
+
+    /**
+     * Split a result array into highest ranked result and the rest
+     *
+     * @param object[] $aResults List of results to split.
+     *
+     * @return array[]
+     */
+    public static function splitResults($aResults)
+    {
+        $aHead = array();
+        $aTail = array();
+        $iMinRank = 10000;
+
+        foreach ($aResults as $oRes) {
+            if ($oRes->iResultRank < $iMinRank) {
+                $aTail = array_merge($aTail, $aHead);
+                $aHead = array($oRes->iId => $oRes);
+                $iMinRank = $oRes->iResultRank;
+            } elseif ($oRes->iResultRank == $iMinRank) {
+                $aHead[$oRes->iId] = $oRes;
+            } else {
+                $aTail[$oRes->iId] = $oRes;
+            }
+        }
+
+        return array('head' => $aHead, 'tail' => $aTail);
+    }
 }

lib/ReverseGeocode.php

@@ -36,8 +36,8 @@ class ReverseGeocode
                       13 => 18,
                       14 => 22, // Suburb
                       15 => 22,
-                      16 => 26, // Street, TODO: major street?
-                      17 => 26,
+                      16 => 26, // major street
+                      17 => 27, // minor street
                       18 => 30, // or >, Building
                       19 => 30, // or >, Building
                      );
@@ -63,12 +63,160 @@ class ReverseGeocode
         $sSQL .= ' and indexed_status = 0 and startnumber is not NULL ';
         $sSQL .= ' ORDER BY distance ASC limit 1';
 
-        return chksql(
-            $this->oDB->getRow($sSQL),
+        return $this->oDB->getRow(
+            $sSQL,
+            null,
             'Could not determine closest housenumber on an osm interpolation line.'
         );
     }
 
+    protected function lookupLargeArea($sPointSQL, $iMaxRank)
+    {
+        $oResult = null;
+
+        if ($iMaxRank > 4) {
+            $aPlace = $this->lookupPolygon($sPointSQL, $iMaxRank);
+            if ($aPlace) {
+                return new Result($aPlace['place_id']);
+            }
+        }
+
+        // If no polygon which contains the searchpoint is found,
+        // searches in the country_osm_grid table for a polygon.
+        return  $this->lookupInCountry($sPointSQL, $iMaxRank);
+    }
+
+    protected function lookupInCountry($sPointSQL, $iMaxRank)
+    {
+        // searches for polygon in table country_osm_grid which contains the searchpoint
+        // and searches for the nearest place node to the searchpoint in this polygon
+        $sSQL = 'SELECT country_code FROM country_osm_grid';
+        $sSQL .= ' WHERE ST_CONTAINS(geometry, '.$sPointSQL.') LIMIT 1';
+
+        $sCountryCode = $this->oDB->getOne(
+            $sSQL,
+            null,
+            'Could not determine country polygon containing the point.'
+        );
+        if ($sCountryCode) {
+            if ($iMaxRank > 4) {
+                // look for place nodes with the given country code
+                $sSQL = 'SELECT place_id FROM';
+                $sSQL .= ' (SELECT place_id, rank_search,';
+                $sSQL .= '         ST_distance('.$sPointSQL.', geometry) as distance';
+                $sSQL .= ' FROM placex';
+                $sSQL .= ' WHERE osm_type = \'N\'';
+                $sSQL .= ' AND country_code = \''.$sCountryCode.'\'';
+                $sSQL .= ' AND rank_search between 5 and ' .min(25, $iMaxRank);
+                $sSQL .= ' AND class = \'place\' AND type != \'postcode\'';
+                $sSQL .= ' AND name IS NOT NULL ';
+                $sSQL .= ' and indexed_status = 0 and linked_place_id is null';
+                $sSQL .= ' AND ST_DWithin('.$sPointSQL.', geometry, 1.8)) p ';
+                $sSQL .= 'WHERE distance <= reverse_place_diameter(rank_search)';
+                $sSQL .= ' ORDER BY rank_search DESC, distance ASC';
+                $sSQL .= ' LIMIT 1';
+
+                if (CONST_Debug) var_dump($sSQL);
+                $aPlace = $this->oDB->getRow($sSQL, null, 'Could not determine place node.');
+                if ($aPlace) {
+                    return new Result($aPlace['place_id']);
+                }
+            }
+
+            // still nothing, then return the country object
+            $sSQL = 'SELECT place_id, ST_distance('.$sPointSQL.', centroid) as distance';
+            $sSQL .= ' FROM placex';
+            $sSQL .= ' WHERE country_code = \''.$sCountryCode.'\'';
+            $sSQL .= ' AND rank_search = 4 AND rank_address = 4';
+            $sSQL .= ' AND class in (\'boundary\',  \'place\')';
+            $sSQL .= ' AND linked_place_id is null';
+            $sSQL .= ' ORDER BY distance ASC';
+
+            if (CONST_Debug) var_dump($sSQL);
+            $aPlace = $this->oDB->getRow($sSQL, null, 'Could not determine place node.');
+            if ($aPlace) {
+                return new Result($aPlace['place_id']);
+            }
+        }
+
+        return null;
+    }
+
+    /**
+     * Search for areas or nodes for areas or nodes between state and suburb level.
+     *
+     * @param string $sPointSQL Search point as SQL string.
+     * @param int    $iMaxRank  Maximum address rank of the feature.
+     *
+     * @return Record of the found feature or null.
+     *
+     * Searches first for polygon that contains the search point.
+     * If such a polygon is found, place nodes with a higher rank are
+     * searched inside the polygon.
+     */
+    protected function lookupPolygon($sPointSQL, $iMaxRank)
+    {
+        // polygon search begins at suburb-level
+        if ($iMaxRank > 25) $iMaxRank = 25;
+        // no polygon search over country-level
+        if ($iMaxRank < 5) $iMaxRank = 5;
+        // search for polygon
+        $sSQL = 'SELECT place_id, parent_place_id, rank_address, rank_search FROM';
+        $sSQL .= '(select place_id, parent_place_id, rank_address, rank_search, country_code, geometry';
+        $sSQL .= ' FROM placex';
+        $sSQL .= ' WHERE ST_GeometryType(geometry) in (\'ST_Polygon\', \'ST_MultiPolygon\')';
+        $sSQL .= ' AND rank_address Between 5 AND ' .$iMaxRank;
+        $sSQL .= ' AND geometry && '.$sPointSQL;
+        $sSQL .= ' AND type != \'postcode\' ';
+        $sSQL .= ' AND name is not null';
+        $sSQL .= ' AND indexed_status = 0 and linked_place_id is null';
+        $sSQL .= ' ORDER BY rank_address DESC LIMIT 50 ) as a';
+        $sSQL .= ' WHERE ST_CONTAINS(geometry, '.$sPointSQL.' )';
+        $sSQL .= ' ORDER BY rank_address DESC LIMIT 1';
+
+        $aPoly = $this->oDB->getRow($sSQL, null, 'Could not determine polygon containing the point.');
+
+        if ($aPoly) {
+        // if a polygon is found, search for placenodes begins ...
+            $iParentPlaceID = $aPoly['parent_place_id'];
+            $iRankAddress = $aPoly['rank_address'];
+            $iRankSearch = $aPoly['rank_search'];
+            $iPlaceID = $aPoly['place_id'];
+
+            if ($iRankAddress != $iMaxRank) {
+                $sSQL = 'SELECT place_id FROM ';
+                $sSQL .= '(SELECT place_id, rank_search, country_code, geometry,';
+                $sSQL .= ' ST_distance('.$sPointSQL.', geometry) as distance';
+                $sSQL .= ' FROM placex';
+                $sSQL .= ' WHERE osm_type = \'N\'';
+                // using rank_search because of a better differentiation
+                // for place nodes at rank_address 16
+                $sSQL .= ' AND rank_search > '.$iRankSearch;
+                $sSQL .= ' AND rank_search <= '.$iMaxRank;
+                $sSQL .= ' AND class = \'place\'';
+                $sSQL .= ' AND type != \'postcode\'';
+                $sSQL .= ' AND name IS NOT NULL ';
+                $sSQL .= ' AND indexed_status = 0 AND linked_place_id is null';
+                $sSQL .= ' AND ST_DWithin('.$sPointSQL.', geometry, reverse_place_diameter('.$iRankSearch.'::smallint))';
+                $sSQL .= ' ORDER BY distance ASC,';
+                $sSQL .= ' rank_address DESC';
+                $sSQL .= ' limit 500) as a';
+                $sSQL .= ' WHERE ST_CONTAINS((SELECT geometry FROM placex WHERE place_id = '.$iPlaceID.'), geometry )';
+                $sSQL .= ' AND distance <= reverse_place_diameter(rank_search)';
+                $sSQL .= ' ORDER BY distance ASC, rank_search DESC';
+                $sSQL .= ' LIMIT 1';
+
+                if (CONST_Debug) var_dump($sSQL);
+                $aPlacNode = $this->oDB->getRow($sSQL, null, 'Could not determine place node.');
+                if ($aPlacNode) {
+                    return $aPlacNode;
+                }
+            }
+        }
+        return $aPoly;
+    }
+
+
     public function lookup($fLat, $fLon, $bDoInterpolation = true)
     {
         return $this->lookupPoint(
@@ -79,145 +227,121 @@ class ReverseGeocode
 
     public function lookupPoint($sPointSQL, $bDoInterpolation = true)
     {
+        // starts if the search is on POI or street level,
+        // searches for the nearest POI or street,
+        // if a street is found and a POI is searched for,
+        // the nearest POI which the found street is a parent of is choosen.
         $iMaxRank = $this->iMaxRank;
 
         // Find the nearest point
-        $fSearchDiam = 0.0004;
+        $fSearchDiam = 0.006;
         $oResult = null;
         $aPlace = null;
-        $fMaxAreaDistance = 1;
-        $bIsTigerStreet = false;
-        while ($oResult === null && $fSearchDiam < $fMaxAreaDistance) {
-            $fSearchDiam = $fSearchDiam * 2;
 
-            // If we have to expand the search area by a large amount then we need a larger feature
-            // then there is a limit to how small the feature should be
-            if ($fSearchDiam > 2 && $iMaxRank > 4) $iMaxRank = 4;
-            if ($fSearchDiam > 1 && $iMaxRank > 9) $iMaxRank = 8;
-            if ($fSearchDiam > 0.8 && $iMaxRank > 10) $iMaxRank = 10;
-            if ($fSearchDiam > 0.6 && $iMaxRank > 12) $iMaxRank = 12;
-            if ($fSearchDiam > 0.2 && $iMaxRank > 17) $iMaxRank = 17;
-            if ($fSearchDiam > 0.1 && $iMaxRank > 18) $iMaxRank = 18;
-            if ($fSearchDiam > 0.008 && $iMaxRank > 22) $iMaxRank = 22;
-            if ($fSearchDiam > 0.001 && $iMaxRank > 26) {
-                // try with interpolations before continuing
-                if ($bDoInterpolation) {
-                    $aHouse = $this->lookupInterpolation($sPointSQL, $fSearchDiam/2);
-
-                    if ($aHouse) {
-                        $oResult = new Result($aHouse['place_id'], Result::TABLE_OSMLINE);
-                        $oResult->iHouseNumber = closestHouseNumber($aHouse);
-
-                        $aPlace = $aHouse;
-                        $iParentPlaceID = $aHouse['parent_place_id']; // the street
-                        $iMaxRank = 30;
-
-                        break;
-                    }
-                }
-                // no interpolation found, continue search
-                $iMaxRank = 26;
-            }
-
-            $sSQL = 'select place_id,parent_place_id,rank_search,country_code,';
-            $sSQL .= '  ST_distance('.$sPointSQL.', geometry) as distance';
+        // for POI or street level
+        if ($iMaxRank >= 26) {
+            $sSQL = 'select place_id,parent_place_id,rank_address,country_code,';
+            $sSQL .= ' ST_distance('.$sPointSQL.', geometry) as distance';
             $sSQL .= ' FROM ';
-            if ($fSearchDiam < 0.01) {
-                $sSQL .= ' placex';
-                $sSQL .= '   WHERE ST_DWithin('.$sPointSQL.', geometry, '.$fSearchDiam.')';
-                $sSQL .= '   AND';
-            } else {
-                $sSQL .= ' (SELECT * FROM placex ';
-                $sSQL .= '   WHERE ST_DWithin('.$sPointSQL.', geometry, '.$fSearchDiam.')';
-                $sSQL .= '   LIMIT 1000) as p WHERE';
-            }
-            $sSQL .= ' rank_search != 28 and rank_search >= '.$iMaxRank;
+            $sSQL .= ' placex';
+            $sSQL .= '   WHERE ST_DWithin('.$sPointSQL.', geometry, '.$fSearchDiam.')';
+            $sSQL .= '   AND';
+            $sSQL .= ' rank_address between 26 and '.$iMaxRank;
             $sSQL .= ' and (name is not null or housenumber is not null';
-            $sSQL .= '      or rank_search between 26 and 27)';
-            $sSQL .= ' and class not in (\'waterway\',\'railway\',\'tunnel\',\'bridge\',\'man_made\')';
+            $sSQL .= ' or rank_address between 26 and 27)';
+            $sSQL .= ' and (rank_address between 26 and 27';
+            $sSQL .= '      or ST_GeometryType(geometry) != \'ST_LineString\')';
+            $sSQL .= ' and class not in (\'boundary\')';
             $sSQL .= ' and indexed_status = 0 and linked_place_id is null';
             $sSQL .= ' and (ST_GeometryType(geometry) not in (\'ST_Polygon\',\'ST_MultiPolygon\') ';
             $sSQL .= ' OR ST_DWithin('.$sPointSQL.', centroid, '.$fSearchDiam.'))';
             $sSQL .= ' ORDER BY distance ASC limit 1';
             if (CONST_Debug) var_dump($sSQL);
-            $aPlace = chksql(
-                $this->oDB->getRow($sSQL),
-                'Could not determine closest place.'
-            );
-            if ($aPlace) {
-                $oResult = new Result($aPlace['place_id']);
-                $iParentPlaceID = $aPlace['parent_place_id'];
-                if ($bDoInterpolation) {
-                    if ($aPlace['rank_search'] == 26 || $aPlace['rank_search'] == 27) {
-                        $bIsTigerStreet = ($aPlace['country_code'] == 'us');
-                    } elseif ($aPlace['rank_search'] == 30) {
-                        // If a house was found, make sure there isn't an
-                        // interpolation line that is closer.
-                        $aHouse = $this->lookupInterpolation(
-                            $sPointSQL,
-                            $aPlace['distance']
-                        );
-                        if ($aHouse && $aPlace['distance'] < $aHouse['distance']) {
-                            $oResult = new Result(
-                                $aHouse['place_id'],
-                                Result::TABLE_OSMLINE
-                            );
-                            $oResult->iHouseNumber = closestHouseNumber($aHouse);
+            $aPlace = $this->oDB->getRow($sSQL, null, 'Could not determine closest place.');
 
-                            $aPlace = $aHouse;
-                            $iParentPlaceID = $aHouse['parent_place_id'];
-                        }
-                    }
+            if (CONST_Debug) var_dump($aPlace);
+            if ($aPlace) {
+                $iPlaceID = $aPlace['place_id'];
+                $oResult = new Result($iPlaceID);
+                $iRankAddress = $aPlace['rank_address'];
+                $iParentPlaceID = $aPlace['parent_place_id'];
+            }
+
+            if ($bDoInterpolation && $iMaxRank >= 30) {
+                $fDistance = $fSearchDiam;
+                if ($aPlace) {
+                    // We can't reliably go from the closest street to an
+                    // interpolation line because the closest interpolation
+                    // may have a different street segments as a parent.
+                    // Therefore allow an interpolation line to take precendence
+                    // even when the street is closer.
+                    $fDistance = $iRankAddress < 28 ? 0.001 : $aPlace['distance'];
+                }
+
+                $aHouse = $this->lookupInterpolation($sPointSQL, $fDistance);
+
+                if ($aHouse) {
+                    $oResult = new Result($aHouse['place_id'], Result::TABLE_OSMLINE);
+                    $oResult->iHouseNumber = closestHouseNumber($aHouse);
+                    $aPlace = $aHouse;
+                    $iRankAddress = 30;
                 }
             }
-        }
 
-        // Only street found? In the US we can check TIGER data for nearest housenumber
-        if (CONST_Use_US_Tiger_Data && $bIsTigerStreet && $this->iMaxRank >= 28) {
-            $fSearchDiam = $aPlace['rank_search'] > 28 ? $aPlace['distance'] : 0.001;
-            $sSQL = 'SELECT place_id,parent_place_id,30 as rank_search,';
-            $sSQL .= 'ST_LineLocatePoint(linegeo,'.$sPointSQL.') as fraction,';
-            $sSQL .= 'ST_distance('.$sPointSQL.', linegeo) as distance,';
-            $sSQL .= 'startnumber,endnumber,interpolationtype';
-            $sSQL .= ' FROM location_property_tiger WHERE parent_place_id = '.$oResult->iId;
-            $sSQL .= ' AND ST_DWithin('.$sPointSQL.', linegeo, '.$fSearchDiam.')';
-            $sSQL .= ' ORDER BY distance ASC limit 1';
+            if ($aPlace) {
+                // if street and maxrank > streetlevel
+                if ($iRankAddress <= 27 && $iMaxRank > 27) {
+                    // find the closest object (up to a certain radius) of which the street is a parent of
+                    $sSQL = ' select place_id,';
+                    $sSQL .= ' ST_distance('.$sPointSQL.', geometry) as distance';
+                    $sSQL .= ' FROM ';
+                    $sSQL .= ' placex';
+                    // radius ?
+                    $sSQL .= ' WHERE ST_DWithin('.$sPointSQL.', geometry, 0.001)';
+                    $sSQL .= ' AND parent_place_id = '.$iPlaceID;
+                    $sSQL .= ' and rank_address > 28';
+                    $sSQL .= ' and ST_GeometryType(geometry) != \'ST_LineString\'';
+                    $sSQL .= ' and (name is not null or housenumber is not null)';
+                    $sSQL .= ' and class not in (\'boundary\')';
+                    $sSQL .= ' and indexed_status = 0 and linked_place_id is null';
+                    $sSQL .= ' ORDER BY distance ASC limit 1';
+                    if (CONST_Debug) var_dump($sSQL);
+                    $aStreet = $this->oDB->getRow($sSQL, null, 'Could not determine closest place.');
+                    if ($aStreet) {
+                        if (CONST_Debug) var_dump($aStreet);
+                        $oResult = new Result($aStreet['place_id']);
+                    }
+                }
 
-            if (CONST_Debug) var_dump($sSQL);
-
-            $aPlaceTiger = chksql(
-                $this->oDB->getRow($sSQL),
-                'Could not determine closest Tiger place.'
-            );
-            if ($aPlaceTiger) {
-                if (CONST_Debug) var_dump('found Tiger housenumber', $aPlaceTiger);
-                $aPlace = $aPlaceTiger;
-                $oResult = new Result($aPlace['place_id'], Result::TABLE_TIGER);
-                $oResult->iHouseNumber = closestHouseNumber($aPlaceTiger);
-                $iParentPlaceID = $aPlace['parent_place_id'];
-                $iMaxRank = 30;
-            }
-        }
-
-        // The point we found might be too small - use the address to find what it is a child of
-        if ($oResult !== null && $iMaxRank < 28) {
-            if ($aPlace['rank_search'] > 28 && $iParentPlaceID) {
-                $iPlaceID = $iParentPlaceID;
+                  // In the US we can check TIGER data for nearest housenumber
+                if (CONST_Use_US_Tiger_Data
+                    && $iRankAddress <= 27
+                    && $aPlace['country_code'] == 'us'
+                    && $this->iMaxRank >= 28
+                ) {
+                    $sSQL = 'SELECT place_id,parent_place_id,30 as rank_search,';
+                    $sSQL .= 'ST_LineLocatePoint(linegeo,'.$sPointSQL.') as fraction,';
+                    $sSQL .= 'ST_distance('.$sPointSQL.', linegeo) as distance,';
+                    $sSQL .= 'startnumber,endnumber,interpolationtype';
+                    $sSQL .= ' FROM location_property_tiger WHERE parent_place_id = '.$oResult->iId;
+                    $sSQL .= ' AND ST_DWithin('.$sPointSQL.', linegeo, 0.001)';
+                    $sSQL .= ' ORDER BY distance ASC limit 1';
+                    if (CONST_Debug) var_dump($sSQL);
+                    $aPlaceTiger = $this->oDB->getRow($sSQL, null, 'Could not determine closest Tiger place.');
+                    if ($aPlaceTiger) {
+                        if (CONST_Debug) var_dump('found Tiger housenumber', $aPlaceTiger);
+                        $oResult = new Result($aPlaceTiger['place_id'], Result::TABLE_TIGER);
+                        $oResult->iHouseNumber = closestHouseNumber($aPlaceTiger);
+                    }
+                }
             } else {
-                $iPlaceID = $oResult->iId;
-            }
-            $sSQL  = 'select a.address_place_id';
-            $sSQL .= ' FROM place_addressline a, placex p';
-            $sSQL .= " WHERE a.place_id = $iPlaceID and a.address_place_id = p.place_id";
-            $sSQL .= '   AND p.linked_place_id is null';
-            $sSQL .= " ORDER BY abs(cached_rank_address - $iMaxRank) asc,cached_rank_address desc,isaddress desc,distance desc";
-            $sSQL .= ' LIMIT 1';
-            $iPlaceID = chksql($this->oDB->getOne($sSQL), 'Could not get parent for place.');
-            if ($iPlaceID) {
-                $oResult = new Result($iPlaceID);
+                // if no POI or street is found ...
+                $oResult = $this->lookupLargeArea($sPointSQL, 25);
             }
+        } else {
+            // lower than street level ($iMaxRank < 26 )
+            $oResult = $this->lookupLargeArea($sPointSQL, $iMaxRank);
         }
-
         return $oResult;
     }
 }

lib/SearchContext.php

@@ -47,7 +47,7 @@ class SearchContext
     /**
      * Get radius around reference point.
      *
-     * @return float Search radius around refernce point.
+     * @return float Search radius around reference point.
      */
     public function nearRadius()
     {
@@ -108,8 +108,8 @@ class SearchContext
             $aViewBox[3]
         );
 
-        $fHeight = $aViewBox[0] - $aViewBox[2];
-        $fWidth = $aViewBox[1] - $aViewBox[3];
+        $fHeight = abs($aViewBox[0] - $aViewBox[2]);
+        $fWidth = abs($aViewBox[1] - $aViewBox[3]);
 
         $this->sqlViewboxLarge = sprintf(
             'ST_SetSRID(ST_MakeBox2D(ST_Point(%F,%F),ST_Point(%F,%F)),4326)',
@@ -126,7 +126,7 @@ class SearchContext
      * The viewbox may be bounded which means that no search results
      * must be outside the viewbox.
      *
-     * @param object   $oDB          DB connection to use for computing the box.
+     * @param object   $oDB          Nominatim::DB instance to use for computing the box.
      * @param string[] $aRoutePoints List of x,y coordinates along a route.
      * @param float    $fRouteWidth  Buffer around the route to use.
      * @param bool     $bBounded     True if the viewbox bounded.
@@ -146,11 +146,11 @@ class SearchContext
         $this->sqlViewboxCentre .= ")'::geometry,4326)";
 
         $sSQL = 'ST_BUFFER('.$this->sqlViewboxCentre.','.($fRouteWidth/69).')';
-        $sGeom = chksql($oDB->getOne('select '.$sSQL), 'Could not get small viewbox');
+        $sGeom = $oDB->getOne('select '.$sSQL, null, 'Could not get small viewbox');
         $this->sqlViewboxSmall = "'".$sGeom."'::geometry";
 
         $sSQL = 'ST_BUFFER('.$this->sqlViewboxCentre.','.($fRouteWidth/30).')';
-        $sGeom = chksql($oDB->getOne('select '.$sSQL), 'Could not get large viewbox');
+        $sGeom = $oDB->getOne('select '.$sSQL, null, 'Could not get large viewbox');
         $this->sqlViewboxLarge = "'".$sGeom."'::geometry";
     }
 
@@ -267,4 +267,18 @@ class SearchContext
 
         return '';
     }
+
+    public function debugInfo()
+    {
+        return array(
+                'Near radius' => $this->fNearRadius,
+                'Near point (SQL)' => $this->sqlNear,
+                'Bounded viewbox' => $this->bViewboxBounded,
+                'Viewbox (SQL, small)' => $this->sqlViewboxSmall,
+                'Viewbox (SQL, large)' => $this->sqlViewboxLarge,
+                'Viewbox (SQL, centre)' => $this->sqlViewboxCentre,
+                'Countries (SQL)' => $this->sqlCountryList,
+                'Excluded IDs (SQL)' => $this->sqlExcludeList
+               );
+    }
 }

lib/SearchDescription.php

@@ -17,6 +17,8 @@ class SearchDescription
     private $sCountryCode = '';
     /// List of word ids making up the name of the object.
     private $aName = array();
+    /// True if the name is rare enough to force index use on name.
+    private $bRareName = false;
     /// List of word ids making up the address of the object.
     private $aAddress = array();
     /// Subset of word ids of full words making up the address.
@@ -43,7 +45,6 @@ class SearchDescription
     /// Index of phrase currently processed.
     private $iNamePhrase = -1;
 
-
     /**
      * Create an empty search description.
      *
@@ -58,7 +59,7 @@ class SearchDescription
     /**
      * Get current search rank.
      *
-     * The higher the search rank the lower the likelyhood that the
+     * The higher the search rank the lower the likelihood that the
      * search is a correct interpretation of the search query.
      *
      * @return integer Search rank.
@@ -94,8 +95,8 @@ class SearchDescription
      */
     public function looksLikeFullAddress()
     {
-        return sizeof($this->aName)
-               && (sizeof($this->aAddress || $this->sCountryCode))
+        return (!empty($this->aName))
+               && (!empty($this->aAddress) || $this->sCountryCode)
                && preg_match('/[0-9]+/', $this->sHouseNumber);
     }
 
@@ -147,7 +148,7 @@ class SearchDescription
      */
     public function isValidSearch()
     {
-        if (!sizeof($this->aName)) {
+        if (empty($this->aName)) {
             if ($this->sHouseNumber) {
                 return false;
             }
@@ -165,30 +166,29 @@ class SearchDescription
     /**
      * Derive new searches by adding a full term to the existing search.
      *
-     * @param mixed[] $aSearchTerm  Description of the token.
-     * @param bool    $bHasPartial  True if there are also tokens of partial terms
-     *                              with the same name.
-     * @param string  $sPhraseType  Type of phrase the token is contained in.
-     * @param bool    $bFirstToken  True if the token is at the beginning of the
-     *                              query.
-     * @param bool    $bFirstPhrase True if the token is in the first phrase of
-     *                              the query.
-     * @param bool    $bLastToken   True if the token is at the end of the query.
+     * @param object $oSearchTerm  Description of the token.
+     * @param bool   $bHasPartial  True if there are also tokens of partial terms
+     *                             with the same name.
+     * @param string $sPhraseType  Type of phrase the token is contained in.
+     * @param bool   $bFirstToken  True if the token is at the beginning of the
+     *                             query.
+     * @param bool   $bFirstPhrase True if the token is in the first phrase of
+     *                             the query.
+     * @param bool   $bLastToken   True if the token is at the end of the query.
      *
      * @return SearchDescription[] List of derived search descriptions.
      */
-    public function extendWithFullTerm($aSearchTerm, $bHasPartial, $sPhraseType, $bFirstToken, $bFirstPhrase, $bLastToken)
+    public function extendWithFullTerm($oSearchTerm, $bHasPartial, $sPhraseType, $bFirstToken, $bFirstPhrase, $bLastToken)
     {
         $aNewSearches = array();
 
         if (($sPhraseType == '' || $sPhraseType == 'country')
-            && !empty($aSearchTerm['country_code'])
-            && $aSearchTerm['country_code'] != '0'
+            && is_a($oSearchTerm, '\Nominatim\Token\Country')
         ) {
             if (!$this->sCountryCode) {
                 $oSearch = clone $this;
                 $oSearch->iSearchRank++;
-                $oSearch->sCountryCode = $aSearchTerm['country_code'];
+                $oSearch->sCountryCode = $oSearchTerm->sCountryCode;
                 // Country is almost always at the end of the string
                 // - increase score for finding it anywhere else (optimisation)
                 if (!$bLastToken) {
@@ -197,15 +197,12 @@ class SearchDescription
                 $aNewSearches[] = $oSearch;
             }
         } elseif (($sPhraseType == '' || $sPhraseType == 'postalcode')
-                  && $aSearchTerm['class'] == 'place' && $aSearchTerm['type'] == 'postcode'
+                  && is_a($oSearchTerm, '\Nominatim\Token\Postcode')
         ) {
             // We need to try the case where the postal code is the primary element
             // (i.e. no way to tell if it is (postalcode, city) OR (city, postalcode)
             // so try both.
-            if (!$this->sPostcode
-                && $aSearchTerm['word']
-                && pg_escape_string($aSearchTerm['word']) == $aSearchTerm['word']
-            ) {
+            if (!$this->sPostcode) {
                 // If we have structured search or this is the first term,
                 // make the postcode the primary search element.
                 if ($this->iOperator == Operator::NONE
@@ -216,73 +213,82 @@ class SearchDescription
                     $oSearch->iOperator = Operator::POSTCODE;
                     $oSearch->aAddress = array_merge($this->aAddress, $this->aName);
                     $oSearch->aName =
-                        array($aSearchTerm['word_id'] => $aSearchTerm['word']);
+                        array($oSearchTerm->iId => $oSearchTerm->sPostcode);
                     $aNewSearches[] = $oSearch;
                 }
 
                 // If we have a structured search or this is not the first term,
                 // add the postcode as an addendum.
                 if ($this->iOperator != Operator::POSTCODE
-                    && ($sPhraseType == 'postalcode' || sizeof($this->aName))
+                    && ($sPhraseType == 'postalcode' || !empty($this->aName))
                 ) {
                     $oSearch = clone $this;
                     $oSearch->iSearchRank++;
-                    $oSearch->sPostcode = $aSearchTerm['word'];
+                    $oSearch->sPostcode = $oSearchTerm->sPostcode;
                     $aNewSearches[] = $oSearch;
                 }
             }
         } elseif (($sPhraseType == '' || $sPhraseType == 'street')
-                 && $aSearchTerm['class'] == 'place' && $aSearchTerm['type'] == 'house'
+                 && is_a($oSearchTerm, '\Nominatim\Token\HouseNumber')
         ) {
             if (!$this->sHouseNumber && $this->iOperator != Operator::POSTCODE) {
                 $oSearch = clone $this;
                 $oSearch->iSearchRank++;
-                $oSearch->sHouseNumber = trim($aSearchTerm['word_token']);
+                $oSearch->sHouseNumber = $oSearchTerm->sToken;
                 // sanity check: if the housenumber is not mainly made
                 // up of numbers, add a penalty
-                if (preg_match_all('/[^0-9]/', $oSearch->sHouseNumber, $aMatches) > 2) {
+                if (preg_match('/\\d/', $oSearch->sHouseNumber) === 0
+                    || preg_match_all('/[^0-9]/', $oSearch->sHouseNumber, $aMatches) > 2) {
                     $oSearch->iSearchRank++;
                 }
-                if (!isset($aSearchTerm['word_id'])) {
+                if (empty($oSearchTerm->iId)) {
                     $oSearch->iSearchRank++;
                 }
                 // also must not appear in the middle of the address
-                if (sizeof($this->aAddress)
-                    || sizeof($this->aAddressNonSearch)
+                if (!empty($this->aAddress)
+                    || (!empty($this->aAddressNonSearch))
                     || $this->sPostcode
                 ) {
                     $oSearch->iSearchRank++;
                 }
                 $aNewSearches[] = $oSearch;
             }
-        } elseif ($sPhraseType == '' && $aSearchTerm['class']) {
+        } elseif ($sPhraseType == ''
+                  && is_a($oSearchTerm, '\Nominatim\Token\SpecialTerm')
+        ) {
             if ($this->iOperator == Operator::NONE) {
                 $oSearch = clone $this;
                 $oSearch->iSearchRank++;
 
-                $iOp = Operator::NEAR; // near == in for the moment
-                if ($aSearchTerm['operator'] == '') {
-                    if (sizeof($this->aName) || $this->oContext->isBoundedSearch()) {
+                $iOp = $oSearchTerm->iOperator;
+                if ($iOp == Operator::NONE) {
+                    if (!empty($this->aName) || $this->oContext->isBoundedSearch()) {
                         $iOp = Operator::NAME;
+                    } else {
+                        $iOp = Operator::NEAR;
                     }
                     $oSearch->iSearchRank += 2;
                 }
 
-                $oSearch->setPoiSearch($iOp, $aSearchTerm['class'], $aSearchTerm['type']);
+                $oSearch->setPoiSearch(
+                    $iOp,
+                    $oSearchTerm->sClass,
+                    $oSearchTerm->sType
+                );
                 $aNewSearches[] = $oSearch;
             }
-        } elseif (isset($aSearchTerm['word_id'])
-                  && $aSearchTerm['word_id']
-                  && $sPhraseType != 'country'
+        } elseif ($sPhraseType != 'country'
+                  && is_a($oSearchTerm, '\Nominatim\Token\Word')
         ) {
-            $iWordID = $aSearchTerm['word_id'];
-            if (sizeof($this->aName)) {
-                if (($sPhraseType == '' || !$bFirstPhrase)
-                    && $sPhraseType != 'country'
-                    && !$bHasPartial
-                ) {
+            $iWordID = $oSearchTerm->iId;
+            // Full words can only be a name if they appear at the beginning
+            // of the phrase. In structured search the name must forcably in
+            // the first phrase. In unstructured search it may be in a later
+            // phrase when the first phrase is a house number.
+            if (!empty($this->aName) || !($bFirstPhrase || $sPhraseType == '')) {
+                if (($sPhraseType == '' || !$bFirstPhrase) && !$bHasPartial) {
                     $oSearch = clone $this;
-                    $oSearch->iSearchRank++;
+                    $oSearch->iSearchRank += 2;
                     $oSearch->aAddress[$iWordID] = $iWordID;
                     $aNewSearches[] = $oSearch;
                 } else {
@@ -292,6 +298,11 @@ class SearchDescription
                 $oSearch = clone $this;
                 $oSearch->iSearchRank++;
                 $oSearch->aName = array($iWordID => $iWordID);
+                if (CONST_Search_NameOnlySearchFrequencyThreshold) {
+                    $oSearch->bRareName =
+                        $oSearchTerm->iSearchNameCount
+                          < CONST_Search_NameOnlySearchFrequencyThreshold;
+                }
                 $aNewSearches[] = $oSearch;
             }
         }
@@ -302,7 +313,8 @@ class SearchDescription
     /**
      * Derive new searches by adding a partial term to the existing search.
      *
-     * @param mixed[] $aSearchTerm        Description of the token.
+     * @param string  $sToken             Term for the token.
+     * @param object  $oSearchTerm        Description of the token.
      * @param bool    $bStructuredPhrases True if the search is structured.
      * @param integer $iPhrase            Number of the phrase the token is in.
      * @param array[] $aFullTokens        List of full term tokens with the
@@ -310,21 +322,21 @@ class SearchDescription
      *
      * @return SearchDescription[] List of derived search descriptions.
      */
-    public function extendWithPartialTerm($aSearchTerm, $bStructuredPhrases, $iPhrase, $aFullTokens)
+    public function extendWithPartialTerm($sToken, $oSearchTerm, $bStructuredPhrases, $iPhrase, $aFullTokens)
     {
         // Only allow name terms.
-        if (!(isset($aSearchTerm['word_id']) && $aSearchTerm['word_id'])) {
+        if (!(is_a($oSearchTerm, '\Nominatim\Token\Word'))) {
             return array();
         }
 
         $aNewSearches = array();
-        $iWordID = $aSearchTerm['word_id'];
+        $iWordID = $oSearchTerm->iId;
 
         if ((!$bStructuredPhrases || $iPhrase > 0)
-            && sizeof($this->aName)
-            && strpos($aSearchTerm['word_token'], ' ') === false
+            && (!empty($this->aName))
+            && strpos($sToken, ' ') === false
         ) {
-            if ($aSearchTerm['search_name_count'] + 1 < CONST_Max_Word_Frequency) {
+            if ($oSearchTerm->iSearchNameCount < CONST_Max_Word_Frequency) {
                 $oSearch = clone $this;
                 $oSearch->iSearchRank += 2;
                 $oSearch->aAddress[$iWordID] = $iWordID;
@@ -333,23 +345,21 @@ class SearchDescription
                 $oSearch = clone $this;
                 $oSearch->iSearchRank++;
                 $oSearch->aAddressNonSearch[$iWordID] = $iWordID;
-                if (preg_match('#^[0-9]+$#', $aSearchTerm['word_token'])) {
+                if (preg_match('#^[0-9]+$#', $sToken)) {
                     $oSearch->iSearchRank += 2;
                 }
-                if (sizeof($aFullTokens)) {
+                if (!empty($aFullTokens)) {
                     $oSearch->iSearchRank++;
                 }
                 $aNewSearches[] = $oSearch;
 
                 // revert to the token version?
-                foreach ($aFullTokens as $aSearchTermToken) {
-                    if (empty($aSearchTermToken['country_code'])
-                        && empty($aSearchTermToken['lat'])
-                        && empty($aSearchTermToken['class'])
-                    ) {
+                foreach ($aFullTokens as $oSearchTermToken) {
+                    if (is_a($oSearchTermToken, '\Nominatim\Token\Word')) {
                         $oSearch = clone $this;
                         $oSearch->iSearchRank++;
-                        $oSearch->aAddress[$aSearchTermToken['word_id']] = $aSearchTermToken['word_id'];
+                        $oSearch->aAddress[$oSearchTermToken->iId]
+                            = $oSearchTermToken->iId;
                         $aNewSearches[] = $oSearch;
                     }
                 }
@@ -357,17 +367,26 @@ class SearchDescription
         }
 
         if ((!$this->sPostcode && !$this->aAddress && !$this->aAddressNonSearch)
-            && (!sizeof($this->aName) || $this->iNamePhrase == $iPhrase)
+            && (empty($this->aName) || $this->iNamePhrase == $iPhrase)
         ) {
             $oSearch = clone $this;
             $oSearch->iSearchRank += 2;
-            if (!sizeof($this->aName)) {
+            if (empty($this->aName)) {
                 $oSearch->iSearchRank += 1;
             }
-            if (preg_match('#^[0-9]+$#', $aSearchTerm['word_token'])) {
+            if (preg_match('#^[0-9]+$#', $sToken)) {
                 $oSearch->iSearchRank += 2;
             }
-            if ($aSearchTerm['search_name_count'] + 1 < CONST_Max_Word_Frequency) {
+            if ($oSearchTerm->iSearchNameCount < CONST_Max_Word_Frequency) {
+                if (empty($this->aName)
+                    && CONST_Search_NameOnlySearchFrequencyThreshold
+                ) {
+                    $oSearch->bRareName =
+                        $oSearchTerm->iSearchNameCount
+                          < CONST_Search_NameOnlySearchFrequencyThreshold;
+                } else {
+                    $oSearch->bRareName = false;
+                }
                 $oSearch->aName[$iWordID] = $iWordID;
             } else {
                 $oSearch->aNameNonSearch[$iWordID] = $iWordID;
@@ -385,26 +404,22 @@ class SearchDescription
     /**
      * Query database for places that match this search.
      *
-     * @param object  $oDB                  Database connection to use.
-     * @param mixed[] $aWordFrequencyScores Number of times tokens appears
-     *                                      overall in a planet database.
-     * @param integer $iMinRank             Minimum address rank to restrict
-     *                                      search to.
-     * @param integer $iMaxRank             Maximum address rank to restrict
-     *                                      search to.
-     * @param integer $iLimit               Maximum number of results.
+     * @param object  $oDB      Nominatim::DB instance to use.
+     * @param integer $iMinRank Minimum address rank to restrict search to.
+     * @param integer $iMaxRank Maximum address rank to restrict search to.
+     * @param integer $iLimit   Maximum number of results.
      *
      * @return mixed[] An array with two fields: IDs contains the list of
      *                 matching place IDs and houseNumber the houseNumber
      *                 if appicable or -1 if not.
      */
-    public function query(&$oDB, &$aWordFrequencyScores, $iMinRank, $iMaxRank, $iLimit)
+    public function query(&$oDB, $iMinRank, $iMaxRank, $iLimit)
     {
         $aResults = array();
         $iHousenumber = -1;
 
         if ($this->sCountryCode
-            && !sizeof($this->aName)
+            && empty($this->aName)
             && !$this->iOperator
             && !$this->sClass
             && !$this->oContext->hasNearPoint()
@@ -413,7 +428,7 @@ class SearchDescription
             if (4 >= $iMinRank && 4 <= $iMaxRank) {
                 $aResults = $this->queryCountry($oDB);
             }
-        } elseif (!sizeof($this->aName) && !sizeof($this->aAddress)) {
+        } elseif (empty($this->aName) && empty($this->aAddress)) {
             // Neither name nor address? Then we must be
             // looking for a POI in a geographic area.
             if ($this->oContext->isBoundedSearch()) {
@@ -427,50 +442,47 @@ class SearchDescription
             // First search for places according to name and address.
             $aResults = $this->queryNamedPlace(
                 $oDB,
-                $aWordFrequencyScores,
                 $iMinRank,
                 $iMaxRank,
                 $iLimit
             );
 
             //now search for housenumber, if housenumber provided
-            if ($this->sHouseNumber && sizeof($aResults)) {
-                $aNamedPlaceIDs = $aResults;
-                $aResults = $this->queryHouseNumber($oDB, $aNamedPlaceIDs, $iLimit);
+            if ($this->sHouseNumber && !empty($aResults)) {
+                // Downgrade the rank of the street results, they are missing
+                // the housenumber.
+                foreach ($aResults as $oRes) {
+                    $oRes->iResultRank++;
+                }
 
-                if (!sizeof($aResults) && $this->looksLikeFullAddress()) {
-                    $aResults = $aNamedPlaceIDs;
+                $aHnResults = $this->queryHouseNumber($oDB, $aResults);
+
+                if (!empty($aHnResults)) {
+                    foreach ($aHnResults as $oRes) {
+                        $aResults[$oRes->iId] = $oRes;
+                    }
                 }
             }
 
             // finally get POIs if requested
-            if ($this->sClass && sizeof($aResults)) {
+            if ($this->sClass && !empty($aResults)) {
                 $aResults = $this->queryPoiByOperator($oDB, $aResults, $iLimit);
             }
         }
 
-        if (CONST_Debug) {
-            echo '<br><b>Place IDs:</b> ';
-            var_dump(array_keys($aResults));
-        }
+        Debug::printDebugTable('Place IDs', $aResults);
 
-        if (sizeof($aResults) && $this->sPostcode) {
+        if (!empty($aResults) && $this->sPostcode) {
             $sPlaceIds = Result::joinIdsByTable($aResults, Result::TABLE_PLACEX);
             if ($sPlaceIds) {
                 $sSQL = 'SELECT place_id FROM placex';
                 $sSQL .= ' WHERE place_id in ('.$sPlaceIds.')';
-                $sSQL .= " AND postcode = '".$this->sPostcode."'";
-                if (CONST_Debug) var_dump($sSQL);
-                $aFilteredPlaceIDs = chksql($oDB->getCol($sSQL));
+                $sSQL .= " AND postcode != '".$this->sPostcode."'";
+                Debug::printSQL($sSQL);
+                $aFilteredPlaceIDs = $oDB->getCol($sSQL);
                 if ($aFilteredPlaceIDs) {
-                    $aNewResults = array();
                     foreach ($aFilteredPlaceIDs as $iPlaceId) {
-                        $aNewResults[$iPlaceId] = $aResults[$iPlaceId];
-                    }
-                    $aResults = $aNewResults;
-                    if (CONST_Debug) {
-                        echo '<br><b>Place IDs after postcode filtering:</b> ';
-                        var_dump(array_keys($aResults));
+                        $aResults[$iPlaceId]->iResultRank++;
                     }
                 }
             }
@@ -490,10 +502,12 @@ class SearchDescription
         }
         $sSQL .= ' ORDER BY st_area(geometry) DESC LIMIT 1';
 
-        if (CONST_Debug) var_dump($sSQL);
+        Debug::printSQL($sSQL);
+
+        $iPlaceId = $oDB->getOne($sSQL);
 
         $aResults = array();
-        foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+        if ($iPlaceId) {
             $aResults[$iPlaceId] = new Result($iPlaceId);
         }
 
@@ -509,8 +523,7 @@ class SearchDescription
         $aDBResults = array();
         $sPoiTable = $this->poiTable();
 
-        $sSQL = 'SELECT count(*) FROM pg_tables WHERE tablename = \''.$sPoiTable."'";
-        if (chksql($oDB->getOne($sSQL))) {
+        if ($oDB->tableExists($sPoiTable)) {
             $sSQL = 'SELECT place_id FROM '.$sPoiTable.' ct';
             if ($this->oContext->sqlCountryList) {
                 $sSQL .= ' JOIN placex USING (place_id)';
@@ -530,14 +543,14 @@ class SearchDescription
             } elseif ($this->oContext->hasNearPoint()) {
                 $sSQL .= ' ORDER BY '.$this->oContext->distanceSQL('ct.centroid').' ASC';
             }
-            $sSQL .= " limit $iLimit";
-            if (CONST_Debug) var_dump($sSQL);
-            $aDBResults = chksql($oDB->getCol($sSQL));
+            $sSQL .= " LIMIT $iLimit";
+            Debug::printSQL($sSQL);
+            $aDBResults = $oDB->getCol($sSQL);
         }
 
         if ($this->oContext->hasNearPoint()) {
             $sSQL = 'SELECT place_id FROM placex WHERE ';
-            $sSQL .= 'class=\''.$this->sClass."' and type='".$this->sType."'";
+            $sSQL .= 'class = :class and type = :type';
             $sSQL .= ' AND '.$this->oContext->withinSQL('geometry');
             $sSQL .= ' AND linked_place_id is null';
             if ($this->oContext->sqlCountryList) {
@@ -545,8 +558,11 @@ class SearchDescription
             }
             $sSQL .= ' ORDER BY '.$this->oContext->distanceSQL('centroid').' ASC';
             $sSQL .= " LIMIT $iLimit";
-            if (CONST_Debug) var_dump($sSQL);
-            $aDBResults = chksql($oDB->getCol($sSQL));
+            Debug::printSQL($sSQL);
+            $aDBResults = $oDB->getCol(
+                $sSQL,
+                array(':class' => $this->sClass, ':type' => $this->sType)
+            );
         }
 
         $aResults = array();
@@ -561,36 +577,43 @@ class SearchDescription
     {
         $sSQL = 'SELECT p.place_id FROM location_postcode p ';
 
-        if (sizeof($this->aAddress)) {
+        if (!empty($this->aAddress)) {
             $sSQL .= ', search_name s ';
             $sSQL .= 'WHERE s.place_id = p.parent_place_id ';
             $sSQL .= 'AND array_cat(s.nameaddress_vector, s.name_vector)';
-            $sSQL .= '      @> '.getArraySQL($this->aAddress).' AND ';
+            $sSQL .= '      @> '.$oDB->getArraySQL($this->aAddress).' AND ';
         } else {
             $sSQL .= 'WHERE ';
         }
 
         $sSQL .= "p.postcode = '".reset($this->aName)."'";
         $sSQL .= $this->countryCodeSQL(' AND p.country_code');
+        if ($this->oContext->bViewboxBounded) {
+            $sSQL .= ' AND ST_Intersects('.$this->oContext->sqlViewboxSmall.', geometry)';
+        }
         $sSQL .= $this->oContext->excludeSQL(' AND p.place_id');
         $sSQL .= " LIMIT $iLimit";
 
-        if (CONST_Debug) var_dump($sSQL);
+        Debug::printSQL($sSQL);
 
         $aResults = array();
-        foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+        foreach ($oDB->getCol($sSQL) as $iPlaceId) {
             $aResults[$iPlaceId] = new Result($iPlaceId, Result::TABLE_POSTCODE);
         }
 
         return $aResults;
     }
 
-    private function queryNamedPlace(&$oDB, $aWordFrequencyScores, $iMinAddressRank, $iMaxAddressRank, $iLimit)
+    private function queryNamedPlace(&$oDB, $iMinAddressRank, $iMaxAddressRank, $iLimit)
     {
         $aTerms = array();
         $aOrder = array();
 
-        if ($this->sHouseNumber && sizeof($this->aAddress)) {
+        // Sort by existence of the requested house number but only if not
+        // too many results are expected for the street, i.e. if the result
+        // will be narrowed down by an address. Remeber that with ordering
+        // every single result has to be checked.
+        if ($this->sHouseNumber && (!empty($this->aAddress) || $this->sPostcode)) {
             $sHouseNumberRegex = '\\\\m'.$this->sHouseNumber.'\\\\M';
             $aOrder[] = ' (';
             $aOrder[0] .= 'EXISTS(';
@@ -616,19 +639,15 @@ class SearchDescription
             $aOrder[0] .= ') DESC';
         }
 
-        if (sizeof($this->aName)) {
-            $aTerms[] = 'name_vector @> '.getArraySQL($this->aName);
+        if (!empty($this->aName)) {
+            $aTerms[] = 'name_vector @> '.$oDB->getArraySQL($this->aName);
         }
-        if (sizeof($this->aAddress)) {
+        if (!empty($this->aAddress)) {
             // For infrequent name terms disable index usage for address
-            if (CONST_Search_NameOnlySearchFrequencyThreshold
-                && sizeof($this->aName) == 1
-                && $aWordFrequencyScores[$this->aName[reset($this->aName)]]
-                     < CONST_Search_NameOnlySearchFrequencyThreshold
-            ) {
-                $aTerms[] = 'array_cat(nameaddress_vector,ARRAY[]::integer[]) @> '.getArraySQL($this->aAddress);
+            if ($this->bRareName) {
+                $aTerms[] = 'array_cat(nameaddress_vector,ARRAY[]::integer[]) @> '.$oDB->getArraySQL($this->aAddress);
             } else {
-                $aTerms[] = 'nameaddress_vector @> '.getArraySQL($this->aAddress);
+                $aTerms[] = 'nameaddress_vector @> '.$oDB->getArraySQL($this->aAddress);
             }
         }
 
@@ -652,7 +671,7 @@ class SearchDescription
             $aTerms[] = $this->oContext->withinSQL('centroid');
             $aOrder[] = $this->oContext->distanceSQL('centroid');
         } elseif ($this->sPostcode) {
-            if (!sizeof($this->aAddress)) {
+            if (empty($this->aAddress)) {
                 $aTerms[] = "EXISTS(SELECT place_id FROM location_postcode p WHERE p.postcode = '".$this->sPostcode."' AND ST_DWithin(search_name.centroid, p.geometry, 0.1))";
             } else {
                 $aOrder[] = "(SELECT min(ST_Distance(search_name.centroid, p.geometry)) FROM location_postcode p WHERE p.postcode = '".$this->sPostcode."')";
@@ -675,15 +694,15 @@ class SearchDescription
         if ($this->sHouseNumber) {
             $sImportanceSQL = '- abs(26 - address_rank) + 3';
         } else {
-            $sImportanceSQL = '(CASE WHEN importance = 0 OR importance IS NULL THEN 0.75-(search_rank::float/40) ELSE importance END)';
+            $sImportanceSQL = '(CASE WHEN importance = 0 OR importance IS NULL THEN 0.75001-(search_rank::float/40) ELSE importance END)';
         }
         $sImportanceSQL .= $this->oContext->viewboxImportanceSQL('centroid');
         $aOrder[] = "$sImportanceSQL DESC";
 
-        if (sizeof($this->aFullNameAddress)) {
+        if (!empty($this->aFullNameAddress)) {
             $sExactMatchSQL = ' ( ';
             $sExactMatchSQL .= ' SELECT count(*) FROM ( ';
-            $sExactMatchSQL .= '  SELECT unnest('.getArraySQL($this->aFullNameAddress).')';
+            $sExactMatchSQL .= '  SELECT unnest('.$oDB->getArraySQL($this->aFullNameAddress).')';
             $sExactMatchSQL .= '    INTERSECT ';
             $sExactMatchSQL .= '  SELECT unnest(nameaddress_vector)';
             $sExactMatchSQL .= ' ) s';
@@ -694,24 +713,21 @@ class SearchDescription
         }
 
         if ($this->sHouseNumber || $this->sClass) {
-            $iLimit = 20;
+            $iLimit = 40;
         }
 
         $aResults = array();
 
-        if (sizeof($aTerms)) {
+        if (!empty($aTerms)) {
             $sSQL = 'SELECT place_id,'.$sExactMatchSQL;
             $sSQL .= ' FROM search_name';
             $sSQL .= ' WHERE '.join(' and ', $aTerms);
             $sSQL .= ' ORDER BY '.join(', ', $aOrder);
             $sSQL .= ' LIMIT '.$iLimit;
 
-            if (CONST_Debug) var_dump($sSQL);
+            Debug::printSQL($sSQL);
 
-            $aDBResults = chksql(
-                $oDB->getAll($sSQL),
-                'Could not get places for search terms.'
-            );
+            $aDBResults = $oDB->getAll($sSQL, null, 'Could not get places for search terms.');
 
             foreach ($aDBResults as $aResult) {
                 $oResult = new Result($aResult['place_id']);
@@ -723,7 +739,7 @@ class SearchDescription
         return $aResults;
     }
 
-    private function queryHouseNumber(&$oDB, $aRoadPlaceIDs, $iLimit)
+    private function queryHouseNumber(&$oDB, $aRoadPlaceIDs)
     {
         $aResults = array();
         $sPlaceIDs = Result::joinIdsByTable($aRoadPlaceIDs, Result::TABLE_PLACEX);
@@ -737,18 +753,17 @@ class SearchDescription
         $sSQL .= 'WHERE parent_place_id in ('.$sPlaceIDs.')';
         $sSQL .= "  AND transliteration(housenumber) ~* E'".$sHouseNumberRegex."'";
         $sSQL .= $this->oContext->excludeSQL(' AND place_id');
-        $sSQL .= " LIMIT $iLimit";
 
-        if (CONST_Debug) var_dump($sSQL);
+        Debug::printSQL($sSQL);
 
         // XXX should inherit the exactMatches from its parent
-        foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+        foreach ($oDB->getCol($sSQL) as $iPlaceId) {
             $aResults[$iPlaceId] = new Result($iPlaceId);
         }
 
         $bIsIntHouseNumber= (bool) preg_match('/[0-9]+/', $this->sHouseNumber);
         $iHousenumber = intval($this->sHouseNumber);
-        if ($bIsIntHouseNumber && !sizeof($aResults)) {
+        if ($bIsIntHouseNumber && empty($aResults)) {
             // if nothing found, search in the interpolation line table
             $sSQL = 'SELECT distinct place_id FROM location_property_osmline';
             $sSQL .= ' WHERE startnumber is not NULL';
@@ -765,11 +780,10 @@ class SearchDescription
             $sSQL .= $iHousenumber.'>=startnumber and ';
             $sSQL .= $iHousenumber.'<=endnumber';
             $sSQL .= $this->oContext->excludeSQL(' AND place_id');
-            $sSQL .= " limit $iLimit";
 
-            if (CONST_Debug) var_dump($sSQL);
+            Debug::printSQL($sSQL);
 
-            foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+            foreach ($oDB->getCol($sSQL) as $iPlaceId) {
                 $oResult = new Result($iPlaceId, Result::TABLE_OSMLINE);
                 $oResult->iHouseNumber = $iHousenumber;
                 $aResults[$iPlaceId] = $oResult;
@@ -777,22 +791,21 @@ class SearchDescription
         }
 
         // If nothing found try the aux fallback table
-        if (CONST_Use_Aux_Location_data && !sizeof($aResults)) {
+        if (CONST_Use_Aux_Location_data && empty($aResults)) {
             $sSQL = 'SELECT place_id FROM location_property_aux';
             $sSQL .= ' WHERE parent_place_id in ('.$sPlaceIDs.')';
             $sSQL .= " AND housenumber = '".$this->sHouseNumber."'";
             $sSQL .= $this->oContext->excludeSQL(' AND place_id');
-            $sSQL .= " limit $iLimit";
 
-            if (CONST_Debug) var_dump($sSQL);
+            Debug::printSQL($sSQL);
 
-            foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+            foreach ($oDB->getCol($sSQL) as $iPlaceId) {
                 $aResults[$iPlaceId] = new Result($iPlaceId, Result::TABLE_AUX);
             }
         }
 
         // If nothing found then search in Tiger data (location_property_tiger)
-        if (CONST_Use_US_Tiger_Data && $bIsIntHouseNumber && !sizeof($aResults)) {
+        if (CONST_Use_US_Tiger_Data && $bIsIntHouseNumber && empty($aResults)) {
             $sSQL = 'SELECT place_id FROM location_property_tiger';
             $sSQL .= ' WHERE parent_place_id in ('.$sPlaceIDs.') and (';
             if ($iHousenumber % 2 == 0) {
@@ -804,11 +817,10 @@ class SearchDescription
             $sSQL .= $iHousenumber.'>=startnumber and ';
             $sSQL .= $iHousenumber.'<=endnumber';
             $sSQL .= $this->oContext->excludeSQL(' AND place_id');
-            $sSQL .= " limit $iLimit";
 
-            if (CONST_Debug) var_dump($sSQL);
+            Debug::printSQL($sSQL);
 
-            foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+            foreach ($oDB->getCol($sSQL) as $iPlaceId) {
                 $oResult = new Result($iPlaceId, Result::TABLE_TIGER);
                 $oResult->iHouseNumber = $iHousenumber;
                 $aResults[$iPlaceId] = $oResult;
@@ -840,9 +852,9 @@ class SearchDescription
             $sSQL .= ' ORDER BY rank_search ASC ';
             $sSQL .= " LIMIT $iLimit";
 
-            if (CONST_Debug) var_dump($sSQL);
+            Debug::printSQL($sSQL);
 
-            foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+            foreach ($oDB->getCol($sSQL) as $iPlaceId) {
                 $aResults[$iPlaceId] = new Result($iPlaceId);
             }
         }
@@ -850,12 +862,11 @@ class SearchDescription
         // NEAR and IN are handled the same
         if ($this->iOperator == Operator::TYPE || $this->iOperator == Operator::NEAR) {
             $sClassTable = $this->poiTable();
-            $sSQL = "SELECT count(*) FROM pg_tables WHERE tablename = '$sClassTable'";
-            $bCacheTable = (bool) chksql($oDB->getOne($sSQL));
+            $bCacheTable = $oDB->tableExists($sClassTable);
 
             $sSQL = "SELECT min(rank_search) FROM placex WHERE place_id in ($sPlaceIDs)";
-            if (CONST_Debug) var_dump($sSQL);
-            $iMaxRank = (int)chksql($oDB->getOne($sSQL));
+            Debug::printSQL($sSQL);
+            $iMaxRank = (int) $oDB->getOne($sSQL);
 
             // For state / country level searches the normal radius search doesn't work very well
             $sPlaceGeom = false;
@@ -867,8 +878,8 @@ class SearchDescription
                 $sSQL .= "   AND ST_GeometryType(geometry) in ('ST_Polygon','ST_MultiPolygon')";
                 $sSQL .= ' ORDER BY rank_search ASC ';
                 $sSQL .= ' LIMIT 1';
-                if (CONST_Debug) var_dump($sSQL);
-                $sPlaceGeom = chksql($oDB->getOne($sSQL));
+                Debug::printSQL($sSQL);
+                $sPlaceGeom = $oDB->getOne($sSQL);
             }
 
             if ($sPlaceGeom) {
@@ -877,8 +888,8 @@ class SearchDescription
                 $iMaxRank += 5;
                 $sSQL = 'SELECT place_id FROM placex';
                 $sSQL .= " WHERE place_id in ($sPlaceIDs) and rank_search < $iMaxRank";
-                if (CONST_Debug) var_dump($sSQL);
-                $aPlaceIDs = chksql($oDB->getCol($sSQL));
+                Debug::printSQL($sSQL);
+                $aPlaceIDs = $oDB->getCol($sSQL);
                 $sPlaceIDs = join(',', $aPlaceIDs);
             }
 
@@ -922,9 +933,9 @@ class SearchDescription
                     }
                     $sSQL .= " limit $iLimit";
 
-                    if (CONST_Debug) var_dump($sSQL);
+                    Debug::printSQL($sSQL);
 
-                    foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+                    foreach ($oDB->getCol($sSQL) as $iPlaceId) {
                         $aResults[$iPlaceId] = new Result($iPlaceId);
                     }
                 } else {
@@ -954,9 +965,9 @@ class SearchDescription
                     }
                     $sSQL .= " limit $iLimit";
 
-                    if (CONST_Debug) var_dump($sSQL);
+                    Debug::printSQL($sSQL);
 
-                    foreach (chksql($oDB->getCol($sSQL)) as $iPlaceId) {
+                    foreach ($oDB->getCol($sSQL) as $iPlaceId) {
                         $aResults[$iPlaceId] = new Result($iPlaceId);
                     }
                 }
@@ -999,6 +1010,24 @@ class SearchDescription
     //////////// Debugging functions
 
 
+    public function debugInfo()
+    {
+        return array(
+                'Search rank' => $this->iSearchRank,
+                'Country code' => $this->sCountryCode,
+                'Name terms' => $this->aName,
+                'Name terms (stop words)' => $this->aNameNonSearch,
+                'Address terms' => $this->aAddress,
+                'Address terms (stop words)' => $this->aAddressNonSearch,
+                'Address terms (full words)' => $this->aFullNameAddress,
+                'Special search' => $this->iOperator,
+                'Class' => $this->sClass,
+                'Type' => $this->sType,
+                'House number' => $this->sHouseNumber,
+                'Postcode' => $this->sPostcode
+               );
+    }
+
     public function dumpAsHtmlTableRow(&$aWordIDs)
     {
         $kf = function ($k) use (&$aWordIDs) {

lib/Status.php

@@ -0,0 +1,59 @@
+<?php
+
+namespace Nominatim;
+
+use Exception;
+
+class Status
+{
+    protected $oDB;
+
+    public function __construct(&$oDB)
+    {
+        $this->oDB =& $oDB;
+    }
+
+    public function status()
+    {
+        if (!$this->oDB) {
+            throw new Exception('No database', 700);
+        }
+
+        try {
+            $this->oDB->connect();
+        } catch (\Nominatim\DatabaseError $e) {
+            throw new Exception('Database connection failed', 700);
+        }
+
+        $sStandardWord = $this->oDB->getOne("SELECT make_standard_name('a')");
+        if ($sStandardWord === false) {
+            throw new Exception('Module failed', 701);
+        }
+
+        if ($sStandardWord != 'a') {
+            throw new Exception('Module call failed', 702);
+        }
+
+        $sSQL = 'SELECT word_id, word_token, word, class, type, country_code, ';
+        $sSQL .= "operator, search_name_count FROM word WHERE word_token IN (' a')";
+        $iWordID = $this->oDB->getOne($sSQL);
+        if ($iWordID === false) {
+            throw new Exception('Query failed', 703);
+        }
+        if (!$iWordID) {
+            throw new Exception('No value', 704);
+        }
+    }
+
+    public function dataDate()
+    {
+        $sSQL = 'SELECT EXTRACT(EPOCH FROM lastimportdate) FROM import_status LIMIT 1';
+        $iDataDateEpoch = $this->oDB->getOne($sSQL);
+
+        if ($iDataDateEpoch === false) {
+            throw Exception('Data date query failed '.$iDataDateEpoch->getMessage(), 705);
+        }
+
+        return $iDataDateEpoch;
+    }
+}

lib/TokenCountry.php

@@ -0,0 +1,29 @@
+<?php
+
+namespace Nominatim\Token;
+
+/**
+ * A country token.
+ */
+class Country
+{
+    /// Database word id, if available.
+    public $iId;
+    /// Two-letter country code (lower-cased).
+    public $sCountryCode;
+
+    public function __construct($iId, $sCountryCode)
+    {
+        $this->iId = $iId;
+        $this->sCountryCode = $sCountryCode;
+    }
+
+    public function debugInfo()
+    {
+        return array(
+                'ID' => $this->iId,
+                'Type' => 'country',
+                'Info' => $this->sCountryCode
+               );
+    }
+}

lib/TokenHousenumber.php

@@ -0,0 +1,29 @@
+<?php
+
+namespace Nominatim\Token;
+
+/**
+ * A house number token.
+ */
+class HouseNumber
+{
+    /// Database word id, if available.
+    public $iId;
+    /// Normalized house number.
+    public $sToken;
+
+    public function __construct($iId, $sToken)
+    {
+        $this->iId = $iId;
+        $this->sToken = $sToken;
+    }
+
+    public function debugInfo()
+    {
+        return array(
+                'ID' => $this->iId,
+                'Type' => 'house number',
+                'Info' => array('nr' => $this->sToken)
+               );
+    }
+}

lib/TokenList.php

@@ -0,0 +1,188 @@
+<?php
+
+namespace Nominatim;
+
+require_once(CONST_BasePath.'/lib/TokenCountry.php');
+require_once(CONST_BasePath.'/lib/TokenHousenumber.php');
+require_once(CONST_BasePath.'/lib/TokenPostcode.php');
+require_once(CONST_BasePath.'/lib/TokenSpecialTerm.php');
+require_once(CONST_BasePath.'/lib/TokenWord.php');
+require_once(CONST_BasePath.'/lib/SpecialSearchOperator.php');
+
+/**
+ * Saves information about the tokens that appear in a search query.
+ *
+ * Tokens are sorted by their normalized form, the token word. There are different
+ * kinds of tokens, represented by different Token* classes. Note that
+ * tokens do not have a common base class. All tokens need to have a field
+ * with the word id that points to an entry in the `word` database table
+ * but otherwise the information saved about a token can be very different.
+ *
+ * There are two different kinds of token words: full words and partial terms.
+ *
+ * Full words start with a space. They represent a complete name of a place.
+ * All special tokens are normally full words.
+ *
+ * Partial terms have no space at the beginning. They may represent a part of
+ * a name of a place (e.g. in the name 'World Trade Center' a partial term
+ * would be 'Trade' or 'Trade Center'). They are only used in TokenWord.
+ */
+class TokenList
+{
+    // List of list of tokens indexed by their word_token.
+    private $aTokens = array();
+
+
+    /**
+     * Return total number of tokens.
+     *
+     * @return Integer
+     */
+    public function count()
+    {
+        return count($this->aTokens);
+    }
+
+    /**
+     * Check if there are tokens for the given token word.
+     *
+     * @param string $sWord Token word to look for.
+     *
+     * @return bool True if there is one or more token for the token word.
+     */
+    public function contains($sWord)
+    {
+        return isset($this->aTokens[$sWord]);
+    }
+
+    /**
+     * Get the list of tokens for the given token word.
+     *
+     * @param string $sWord Token word to look for.
+     *
+     * @return object[] Array of tokens for the given token word or an
+     *                  empty array if no tokens could be found.
+     */
+    public function get($sWord)
+    {
+        return isset($this->aTokens[$sWord]) ? $this->aTokens[$sWord] : array();
+    }
+
+    /**
+     * Add token information from the word table in the database.
+     *
+     * @param object   $oDB           Nominatim::DB instance.
+     * @param string[] $aTokens       List of tokens to look up in the database.
+     * @param string[] $aCountryCodes List of country restrictions.
+     * @param string   $sNormQuery    Normalized query string.
+     * @param object   $oNormalizer   Normalizer function to use on tokens.
+     *
+     * @return void
+     */
+    public function addTokensFromDB(&$oDB, &$aTokens, &$aCountryCodes, $sNormQuery, $oNormalizer)
+    {
+        // Check which tokens we have, get the ID numbers
+        $sSQL = 'SELECT word_id, word_token, word, class, type, country_code,';
+        $sSQL .= ' operator, coalesce(search_name_count, 0) as count';
+        $sSQL .= ' FROM word WHERE word_token in (';
+        $sSQL .= join(',', $oDB->getDBQuotedList($aTokens)).')';
+
+        Debug::printSQL($sSQL);
+
+        $aDBWords = $oDB->getAll($sSQL, null, 'Could not get word tokens.');
+
+        foreach ($aDBWords as $aWord) {
+            $oToken = null;
+            $iId = (int) $aWord['word_id'];
+
+            if ($aWord['class']) {
+                // Special terms need to appear in their normalized form.
+                if ($aWord['word']) {
+                    $sNormWord = $aWord['word'];
+                    if ($oNormalizer != null) {
+                        $sNormWord = $oNormalizer->transliterate($aWord['word']);
+                    }
+                    if (strpos($sNormQuery, $sNormWord) === false) {
+                        continue;
+                    }
+                }
+
+                if ($aWord['class'] == 'place' && $aWord['type'] == 'house') {
+                    $oToken = new Token\HouseNumber($iId, trim($aWord['word_token']));
+                } elseif ($aWord['class'] == 'place' && $aWord['type'] == 'postcode') {
+                    if ($aWord['word']
+                        && pg_escape_string($aWord['word']) == $aWord['word']
+                    ) {
+                        $oToken = new Token\Postcode(
+                            $iId,
+                            $aWord['word'],
+                            $aWord['country_code']
+                        );
+                    }
+                } else {
+                    // near and in operator the same at the moment
+                    $oToken = new Token\SpecialTerm(
+                        $iId,
+                        $aWord['class'],
+                        $aWord['type'],
+                        $aWord['operator'] ? Operator::NEAR : Operator::NONE
+                    );
+                }
+            } elseif ($aWord['country_code']) {
+                // Filter country tokens that do not match restricted countries.
+                if (!$aCountryCodes
+                    || in_array($aWord['country_code'], $aCountryCodes)
+                ) {
+                    $oToken = new Token\Country($iId, $aWord['country_code']);
+                }
+            } else {
+                $oToken = new Token\Word(
+                    $iId,
+                    $aWord['word_token'][0] != ' ',
+                    (int) $aWord['count']
+                );
+            }
+
+            if ($oToken) {
+                $this->addToken($aWord['word_token'], $oToken);
+            }
+        }
+    }
+
+    /**
+     * Add a new token for the given word.
+     *
+     * @param string $sWord  Word the token describes.
+     * @param object $oToken Token object to add.
+     *
+     * @return void
+     */
+    public function addToken($sWord, $oToken)
+    {
+        if (isset($this->aTokens[$sWord])) {
+            $this->aTokens[$sWord][] = $oToken;
+        } else {
+            $this->aTokens[$sWord] = array($oToken);
+        }
+    }
+
+    public function debugTokenByWordIdList()
+    {
+        $aWordsIDs = array();
+        foreach ($this->aTokens as $sToken => $aWords) {
+            foreach ($aWords as $aToken) {
+                if ($aToken->iId !== null) {
+                    $aWordsIDs[$aToken->iId] =
+                        '#'.$sToken.'('.$aToken->iId.')#';
+                }
+            }
+        }
+
+        return $aWordsIDs;
+    }
+
+    public function debugInfo()
+    {
+        return $this->aTokens;
+    }
+}

lib/TokenPostcode.php

@@ -0,0 +1,32 @@
+<?php
+
+namespace Nominatim\Token;
+
+/**
+ * A postcode token.
+ */
+class Postcode
+{
+    /// Database word id, if available.
+    public $iId;
+    /// Full nomralized postcode (upper cased).
+    public $sPostcode;
+    // Optional country code the postcode belongs to (currently unused).
+    public $sCountryCode;
+
+    public function __construct($iId, $sPostcode, $sCountryCode = '')
+    {
+        $this->iId = $iId;
+        $this->sPostcode = $sPostcode;
+        $this->sCountryCode = empty($sCountryCode) ? '' : $sCountryCode;
+    }
+
+    public function debugInfo()
+    {
+        return array(
+                'ID' => $this->iId,
+                'Type' => 'postcode',
+                'Info' => $this->sPostcode.'('.$this->sCountryCode.')'
+               );
+    }
+}

lib/TokenSpecialTerm.php

@@ -0,0 +1,41 @@
+<?php
+
+namespace Nominatim\Token;
+
+require_once(CONST_BasePath.'/lib/SpecialSearchOperator.php');
+
+/**
+ * A word token describing a place type.
+ */
+class SpecialTerm
+{
+    /// Database word id, if applicable.
+    public $iId;
+    /// Class (or OSM tag key) of the place to look for.
+    public $sClass;
+    /// Type (or OSM tag value) of the place to look for.
+    public $sType;
+    /// Relationship of the operator to the object (see Operator class).
+    public $iOperator;
+
+    public function __construct($iID, $sClass, $sType, $iOperator)
+    {
+        $this->iId = $iID;
+        $this->sClass = $sClass;
+        $this->sType = $sType;
+        $this->iOperator = $iOperator;
+    }
+
+    public function debugInfo()
+    {
+        return array(
+                'ID' => $this->iId,
+                'Type' => 'special term',
+                'Info' => array(
+                           'class' => $this->sClass,
+                           'type' => $this->sType,
+                           'operator' => \Nominatim\Operator::toString($this->iOperator)
+                          )
+               );
+    }
+}

lib/TokenWord.php

@@ -0,0 +1,35 @@
+<?php
+
+namespace Nominatim\Token;
+
+/**
+ * A standard word token.
+ */
+class Word
+{
+    /// Database word id, if applicable.
+    public $iId;
+    /// If true, the word may represent only part of a place name.
+    public $bPartial;
+    /// Number of appearances in the database.
+    public $iSearchNameCount;
+
+    public function __construct($iId, $bPartial, $iSearchNameCount)
+    {
+        $this->iId = $iId;
+        $this->bPartial = $bPartial;
+        $this->iSearchNameCount = $iSearchNameCount;
+    }
+
+    public function debugInfo()
+    {
+        return array(
+                'ID' => $this->iId,
+                'Type' => 'word',
+                'Info' => array(
+                           'partial' => $this->bPartial,
+                           'count' => $this->iSearchNameCount
+                          )
+               );
+    }
+}

lib/cmd.php

@@ -16,7 +16,7 @@ function getCmdOpt($aArg, $aSpec, &$aResult, $bExitOnError = false, $bExitOnUnkn
 
     $aResult = array();
     $bUnknown = false;
-    $iSize = sizeof($aArg);
+    $iSize = count($aArg);
     for ($i = 1; $i < $iSize; $i++) {
         if (isset($aQuick[$aArg[$i]])) {
             $aLine = $aQuick[$aArg[$i]];
@@ -120,21 +120,12 @@ function showUsage($aSpec, $bExit = false, $sError = false)
     exit;
 }
 
-function chksql($oSql, $sMsg = false)
-{
-    if (PEAR::isError($oSql)) {
-        fail($sMsg || $oSql->getMessage(), $oSql->userinfo);
-    }
-
-    return $oSql;
-}
-
 function info($sMsg)
 {
     echo date('Y-m-d H:i:s == ').$sMsg."\n";
 }
 
-$aWarnings = [];
+$aWarnings = array();
 
 
 function warn($sMsg)
@@ -155,9 +146,19 @@ function repeatWarnings()
 function runSQLScript($sScript, $bfatal = true, $bVerbose = false, $bIgnoreErrors = false)
 {
     // Convert database DSN to psql parameters
-    $aDSNInfo = DB::parseDSN(CONST_Database_DSN);
+    $aDSNInfo = \Nominatim\DB::parseDSN(CONST_Database_DSN);
     if (!isset($aDSNInfo['port']) || !$aDSNInfo['port']) $aDSNInfo['port'] = 5432;
     $sCMD = 'psql -p '.$aDSNInfo['port'].' -d '.$aDSNInfo['database'];
+    if (isset($aDSNInfo['hostspec']) && $aDSNInfo['hostspec']) {
+        $sCMD .= ' -h ' . $aDSNInfo['hostspec'];
+    }
+    if (isset($aDSNInfo['username']) && $aDSNInfo['username']) {
+        $sCMD .= ' -U ' . $aDSNInfo['username'];
+    }
+    $aProcEnv = null;
+    if (isset($aDSNInfo['password']) && $aDSNInfo['password']) {
+        $aProcEnv = array_merge(array('PGPASSWORD' => $aDSNInfo['password']), $_ENV);
+    }
     if (!$bVerbose) {
         $sCMD .= ' -q';
     }
@@ -170,15 +171,19 @@ function runSQLScript($sScript, $bfatal = true, $bVerbose = false, $bIgnoreError
                      2 => STDERR
                     );
     $ahPipes = null;
-    $hProcess = @proc_open($sCMD, $aDescriptors, $ahPipes);
+    $hProcess = @proc_open($sCMD, $aDescriptors, $ahPipes, null, $aProcEnv);
     if (!is_resource($hProcess)) {
         fail('unable to start pgsql');
     }
 
+    if (!$bVerbose) {
+        fwrite($ahPipes[0], 'set client_min_messages to WARNING;');
+    }
+
     while (strlen($sScript)) {
-        $written = fwrite($ahPipes[0], $sScript);
-        if ($written <= 0) break;
-        $sScript = substr($sScript, $written);
+        $iWritten = fwrite($ahPipes[0], $sScript);
+        if ($iWritten <= 0) break;
+        $sScript = substr($sScript, $iWritten);
     }
     fclose($ahPipes[0]);
     $iReturn = proc_close($hProcess);
@@ -186,3 +191,23 @@ function runSQLScript($sScript, $bfatal = true, $bVerbose = false, $bIgnoreError
         fail("pgsql returned with error code ($iReturn)");
     }
 }
+
+
+function runWithEnv($sCmd, $aEnv)
+{
+    $aFDs = array(
+             0 => array('pipe', 'r'),
+             1 => STDOUT,
+             2 => STDERR
+            );
+    $aPipes = null;
+    $hProc = @proc_open($sCmd, $aFDs, $aPipes, null, $aEnv);
+    if (!is_resource($hProc)) {
+        fail('unable to run command:' . $sCmd);
+    }
+
+    fclose($aPipes[0]); // no stdin
+
+    $iStat = proc_close($hProc);
+    return $iStat;
+}

lib/db.php

@@ -1,43 +0,0 @@
-<?php
-
-require_once('DB.php');
-
-
-function &getDB($bNew = false, $bPersistent = false)
-{
-    // Get the database object
-    $oDB = chksql(
-        DB::connect(CONST_Database_DSN.($bNew?'?new_link=true':''), $bPersistent),
-        'Failed to establish database connection'
-    );
-    $oDB->setFetchMode(DB_FETCHMODE_ASSOC);
-    $oDB->query("SET DateStyle TO 'sql,european'");
-    $oDB->query("SET client_encoding TO 'utf-8'");
-    $iMaxExecution = ini_get('max_execution_time') * 1000;
-    if ($iMaxExecution > 0) $oDB->query("SET statement_timeout TO $iMaxExecution");
-    return $oDB;
-}
-
-function getDBQuoted($s)
-{
-    return "'".pg_escape_string($s)."'";
-}
-
-function getArraySQL($a)
-{
-    return 'ARRAY['.join(',', $a).']';
-}
-
-function getPostgresVersion(&$oDB)
-{
-    $sVersionString = $oDB->getOne('select version()');
-    preg_match('#PostgreSQL ([0-9]+)[.]([0-9]+)[^0-9]#', $sVersionString, $aMatches);
-    return (float) ($aMatches[1].'.'.$aMatches[2]);
-}
-
-function getPostgisVersion(&$oDB)
-{
-    $sVersionString = $oDB->getOne('select postgis_full_version()');
-    preg_match('#POSTGIS="([0-9]+)[.]([0-9]+)[.]([0-9]+)( r([0-9]+))?"#', $sVersionString, $aMatches);
-    return (float) ($aMatches[1].'.'.$aMatches[2]);
-}

lib/init-cmd.php

@@ -2,6 +2,7 @@
 
 require_once('init.php');
 require_once('cmd.php');
+require_once('DebugNone.php');
 
 // handle http proxy when using file_get_contents
 if (CONST_HTTP_Proxy) {

lib/init-website.php

@@ -2,6 +2,7 @@
 
 require_once('init.php');
 require_once('ParameterParser.php');
+require_once(CONST_Debug ? 'DebugHtml.php' : 'DebugNone.php');
 
 /***************************************************************************
  *
@@ -9,79 +10,47 @@ require_once('ParameterParser.php');
  *
  */
 
-
-function chksql($oSql, $sMsg = 'Database request failed')
+function userError($sMsg)
 {
-    if (!PEAR::isError($oSql)) return $oSql;
+    throw new Exception($sMsg, 400);
+}
 
-    header('HTTP/1.0 500 Internal Server Error');
-    header('Content-type: text/html; charset=utf-8');
 
-    $sSqlError = $oSql->getMessage();
+function exception_handler_html($exception)
+{
+    http_response_code($exception->getCode());
+    header('Content-type: text/html; charset=UTF-8');
+    include(CONST_BasePath.'/lib/template/error-html.php');
+}
 
-    echo <<<INTERNALFAIL
-<html>
-  <head><title>Internal Server Error</title></head>
-  <body>
-    <h1>Internal Server Error</h1>
-    <p>Nominatim has encountered an internal error while accessing the database.
-       This may happen because the database is broken or because of a bug in
-       the software. If you think it is a bug, feel free to report
-       it over on <a href="https://github.com/openstreetmap/Nominatim/issues">
-       Github</a>. Please include the URL that caused the problem and the
-       complete error details below.</p>
-    <p><b>Message:</b> $sMsg</p>
-    <p><b>SQL Error:</b> $sSqlError</p>
-    <p><b>Details:</b> <pre>
-INTERNALFAIL;
+function exception_handler_json($exception)
+{
+    http_response_code($exception->getCode());
+    header('Content-type: application/json; charset=utf-8');
+    include(CONST_BasePath.'/lib/template/error-json.php');
+}
 
-    if (CONST_Debug) {
-        var_dump($oSql);
+function exception_handler_xml($exception)
+{
+    http_response_code($exception->getCode());
+    header('Content-type: text/xml; charset=utf-8');
+    echo '<?xml version="1.0" encoding="UTF-8" ?>'."\n";
+    include(CONST_BasePath.'/lib/template/error-xml.php');
+}
+
+
+function set_exception_handler_by_format($sFormat = 'html')
+{
+    if ($sFormat == 'html') {
+        set_exception_handler('exception_handler_html');
+    } elseif ($sFormat == 'xml') {
+        set_exception_handler('exception_handler_xml');
     } else {
-        echo "<pre>\n".$oSql->getUserInfo().'</pre>';
+        set_exception_handler('exception_handler_json');
     }
-
-    echo '</pre></p></body></html>';
-    exit;
-}
-
-function failInternalError($sError, $sSQL = false, $vDumpVar = false)
-{
-    header('HTTP/1.0 500 Internal Server Error');
-    header('Content-type: text/html; charset=utf-8');
-    echo '<html><body><h1>Internal Server Error</h1>';
-    echo '<p>Nominatim has encountered an internal error while processing your request. This is most likely because of a bug in the software.</p>';
-    echo '<p><b>Details:</b> '.$sError,'</p>';
-    echo '<p>Feel free to file an issue on <a href="https://github.com/openstreetmap/Nominatim/issues">Github</a>. ';
-    echo 'Please include the error message above and the URL you used.</p>';
-    if (CONST_Debug) {
-        echo '<hr><h2>Debugging Information</h2><br>';
-        if ($sSQL) {
-            echo '<h3>SQL query</h3><code>'.$sSQL.'</code>';
-        }
-        if ($vDumpVar) {
-            echo '<h3>Result</h3> <code>';
-            var_dump($vDumpVar);
-            echo '</code>';
-        }
-    }
-    echo "\n</body></html>\n";
-    exit;
-}
-
-
-function userError($sError)
-{
-    header('HTTP/1.0 400 Bad Request');
-    header('Content-type: text/html; charset=utf-8');
-    echo '<html><body><h1>Bad Request</h1>';
-    echo '<p>Nominatim has encountered an error with your request.</p>';
-    echo '<p><b>Details:</b> '.$sError.'</p>';
-    echo '<p>If you feel this error is incorrect feel file an issue on <a href="https://github.com/openstreetmap/Nominatim/issues">Github</a>. ';
-    echo 'Please include the error message above and the URL you used.</p>';
-    echo "\n</body></html>\n";
-    exit;
 }
+// set a default
+set_exception_handler_by_format();
 
 
 /***************************************************************************
@@ -95,6 +64,6 @@ if (CONST_NoAccessControl) {
         header('Access-Control-Allow-Headers: '.$_SERVER['HTTP_ACCESS_CONTROL_REQUEST_HEADERS']);
     }
 }
-if ($_SERVER['REQUEST_METHOD'] == 'OPTIONS') exit;
+if (isset($_SERVER['REQUEST_METHOD']) && $_SERVER['REQUEST_METHOD'] == 'OPTIONS') exit;
 
 if (CONST_Debug) header('Content-type: text/html; charset=utf-8');

lib/init.php

@@ -1,7 +1,7 @@
 <?php
 
 require_once(CONST_BasePath.'/lib/lib.php');
-require_once(CONST_BasePath.'/lib/db.php');
+require_once(CONST_BasePath.'/lib/DB.php');
 
 if (get_magic_quotes_gpc()) {
     echo "Please disable magic quotes in your php.ini configuration\n";

lib/lib.php

@@ -1,11 +1,10 @@
 <?php
 
-
 function fail($sError, $sUserError = false)
 {
     if (!$sUserError) $sUserError = $sError;
     error_log('ERROR: '.$sError);
-    echo $sUserError."\n";
+    var_dump($sUserError)."\n";
     exit(-1);
 }
 
@@ -14,7 +13,7 @@ function getProcessorCount()
 {
     $sCPU = file_get_contents('/proc/cpuinfo');
     preg_match_all('#processor\s+: [0-9]+#', $sCPU, $aMatches);
-    return sizeof($aMatches[0]);
+    return count($aMatches[0]);
 }
 
 
@@ -60,438 +59,31 @@ function byImportance($a, $b)
 }
 
 
-function getClassTypes()
-{
-    return array(
-            'boundary:administrative:1' => array('label' => 'Continent', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'boundary:administrative:2' => array('label' => 'Country', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'place:country' => array('label' => 'Country', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defzoom' => 6, 'defdiameter' => 15),
-            'boundary:administrative:3' => array('label' => 'State', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'boundary:administrative:4' => array('label' => 'State', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'place:state' => array('label' => 'State', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defzoom' => 8, 'defdiameter' => 5.12),
-            'boundary:administrative:5' => array('label' => 'State District', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'boundary:administrative:6' => array('label' => 'County', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'boundary:administrative:7' => array('label' => 'County', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'place:county' => array('label' => 'County', 'frequency' => 108, 'icon' => 'poi_boundary_administrative', 'defzoom' => 10, 'defdiameter' => 1.28),
-            'boundary:administrative:8' => array('label' => 'City', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'place:city' => array('label' => 'City', 'frequency' => 66, 'icon' => 'poi_place_city', 'defzoom' => 12, 'defdiameter' => 0.32),
-            'boundary:administrative:9' => array('label' => 'City District', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'boundary:administrative:10' => array('label' => 'Suburb', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'boundary:administrative:11' => array('label' => 'Neighbourhood', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'place:region' => array('label' => 'Region', 'frequency' => 0, 'icon' => 'poi_boundary_administrative', 'defzoom' => 8, 'defdiameter' => 0.04),
-            'place:island' => array('label' => 'Island', 'frequency' => 288, 'icon' => '', 'defzoom' => 11, 'defdiameter' => 0.64),
-            'boundary:administrative' => array('label' => 'Administrative', 'frequency' => 413, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'boundary:postal_code' => array('label' => 'Postcode', 'frequency' => 413, 'icon' => 'poi_boundary_administrative', 'defdiameter' => 0.32),
-            'place:town' => array('label' => 'Town', 'frequency' => 1497, 'icon' => 'poi_place_town', 'defzoom' => 14, 'defdiameter' => 0.08),
-            'place:village' => array('label' => 'Village', 'frequency' => 11230, 'icon' => 'poi_place_village', 'defzoom' => 15, 'defdiameter' => 0.04),
-            'place:hamlet' => array('label' => 'Hamlet', 'frequency' => 7075, 'icon' => 'poi_place_village', 'defzoom' => 15, 'defdiameter' => 0.04),
-            'place:suburb' => array('label' => 'Suburb', 'frequency' => 2528, 'icon' => 'poi_place_village', 'defdiameter' => 0.04),
-            'place:locality' => array('label' => 'Locality', 'frequency' => 4113, 'icon' => 'poi_place_village', 'defdiameter' => 0.02),
-            'landuse:farm' => array('label' => 'Farm', 'frequency' => 1201, 'icon' => '', 'defdiameter' => 0.02),
-            'place:farm' => array('label' => 'Farm', 'frequency' => 1162, 'icon' => '', 'defdiameter' => 0.02),
-
-            'highway:motorway_junction' => array('label' => 'Motorway Junction', 'frequency' => 1126, 'icon' => '', 'simplelabel' => 'Junction'),
-            'highway:motorway' => array('label' => 'Motorway', 'frequency' => 4627, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:trunk' => array('label' => 'Trunk', 'frequency' => 23084, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:primary' => array('label' => 'Primary', 'frequency' => 32138, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:secondary' => array('label' => 'Secondary', 'frequency' => 25807, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:tertiary' => array('label' => 'Tertiary', 'frequency' => 29829, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:residential' => array('label' => 'Residential', 'frequency' => 361498, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:unclassified' => array('label' => 'Unclassified', 'frequency' => 66441, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:living_street' => array('label' => 'Living Street', 'frequency' => 710, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:service' => array('label' => 'Service', 'frequency' => 9963, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:track' => array('label' => 'Track', 'frequency' => 2565, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:road' => array('label' => 'Road', 'frequency' => 591, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:byway' => array('label' => 'Byway', 'frequency' => 346, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:bridleway' => array('label' => 'Bridleway', 'frequency' => 1556, 'icon' => ''),
-            'highway:cycleway' => array('label' => 'Cycleway', 'frequency' => 2419, 'icon' => ''),
-            'highway:pedestrian' => array('label' => 'Pedestrian', 'frequency' => 2757, 'icon' => ''),
-            'highway:footway' => array('label' => 'Footway', 'frequency' => 15008, 'icon' => ''),
-            'highway:steps' => array('label' => 'Steps', 'frequency' => 444, 'icon' => '', 'simplelabel' => 'Footway'),
-            'highway:motorway_link' => array('label' => 'Motorway Link', 'frequency' => 795, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:trunk_link' => array('label' => 'Trunk Link', 'frequency' => 1258, 'icon' => '', 'simplelabel' => 'Road'),
-            'highway:primary_link' => array('label' => 'Primary Link', 'frequency' => 313, 'icon' => '', 'simplelabel' => 'Road'),
-
-            'landuse:industrial' => array('label' => 'Industrial', 'frequency' => 1062, 'icon' => ''),
-            'landuse:residential' => array('label' => 'Residential', 'frequency' => 886, 'icon' => ''),
-            'landuse:retail' => array('label' => 'Retail', 'frequency' => 754, 'icon' => ''),
-            'landuse:commercial' => array('label' => 'Commercial', 'frequency' => 657, 'icon' => ''),
-
-            'place:airport' => array('label' => 'Airport', 'frequency' => 36, 'icon' => 'transport_airport2', 'defdiameter' => 0.03),
-            'aeroway:aerodrome' => array('label' => 'Aerodrome', 'frequency' => 36, 'icon' => 'transport_airport2', 'defdiameter' => 0.03),
-            'aeroway' => array('label' => 'Aeroway', 'frequency' => 36, 'icon' => 'transport_airport2', 'defdiameter' => 0.03),
-            'railway:station' => array('label' => 'Station', 'frequency' => 3431, 'icon' => 'transport_train_station2', 'defdiameter' => 0.01),
-            'amenity:place_of_worship' => array('label' => 'Place Of Worship', 'frequency' => 9049, 'icon' => 'place_of_worship_unknown3'),
-            'amenity:pub' => array('label' => 'Pub', 'frequency' => 18969, 'icon' => 'food_pub'),
-            'amenity:bar' => array('label' => 'Bar', 'frequency' => 164, 'icon' => 'food_bar'),
-            'amenity:university' => array('label' => 'University', 'frequency' => 607, 'icon' => 'education_university'),
-            'tourism:museum' => array('label' => 'Museum', 'frequency' => 543, 'icon' => 'tourist_museum'),
-            'amenity:arts_centre' => array('label' => 'Arts Centre', 'frequency' => 136, 'icon' => 'tourist_art_gallery2'),
-            'tourism:zoo' => array('label' => 'Zoo', 'frequency' => 47, 'icon' => 'tourist_zoo'),
-            'tourism:theme_park' => array('label' => 'Theme Park', 'frequency' => 24, 'icon' => 'poi_point_of_interest'),
-            'tourism:attraction' => array('label' => 'Attraction', 'frequency' => 1463, 'icon' => 'poi_point_of_interest'),
-            'leisure:golf_course' => array('label' => 'Golf Course', 'frequency' => 712, 'icon' => 'sport_golf'),
-            'historic:castle' => array('label' => 'Castle', 'frequency' => 316, 'icon' => 'tourist_castle'),
-            'amenity:hospital' => array('label' => 'Hospital', 'frequency' => 879, 'icon' => 'health_hospital'),
-            'amenity:school' => array('label' => 'School', 'frequency' => 8192, 'icon' => 'education_school'),
-            'amenity:theatre' => array('label' => 'Theatre', 'frequency' => 371, 'icon' => 'tourist_theatre'),
-            'amenity:public_building' => array('label' => 'Public Building', 'frequency' => 985, 'icon' => ''),
-            'amenity:library' => array('label' => 'Library', 'frequency' => 794, 'icon' => 'amenity_library'),
-            'amenity:townhall' => array('label' => 'Townhall', 'frequency' => 242, 'icon' => ''),
-            'amenity:community_centre' => array('label' => 'Community Centre', 'frequency' => 157, 'icon' => ''),
-            'amenity:fire_station' => array('label' => 'Fire Station', 'frequency' => 221, 'icon' => 'amenity_firestation3'),
-            'amenity:police' => array('label' => 'Police', 'frequency' => 334, 'icon' => 'amenity_police2'),
-            'amenity:bank' => array('label' => 'Bank', 'frequency' => 1248, 'icon' => 'money_bank2'),
-            'amenity:post_office' => array('label' => 'Post Office', 'frequency' => 859, 'icon' => 'amenity_post_office'),
-            'leisure:park' => array('label' => 'Park', 'frequency' => 2378, 'icon' => ''),
-            'amenity:park' => array('label' => 'Park', 'frequency' => 53, 'icon' => ''),
-            'landuse:park' => array('label' => 'Park', 'frequency' => 50, 'icon' => ''),
-            'landuse:recreation_ground' => array('label' => 'Recreation Ground', 'frequency' => 517, 'icon' => ''),
-            'tourism:hotel' => array('label' => 'Hotel', 'frequency' => 2150, 'icon' => 'accommodation_hotel2'),
-            'tourism:motel' => array('label' => 'Motel', 'frequency' => 43, 'icon' => ''),
-            'amenity:cinema' => array('label' => 'Cinema', 'frequency' => 277, 'icon' => 'tourist_cinema'),
-            'tourism:artwork' => array('label' => 'Artwork', 'frequency' => 171, 'icon' => 'tourist_art_gallery2'),
-            'historic:archaeological_site' => array('label' => 'Archaeological Site', 'frequency' => 407, 'icon' => 'tourist_archaeological2'),
-            'amenity:doctors' => array('label' => 'Doctors', 'frequency' => 581, 'icon' => 'health_doctors'),
-            'leisure:sports_centre' => array('label' => 'Sports Centre', 'frequency' => 767, 'icon' => 'sport_leisure_centre'),
-            'leisure:swimming_pool' => array('label' => 'Swimming Pool', 'frequency' => 24, 'icon' => 'sport_swimming_outdoor'),
-            'shop:supermarket' => array('label' => 'Supermarket', 'frequency' => 2673, 'icon' => 'shopping_supermarket'),
-            'shop:convenience' => array('label' => 'Convenience', 'frequency' => 1469, 'icon' => 'shopping_convenience'),
-            'amenity:restaurant' => array('label' => 'Restaurant', 'frequency' => 3179, 'icon' => 'food_restaurant'),
-            'amenity:fast_food' => array('label' => 'Fast Food', 'frequency' => 2289, 'icon' => 'food_fastfood'),
-            'amenity:cafe' => array('label' => 'Cafe', 'frequency' => 1780, 'icon' => 'food_cafe'),
-            'tourism:guest_house' => array('label' => 'Guest House', 'frequency' => 223, 'icon' => 'accommodation_bed_and_breakfast'),
-            'amenity:pharmacy' => array('label' => 'Pharmacy', 'frequency' => 733, 'icon' => 'health_pharmacy_dispensing'),
-            'amenity:fuel' => array('label' => 'Fuel', 'frequency' => 1308, 'icon' => 'transport_fuel'),
-            'natural:peak' => array('label' => 'Peak', 'frequency' => 3212, 'icon' => 'poi_peak'),
-            'waterway:waterfall' => array('label' => 'Waterfall', 'frequency' => 24, 'icon' => ''),
-            'natural:wood' => array('label' => 'Wood', 'frequency' => 1845, 'icon' => 'landuse_coniferous_and_deciduous'),
-            'natural:water' => array('label' => 'Water', 'frequency' => 1790, 'icon' => ''),
-            'landuse:forest' => array('label' => 'Forest', 'frequency' => 467, 'icon' => ''),
-            'landuse:cemetery' => array('label' => 'Cemetery', 'frequency' => 463, 'icon' => ''),
-            'landuse:allotments' => array('label' => 'Allotments', 'frequency' => 408, 'icon' => ''),
-            'landuse:farmyard' => array('label' => 'Farmyard', 'frequency' => 397, 'icon' => ''),
-            'railway:rail' => array('label' => 'Rail', 'frequency' => 4894, 'icon' => ''),
-            'waterway:canal' => array('label' => 'Canal', 'frequency' => 1723, 'icon' => ''),
-            'waterway:river' => array('label' => 'River', 'frequency' => 4089, 'icon' => ''),
-            'waterway:stream' => array('label' => 'Stream', 'frequency' => 2684, 'icon' => ''),
-            'shop:bicycle' => array('label' => 'Bicycle', 'frequency' => 349, 'icon' => 'shopping_bicycle'),
-            'shop:clothes' => array('label' => 'Clothes', 'frequency' => 315, 'icon' => 'shopping_clothes'),
-            'shop:hairdresser' => array('label' => 'Hairdresser', 'frequency' => 312, 'icon' => 'shopping_hairdresser'),
-            'shop:doityourself' => array('label' => 'Doityourself', 'frequency' => 247, 'icon' => 'shopping_diy'),
-            'shop:estate_agent' => array('label' => 'Estate Agent', 'frequency' => 162, 'icon' => 'shopping_estateagent2'),
-            'shop:car' => array('label' => 'Car', 'frequency' => 159, 'icon' => 'shopping_car'),
-            'shop:garden_centre' => array('label' => 'Garden Centre', 'frequency' => 143, 'icon' => 'shopping_garden_centre'),
-            'shop:car_repair' => array('label' => 'Car Repair', 'frequency' => 141, 'icon' => 'shopping_car_repair'),
-            'shop:newsagent' => array('label' => 'Newsagent', 'frequency' => 132, 'icon' => ''),
-            'shop:bakery' => array('label' => 'Bakery', 'frequency' => 129, 'icon' => 'shopping_bakery'),
-            'shop:furniture' => array('label' => 'Furniture', 'frequency' => 124, 'icon' => ''),
-            'shop:butcher' => array('label' => 'Butcher', 'frequency' => 105, 'icon' => 'shopping_butcher'),
-            'shop:apparel' => array('label' => 'Apparel', 'frequency' => 98, 'icon' => 'shopping_clothes'),
-            'shop:electronics' => array('label' => 'Electronics', 'frequency' => 96, 'icon' => ''),
-            'shop:department_store' => array('label' => 'Department Store', 'frequency' => 86, 'icon' => ''),
-            'shop:books' => array('label' => 'Books', 'frequency' => 85, 'icon' => ''),
-            'shop:yes' => array('label' => 'Shop', 'frequency' => 68, 'icon' => ''),
-            'shop:outdoor' => array('label' => 'Outdoor', 'frequency' => 67, 'icon' => ''),
-            'shop:mall' => array('label' => 'Mall', 'frequency' => 63, 'icon' => ''),
-            'shop:florist' => array('label' => 'Florist', 'frequency' => 61, 'icon' => ''),
-            'shop:charity' => array('label' => 'Charity', 'frequency' => 60, 'icon' => ''),
-            'shop:hardware' => array('label' => 'Hardware', 'frequency' => 59, 'icon' => ''),
-            'shop:laundry' => array('label' => 'Laundry', 'frequency' => 51, 'icon' => 'shopping_laundrette'),
-            'shop:shoes' => array('label' => 'Shoes', 'frequency' => 49, 'icon' => ''),
-            'shop:beverages' => array('label' => 'Beverages', 'frequency' => 48, 'icon' => 'shopping_alcohol'),
-            'shop:dry_cleaning' => array('label' => 'Dry Cleaning', 'frequency' => 46, 'icon' => ''),
-            'shop:carpet' => array('label' => 'Carpet', 'frequency' => 45, 'icon' => ''),
-            'shop:computer' => array('label' => 'Computer', 'frequency' => 44, 'icon' => ''),
-            'shop:alcohol' => array('label' => 'Alcohol', 'frequency' => 44, 'icon' => 'shopping_alcohol'),
-            'shop:optician' => array('label' => 'Optician', 'frequency' => 55, 'icon' => 'health_opticians'),
-            'shop:chemist' => array('label' => 'Chemist', 'frequency' => 42, 'icon' => 'health_pharmacy'),
-            'shop:gallery' => array('label' => 'Gallery', 'frequency' => 38, 'icon' => 'tourist_art_gallery2'),
-            'shop:mobile_phone' => array('label' => 'Mobile Phone', 'frequency' => 37, 'icon' => ''),
-            'shop:sports' => array('label' => 'Sports', 'frequency' => 37, 'icon' => ''),
-            'shop:jewelry' => array('label' => 'Jewelry', 'frequency' => 32, 'icon' => 'shopping_jewelry'),
-            'shop:pet' => array('label' => 'Pet', 'frequency' => 29, 'icon' => ''),
-            'shop:beauty' => array('label' => 'Beauty', 'frequency' => 28, 'icon' => ''),
-            'shop:stationery' => array('label' => 'Stationery', 'frequency' => 25, 'icon' => ''),
-            'shop:shopping_centre' => array('label' => 'Shopping Centre', 'frequency' => 25, 'icon' => ''),
-            'shop:general' => array('label' => 'General', 'frequency' => 25, 'icon' => ''),
-            'shop:electrical' => array('label' => 'Electrical', 'frequency' => 25, 'icon' => ''),
-            'shop:toys' => array('label' => 'Toys', 'frequency' => 23, 'icon' => ''),
-            'shop:jeweller' => array('label' => 'Jeweller', 'frequency' => 23, 'icon' => ''),
-            'shop:betting' => array('label' => 'Betting', 'frequency' => 23, 'icon' => ''),
-            'shop:household' => array('label' => 'Household', 'frequency' => 21, 'icon' => ''),
-            'shop:travel_agency' => array('label' => 'Travel Agency', 'frequency' => 21, 'icon' => ''),
-            'shop:hifi' => array('label' => 'Hifi', 'frequency' => 21, 'icon' => ''),
-            'amenity:shop' => array('label' => 'Shop', 'frequency' => 61, 'icon' => ''),
-            'tourism:information' => array('label' => 'Information', 'frequency' => 224, 'icon' => 'amenity_information'),
-
-            'place:house' => array('label' => 'House', 'frequency' => 2086, 'icon' => '', 'defzoom' => 18),
-            'place:house_name' => array('label' => 'House', 'frequency' => 2086, 'icon' => '', 'defzoom' => 18),
-            'place:house_number' => array('label' => 'House Number', 'frequency' => 2086, 'icon' => '', 'defzoom' => 18),
-            'place:country_code' => array('label' => 'Country Code', 'frequency' => 2086, 'icon' => '', 'defzoom' => 18),
-
-            //
-
-            'leisure:pitch' => array('label' => 'Pitch', 'frequency' => 762, 'icon' => ''),
-            'highway:unsurfaced' => array('label' => 'Unsurfaced', 'frequency' => 492, 'icon' => ''),
-            'historic:ruins' => array('label' => 'Ruins', 'frequency' => 483, 'icon' => 'tourist_ruin'),
-            'amenity:college' => array('label' => 'College', 'frequency' => 473, 'icon' => 'education_school'),
-            'historic:monument' => array('label' => 'Monument', 'frequency' => 470, 'icon' => 'tourist_monument'),
-            'railway:subway' => array('label' => 'Subway', 'frequency' => 385, 'icon' => ''),
-            'historic:memorial' => array('label' => 'Memorial', 'frequency' => 382, 'icon' => 'tourist_monument'),
-            'leisure:nature_reserve' => array('label' => 'Nature Reserve', 'frequency' => 342, 'icon' => ''),
-            'leisure:common' => array('label' => 'Common', 'frequency' => 322, 'icon' => ''),
-            'waterway:lock_gate' => array('label' => 'Lock Gate', 'frequency' => 321, 'icon' => ''),
-            'natural:fell' => array('label' => 'Fell', 'frequency' => 308, 'icon' => ''),
-            'amenity:nightclub' => array('label' => 'Nightclub', 'frequency' => 292, 'icon' => ''),
-            'highway:path' => array('label' => 'Path', 'frequency' => 287, 'icon' => ''),
-            'leisure:garden' => array('label' => 'Garden', 'frequency' => 285, 'icon' => ''),
-            'landuse:reservoir' => array('label' => 'Reservoir', 'frequency' => 276, 'icon' => ''),
-            'leisure:playground' => array('label' => 'Playground', 'frequency' => 264, 'icon' => ''),
-            'leisure:stadium' => array('label' => 'Stadium', 'frequency' => 212, 'icon' => ''),
-            'historic:mine' => array('label' => 'Mine', 'frequency' => 193, 'icon' => 'poi_mine'),
-            'natural:cliff' => array('label' => 'Cliff', 'frequency' => 193, 'icon' => ''),
-            'tourism:caravan_site' => array('label' => 'Caravan Site', 'frequency' => 183, 'icon' => 'accommodation_caravan_park'),
-            'amenity:bus_station' => array('label' => 'Bus Station', 'frequency' => 181, 'icon' => 'transport_bus_station'),
-            'amenity:kindergarten' => array('label' => 'Kindergarten', 'frequency' => 179, 'icon' => ''),
-            'highway:construction' => array('label' => 'Construction', 'frequency' => 176, 'icon' => ''),
-            'amenity:atm' => array('label' => 'Atm', 'frequency' => 172, 'icon' => 'money_atm2'),
-            'amenity:emergency_phone' => array('label' => 'Emergency Phone', 'frequency' => 164, 'icon' => ''),
-            'waterway:lock' => array('label' => 'Lock', 'frequency' => 146, 'icon' => ''),
-            'waterway:riverbank' => array('label' => 'Riverbank', 'frequency' => 143, 'icon' => ''),
-            'natural:coastline' => array('label' => 'Coastline', 'frequency' => 142, 'icon' => ''),
-            'tourism:viewpoint' => array('label' => 'Viewpoint', 'frequency' => 140, 'icon' => 'tourist_view_point'),
-            'tourism:hostel' => array('label' => 'Hostel', 'frequency' => 140, 'icon' => ''),
-            'tourism:bed_and_breakfast' => array('label' => 'Bed And Breakfast', 'frequency' => 140, 'icon' => 'accommodation_bed_and_breakfast'),
-            'railway:halt' => array('label' => 'Halt', 'frequency' => 135, 'icon' => ''),
-            'railway:platform' => array('label' => 'Platform', 'frequency' => 134, 'icon' => ''),
-            'railway:tram' => array('label' => 'Tram', 'frequency' => 130, 'icon' => 'transport_tram_stop'),
-            'amenity:courthouse' => array('label' => 'Courthouse', 'frequency' => 129, 'icon' => 'amenity_court'),
-            'amenity:recycling' => array('label' => 'Recycling', 'frequency' => 126, 'icon' => 'amenity_recycling'),
-            'amenity:dentist' => array('label' => 'Dentist', 'frequency' => 124, 'icon' => 'health_dentist'),
-            'natural:beach' => array('label' => 'Beach', 'frequency' => 121, 'icon' => 'tourist_beach'),
-            'place:moor' => array('label' => 'Moor', 'frequency' => 118, 'icon' => ''),
-            'amenity:grave_yard' => array('label' => 'Grave Yard', 'frequency' => 110, 'icon' => ''),
-            'waterway:drain' => array('label' => 'Drain', 'frequency' => 108, 'icon' => ''),
-            'landuse:grass' => array('label' => 'Grass', 'frequency' => 106, 'icon' => ''),
-            'landuse:village_green' => array('label' => 'Village Green', 'frequency' => 106, 'icon' => ''),
-            'natural:bay' => array('label' => 'Bay', 'frequency' => 102, 'icon' => ''),
-            'railway:tram_stop' => array('label' => 'Tram Stop', 'frequency' => 101, 'icon' => 'transport_tram_stop'),
-            'leisure:marina' => array('label' => 'Marina', 'frequency' => 98, 'icon' => ''),
-            'highway:stile' => array('label' => 'Stile', 'frequency' => 97, 'icon' => ''),
-            'natural:moor' => array('label' => 'Moor', 'frequency' => 95, 'icon' => ''),
-            'railway:light_rail' => array('label' => 'Light Rail', 'frequency' => 91, 'icon' => ''),
-            'railway:narrow_gauge' => array('label' => 'Narrow Gauge', 'frequency' => 90, 'icon' => ''),
-            'natural:land' => array('label' => 'Land', 'frequency' => 86, 'icon' => ''),
-            'amenity:village_hall' => array('label' => 'Village Hall', 'frequency' => 82, 'icon' => ''),
-            'waterway:dock' => array('label' => 'Dock', 'frequency' => 80, 'icon' => ''),
-            'amenity:veterinary' => array('label' => 'Veterinary', 'frequency' => 79, 'icon' => ''),
-            'landuse:brownfield' => array('label' => 'Brownfield', 'frequency' => 77, 'icon' => ''),
-            'leisure:track' => array('label' => 'Track', 'frequency' => 76, 'icon' => ''),
-            'railway:historic_station' => array('label' => 'Historic Station', 'frequency' => 74, 'icon' => ''),
-            'landuse:construction' => array('label' => 'Construction', 'frequency' => 72, 'icon' => ''),
-            'amenity:prison' => array('label' => 'Prison', 'frequency' => 71, 'icon' => 'amenity_prison'),
-            'landuse:quarry' => array('label' => 'Quarry', 'frequency' => 71, 'icon' => ''),
-            'amenity:telephone' => array('label' => 'Telephone', 'frequency' => 70, 'icon' => ''),
-            'highway:traffic_signals' => array('label' => 'Traffic Signals', 'frequency' => 66, 'icon' => ''),
-            'natural:heath' => array('label' => 'Heath', 'frequency' => 62, 'icon' => ''),
-            'historic:house' => array('label' => 'House', 'frequency' => 61, 'icon' => ''),
-            'amenity:social_club' => array('label' => 'Social Club', 'frequency' => 61, 'icon' => ''),
-            'landuse:military' => array('label' => 'Military', 'frequency' => 61, 'icon' => ''),
-            'amenity:health_centre' => array('label' => 'Health Centre', 'frequency' => 59, 'icon' => ''),
-            'historic:building' => array('label' => 'Building', 'frequency' => 58, 'icon' => ''),
-            'amenity:clinic' => array('label' => 'Clinic', 'frequency' => 57, 'icon' => ''),
-            'highway:services' => array('label' => 'Services', 'frequency' => 56, 'icon' => ''),
-            'amenity:ferry_terminal' => array('label' => 'Ferry Terminal', 'frequency' => 55, 'icon' => ''),
-            'natural:marsh' => array('label' => 'Marsh', 'frequency' => 55, 'icon' => ''),
-            'natural:hill' => array('label' => 'Hill', 'frequency' => 54, 'icon' => ''),
-            'highway:raceway' => array('label' => 'Raceway', 'frequency' => 53, 'icon' => ''),
-            'amenity:taxi' => array('label' => 'Taxi', 'frequency' => 47, 'icon' => ''),
-            'amenity:take_away' => array('label' => 'Take Away', 'frequency' => 45, 'icon' => ''),
-            'amenity:car_rental' => array('label' => 'Car Rental', 'frequency' => 44, 'icon' => ''),
-            'place:islet' => array('label' => 'Islet', 'frequency' => 44, 'icon' => ''),
-            'amenity:nursery' => array('label' => 'Nursery', 'frequency' => 44, 'icon' => ''),
-            'amenity:nursing_home' => array('label' => 'Nursing Home', 'frequency' => 43, 'icon' => ''),
-            'amenity:toilets' => array('label' => 'Toilets', 'frequency' => 38, 'icon' => ''),
-            'amenity:hall' => array('label' => 'Hall', 'frequency' => 38, 'icon' => ''),
-            'waterway:boatyard' => array('label' => 'Boatyard', 'frequency' => 36, 'icon' => ''),
-            'highway:mini_roundabout' => array('label' => 'Mini Roundabout', 'frequency' => 35, 'icon' => ''),
-            'historic:manor' => array('label' => 'Manor', 'frequency' => 35, 'icon' => ''),
-            'tourism:chalet' => array('label' => 'Chalet', 'frequency' => 34, 'icon' => ''),
-            'amenity:bicycle_parking' => array('label' => 'Bicycle Parking', 'frequency' => 34, 'icon' => ''),
-            'amenity:hotel' => array('label' => 'Hotel', 'frequency' => 34, 'icon' => ''),
-            'waterway:weir' => array('label' => 'Weir', 'frequency' => 33, 'icon' => ''),
-            'natural:wetland' => array('label' => 'Wetland', 'frequency' => 33, 'icon' => ''),
-            'natural:cave_entrance' => array('label' => 'Cave Entrance', 'frequency' => 32, 'icon' => ''),
-            'amenity:crematorium' => array('label' => 'Crematorium', 'frequency' => 31, 'icon' => ''),
-            'tourism:picnic_site' => array('label' => 'Picnic Site', 'frequency' => 31, 'icon' => ''),
-            'landuse:wood' => array('label' => 'Wood', 'frequency' => 30, 'icon' => ''),
-            'landuse:basin' => array('label' => 'Basin', 'frequency' => 30, 'icon' => ''),
-            'natural:tree' => array('label' => 'Tree', 'frequency' => 30, 'icon' => ''),
-            'leisure:slipway' => array('label' => 'Slipway', 'frequency' => 29, 'icon' => ''),
-            'landuse:meadow' => array('label' => 'Meadow', 'frequency' => 29, 'icon' => ''),
-            'landuse:piste' => array('label' => 'Piste', 'frequency' => 28, 'icon' => ''),
-            'amenity:care_home' => array('label' => 'Care Home', 'frequency' => 28, 'icon' => ''),
-            'amenity:club' => array('label' => 'Club', 'frequency' => 28, 'icon' => ''),
-            'amenity:medical_centre' => array('label' => 'Medical Centre', 'frequency' => 27, 'icon' => ''),
-            'historic:roman_road' => array('label' => 'Roman Road', 'frequency' => 27, 'icon' => ''),
-            'historic:fort' => array('label' => 'Fort', 'frequency' => 26, 'icon' => ''),
-            'railway:subway_entrance' => array('label' => 'Subway Entrance', 'frequency' => 26, 'icon' => ''),
-            'historic:yes' => array('label' => 'Historic', 'frequency' => 25, 'icon' => ''),
-            'highway:gate' => array('label' => 'Gate', 'frequency' => 25, 'icon' => ''),
-            'leisure:fishing' => array('label' => 'Fishing', 'frequency' => 24, 'icon' => ''),
-            'historic:museum' => array('label' => 'Museum', 'frequency' => 24, 'icon' => ''),
-            'amenity:car_wash' => array('label' => 'Car Wash', 'frequency' => 24, 'icon' => ''),
-            'railway:level_crossing' => array('label' => 'Level Crossing', 'frequency' => 23, 'icon' => ''),
-            'leisure:bird_hide' => array('label' => 'Bird Hide', 'frequency' => 23, 'icon' => ''),
-            'natural:headland' => array('label' => 'Headland', 'frequency' => 21, 'icon' => ''),
-            'tourism:apartments' => array('label' => 'Apartments', 'frequency' => 21, 'icon' => ''),
-            'amenity:shopping' => array('label' => 'Shopping', 'frequency' => 21, 'icon' => ''),
-            'natural:scrub' => array('label' => 'Scrub', 'frequency' => 20, 'icon' => ''),
-            'natural:fen' => array('label' => 'Fen', 'frequency' => 20, 'icon' => ''),
-            'building:yes' => array('label' => 'Building', 'frequency' => 200, 'icon' => ''),
-            'mountain_pass:yes' => array('label' => 'Mountain Pass', 'frequency' => 200, 'icon' => ''),
-
-            'amenity:parking' => array('label' => 'Parking', 'frequency' => 3157, 'icon' => ''),
-            'highway:bus_stop' => array('label' => 'Bus Stop', 'frequency' => 35777, 'icon' => 'transport_bus_stop2'),
-            'place:postcode' => array('label' => 'Postcode', 'frequency' => 27267, 'icon' => ''),
-            'amenity:post_box' => array('label' => 'Post Box', 'frequency' => 9613, 'icon' => ''),
-
-            'place:houses' => array('label' => 'Houses', 'frequency' => 85, 'icon' => ''),
-            'railway:preserved' => array('label' => 'Preserved', 'frequency' => 227, 'icon' => ''),
-            'waterway:derelict_canal' => array('label' => 'Derelict Canal', 'frequency' => 21, 'icon' => ''),
-            'amenity:dead_pub' => array('label' => 'Dead Pub', 'frequency' => 20, 'icon' => ''),
-            'railway:disused_station' => array('label' => 'Disused Station', 'frequency' => 114, 'icon' => ''),
-            'railway:abandoned' => array('label' => 'Abandoned', 'frequency' => 641, 'icon' => ''),
-            'railway:disused' => array('label' => 'Disused', 'frequency' => 72, 'icon' => ''),
-           );
-}
-
-
-function getClassTypesWithImportance()
-{
-    $aOrders = getClassTypes();
-    $i = 1;
-    foreach ($aOrders as $sID => $a) {
-        $aOrders[$sID]['importance'] = $i++;
-    }
-    return $aOrders;
-}
-
-function getResultDiameter($aResult)
-{
-    $aClassType = getClassTypes();
-
-    $fDiameter = 0.0001;
-
-    if (isset($aResult['class'])
-        && isset($aResult['type'])
-        && isset($aResult['admin_level'])
-        && isset($aClassType[$aResult['class'].':'.$aResult['type'].':'.$aResult['admin_level']]['defdiameter'])
-        && $aClassType[$aResult['class'].':'.$aResult['type'].':'.$aResult['admin_level']]['defdiameter']
-    ) {
-        $fDiameter = $aClassType[$aResult['class'].':'.$aResult['type'].':'.$aResult['admin_level']]['defdiameter'];
-    } elseif (isset($aResult['class'])
-        && isset($aResult['type'])
-        && isset($aClassType[$aResult['class'].':'.$aResult['type']]['defdiameter'])
-        && $aClassType[$aResult['class'].':'.$aResult['type']]['defdiameter']
-    ) {
-        $fDiameter = $aClassType[$aResult['class'].':'.$aResult['type']]['defdiameter'];
-    }
-
-    return $fDiameter;
-}
-
-
 function javascript_renderData($xVal, $iOptions = 0)
 {
-    $iOptions |= JSON_UNESCAPED_UNICODE;
+    $sCallback = isset($_GET['json_callback']) ? $_GET['json_callback'] : '';
+    if ($sCallback && !preg_match('/^[$_\p{L}][$_\p{L}\p{Nd}.[\]]*$/u', $sCallback)) {
+        // Unset, we call javascript_renderData again during exception handling
+        unset($_GET['json_callback']);
+        throw new Exception('Invalid json_callback value', 400);
+    }
+
+    $iOptions |= JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES;
     if (isset($_GET['pretty']) && in_array(strtolower($_GET['pretty']), array('1', 'true'))) {
         $iOptions |= JSON_PRETTY_PRINT;
     }
 
     $jsonout = json_encode($xVal, $iOptions);
 
-    if (!isset($_GET['json_callback'])) {
+    if ($sCallback) {
+        header('Content-Type: application/javascript; charset=UTF-8');
+        echo $_GET['json_callback'].'('.$jsonout.')';
+    } else {
         header('Content-Type: application/json; charset=UTF-8');
         echo $jsonout;
-    } else {
-        if (preg_match('/^[$_\p{L}][$_\p{L}\p{Nd}.[\]]*$/u', $_GET['json_callback'])) {
-            header('Content-Type: application/javascript; charset=UTF-8');
-            echo $_GET['json_callback'].'('.$jsonout.')';
-        } else {
-            header('HTTP/1.0 400 Bad Request');
-        }
     }
 }
 
-
-function _debugDumpGroupedSearches($aData, $aTokens)
-{
-    $aWordsIDs = array();
-    if ($aTokens) {
-        foreach ($aTokens as $sToken => $aWords) {
-            if ($aWords) {
-                foreach ($aWords as $aToken) {
-                    $aWordsIDs[$aToken['word_id']] =
-                        '#'.$sToken.'('.$aToken['word_id'].')#';
-                }
-            }
-        }
-    }
-    echo '<table border="1">';
-    echo '<tr><th>rank</th><th>Name Tokens</th><th>Name Not</th>';
-    echo '<th>Address Tokens</th><th>Address Not</th><th>country</th><th>operator</th>';
-    echo '<th>class</th><th>type</th><th>postcode</th><th>housenumber</th></tr>';
-    foreach ($aData as $iRank => $aRankedSet) {
-        foreach ($aRankedSet as $aRow) {
-            $aRow->dumpAsHtmlTableRow($aWordsIDs);
-        }
-    }
-    echo '</table>';
-}
-
-
-function getAddressDetails(&$oDB, $sLanguagePrefArraySQL, $iPlaceID, $sCountryCode = false, $housenumber = -1, $bRaw = false)
-{
-    $sSQL = "select *,get_name_by_language(name,$sLanguagePrefArraySQL) as localname from get_addressdata($iPlaceID, $housenumber)";
-    if (!$bRaw) $sSQL .= " WHERE isaddress OR type = 'country_code'";
-    $sSQL .= ' order by rank_address desc,isaddress desc';
-
-    $aAddressLines = chksql($oDB->getAll($sSQL));
-    if ($bRaw) return $aAddressLines;
-    //echo "<pre>";
-    //var_dump($aAddressLines);
-    $aAddress = array();
-    $aFallback = array();
-    $aClassType = getClassTypes();
-    foreach ($aAddressLines as $aLine) {
-        $bFallback = false;
-        $aTypeLabel = false;
-        if (isset($aClassType[$aLine['class'].':'.$aLine['type'].':'.$aLine['admin_level']])) {
-            $aTypeLabel = $aClassType[$aLine['class'].':'.$aLine['type'].':'.$aLine['admin_level']];
-        } elseif (isset($aClassType[$aLine['class'].':'.$aLine['type']])) {
-            $aTypeLabel = $aClassType[$aLine['class'].':'.$aLine['type']];
-        } elseif (isset($aClassType['boundary:administrative:'.((int)($aLine['rank_address']/2))])) {
-            $aTypeLabel = $aClassType['boundary:administrative:'.((int)($aLine['rank_address']/2))];
-            $bFallback = true;
-        } else {
-            $aTypeLabel = array('simplelabel' => 'address'.$aLine['rank_address']);
-            $bFallback = true;
-        }
-        if ($aTypeLabel && ((isset($aLine['localname']) && $aLine['localname']) || (isset($aLine['housenumber']) && $aLine['housenumber']))) {
-            $sTypeLabel = strtolower(isset($aTypeLabel['simplelabel'])?$aTypeLabel['simplelabel']:$aTypeLabel['label']);
-            $sTypeLabel = str_replace(' ', '_', $sTypeLabel);
-            if (!isset($aAddress[$sTypeLabel]) || (isset($aFallback[$sTypeLabel]) && $aFallback[$sTypeLabel]) || $aLine['class'] == 'place') {
-                $aAddress[$sTypeLabel] = $aLine['localname']?$aLine['localname']:$aLine['housenumber'];
-            }
-            $aFallback[$sTypeLabel] = $bFallback;
-        }
-    }
-    return $aAddress;
-}
-
-
 function addQuotes($s)
 {
     return "'".$s."'";
@@ -530,11 +122,12 @@ function parseLatLon($sQuery)
         $sFound    = $aData[0];
         $fQueryLat = ($aData[1]=='N'?1:-1) * ($aData[2] + $aData[3]/60 + $aData[4]/3600);
         $fQueryLon = ($aData[5]=='E'?1:-1) * ($aData[6] + $aData[7]/60 + $aData[8]/3600);
-    } elseif (preg_match('/\\s*([0-9]+)[° ]+([0-9]+)[′\' ]+([0-9]+)[″" ]+([NS])[, ]+([0-9]+)[° ]+([0-9]+)[′\' ]+([0-9]+)[″" ]+([EW])\\s*/', $sQuery, $aData)) {
-        /*                     1            2              3             4          5            6              7             8
+    } elseif (preg_match('/\\s*([0-9]+)[° ]+([0-9]+)[′\' ]+([0-9]+[0-9.]*)[″" ]+([NS])[, ]+([0-9]+)[° ]+([0-9]+)[′\' ]+([0-9]+[0-9.]*)[″" ]+([EW])\\s*/', $sQuery, $aData)) {
+        /*                     1            2              3                    4          5            6              7                     8
          * degrees decimal seconds
          * 40 26 46 N 79 58 56 W
          * 40° 26′ 46″ N, 79° 58′ 56″ W
+         * 40° 26′ 46.78″ N, 79° 58′ 56.89″ W
          */
         $sFound    = $aData[0];
         $fQueryLat = ($aData[4]=='N'?1:-1) * ($aData[1] + $aData[2]/60 + $aData[3]/3600);
@@ -635,3 +228,25 @@ function closestHouseNumber($aRow)
 
     return max(min($aRow['endnumber'], $iHn), $aRow['startnumber']);
 }
+
+function getSearchRankLabel($iRank)
+{
+    if (!isset($iRank)) return 'unknown';
+    if ($iRank < 2) return 'continent';
+    if ($iRank < 4) return 'sea';
+    if ($iRank < 8) return 'country';
+    if ($iRank < 12) return 'state';
+    if ($iRank < 16) return 'county';
+    if ($iRank == 16) return 'city';
+    if ($iRank == 17) return 'town / island';
+    if ($iRank == 18) return 'village / hamlet';
+    if ($iRank == 20) return 'suburb';
+    if ($iRank == 21) return 'postcode area';
+    if ($iRank == 22) return 'croft / farm / locality / islet';
+    if ($iRank == 23) return 'postcode area';
+    if ($iRank == 25) return 'postcode point';
+    if ($iRank == 26) return 'street / major landmark';
+    if ($iRank == 27) return 'minory street / path';
+    if ($iRank == 28) return 'house / building';
+    return 'other: ' . $iRank;
+}

lib/log.php

@@ -36,8 +36,18 @@ function logStart(&$oDB, $sType = '', $sQuery = '', $aLanguageList = array())
             $sUserAgent = $_SERVER['HTTP_USER_AGENT'];
         else $sUserAgent = '';
         $sSQL = 'insert into new_query_log (type,starttime,query,ipaddress,useragent,language,format,searchterm)';
-        $sSQL .= ' values ('.getDBQuoted($sType).','.getDBQuoted($hLog[0]).','.getDBQuoted($hLog[2]);
-        $sSQL .= ','.getDBQuoted($hLog[1]).','.getDBQuoted($sUserAgent).','.getDBQuoted(join(',', $aLanguageList)).','.getDBQuoted($sOutputFormat).','.getDBQuoted($hLog[3]).')';
+        $sSQL .= ' values ('.
+        $sSQL .= join(',', $oDB->getDBQuotedList(array(
+            $sType,
+            $hLog[0],
+            $hLog[2],
+            $hLog[1],
+            $sUserAgent,
+            join(',', $aLanguageList),
+            $sOutputFormat,
+            $hLog[3]
+        )));
+        $sSQL .= ')';
         $oDB->query($sSQL);
     }
 
@@ -53,10 +63,10 @@ function logEnd(&$oDB, $hLog, $iNumResults)
         if (!$aEndTime[1]) $aEndTime[1] = '0';
         $sEndTime = date('Y-m-d H:i:s', $aEndTime[0]).'.'.$aEndTime[1];
 
-        $sSQL = 'update new_query_log set endtime = '.getDBQuoted($sEndTime).', results = '.$iNumResults;
-        $sSQL .= ' where starttime = '.getDBQuoted($hLog[0]);
-        $sSQL .= ' and ipaddress = '.getDBQuoted($hLog[1]);
-        $sSQL .= ' and query = '.getDBQuoted($hLog[2]);
+        $sSQL = 'update new_query_log set endtime = '.$oDB->getDBQuoted($sEndTime).', results = '.$iNumResults;
+        $sSQL .= ' where starttime = '.$oDB->getDBQuoted($hLog[0]);
+        $sSQL .= ' and ipaddress = '.$oDB->getDBQuoted($hLog[1]);
+        $sSQL .= ' and query = '.$oDB->getDBQuoted($hLog[2]);
         $oDB->query($sSQL);
     }
 

lib/output.php

@@ -39,3 +39,14 @@ function detailsLink($aFeature, $sTitle = false)
 
     return '<a href="details.php?place_id='.$aFeature['place_id'].'">'.($sTitle?$sTitle:$aFeature['place_id']).'</a>';
 }
+
+function detailsPermaLink($aFeature, $sRefText = false)
+{
+    $sOSMType = formatOSMType($aFeature['osm_type'], false);
+
+    if ($sOSMType) {
+        $sLabel = $sRefText ? $sRefText : $sOSMType.' '.$aFeature['osm_id'];
+        return '<a href="details.php?osmtype='.$aFeature['osm_type'].'&osmid='.$aFeature['osm_id'].'&class='.$aFeature['class'].'">'.$sLabel.'</a>';
+    }
+    return '';
+}

lib/setup/AddressLevelParser.php

@@ -0,0 +1,98 @@
+<?php
+
+namespace Nominatim\Setup;
+
+/**
+ * Parses an address level description.
+ */
+class AddressLevelParser
+{
+    private $aLevels;
+
+    public function __construct($sDescriptionFile)
+    {
+        $sJson = file_get_contents($sDescriptionFile);
+        $this->aLevels = json_decode($sJson, true);
+        if (!$this->aLevels) {
+            switch (json_last_error()) {
+                case JSON_ERROR_NONE:
+                    break;
+                case JSON_ERROR_DEPTH:
+                    fail('JSON error - Maximum stack depth exceeded');
+                    break;
+                case JSON_ERROR_STATE_MISMATCH:
+                    fail('JSON error - Underflow or the modes mismatch');
+                    break;
+                case JSON_ERROR_CTRL_CHAR:
+                    fail('JSON error - Unexpected control character found');
+                    break;
+                case JSON_ERROR_SYNTAX:
+                    fail('JSON error - Syntax error, malformed JSON');
+                    break;
+                case JSON_ERROR_UTF8:
+                    fail('JSON error - Malformed UTF-8 characters, possibly incorrectly encoded');
+                    break;
+                default:
+                    fail('JSON error - Unknown error');
+                    break;
+            }
+        }
+    }
+
+    /**
+     * Dump the description into a database table.
+     *
+     * @param object $oDB    Database conneciton to use.
+     * @param string $sTable Name of table to create.
+     *
+     * @return null
+     *
+     * A new table is created. Any previously existing table is dropped.
+     * The table has the following columns:
+     * country, class, type, rank_search, rank_address.
+     */
+    public function createTable($oDB, $sTable)
+    {
+        $oDB->exec('DROP TABLE IF EXISTS '.$sTable);
+        $sSql = 'CREATE TABLE '.$sTable;
+        $sSql .= '(country_code varchar(2), class TEXT, type TEXT,';
+        $sSql .= ' rank_search SMALLINT, rank_address SMALLINT)';
+        $oDB->exec($sSql);
+
+        $sSql = 'CREATE UNIQUE INDEX ON '.$sTable.' (country_code, class, type)';
+        $oDB->exec($sSql);
+
+        $sSql = 'INSERT INTO '.$sTable.' VALUES ';
+        foreach ($this->aLevels as $aLevel) {
+            $aCountries = array();
+            if (isset($aLevel['countries'])) {
+                foreach ($aLevel['countries'] as $sCountry) {
+                    $aCountries[$sCountry] = $oDB->getDBQuoted($sCountry);
+                }
+            } else {
+                $aCountries['NULL'] = 'NULL';
+            }
+            foreach ($aLevel['tags'] as $sKey => $aValues) {
+                foreach ($aValues as $sValue => $mRanks) {
+                    $aFields = array(
+                        $oDB->getDBQuoted($sKey),
+                        $sValue ? $oDB->getDBQuoted($sValue) : 'NULL'
+                    );
+                    if (is_array($mRanks)) {
+                        $aFields[] = (string) $mRanks[0];
+                        $aFields[] = (string) $mRanks[1];
+                    } else {
+                        $aFields[] = (string) $mRanks;
+                        $aFields[] = (string) $mRanks;
+                    }
+                    $sLine = ','.join(',', $aFields).'),';
+
+                    foreach ($aCountries as $sCountries) {
+                        $sSql .= '('.$sCountries.$sLine;
+                    }
+                }
+            }
+        }
+        $oDB->exec(rtrim($sSql, ','));
+    }
+}

lib/setup/SetupClass.php

@@ -0,0 +1,928 @@
+<?php
+
+namespace Nominatim\Setup;
+
+require_once(CONST_BasePath.'/lib/setup/AddressLevelParser.php');
+
+class SetupFunctions
+{
+    protected $iCacheMemory;
+    protected $iInstances;
+    protected $sModulePath;
+    protected $aDSNInfo;
+    protected $bVerbose;
+    protected $sIgnoreErrors;
+    protected $bEnableDiffUpdates;
+    protected $bEnableDebugStatements;
+    protected $bNoPartitions;
+    protected $oDB = null;
+
+    public function __construct(array $aCMDResult)
+    {
+        // by default, use all but one processor, but never more than 15.
+        $this->iInstances = isset($aCMDResult['threads'])
+            ? $aCMDResult['threads']
+            : (min(16, getProcessorCount()) - 1);
+
+        if ($this->iInstances < 1) {
+            $this->iInstances = 1;
+            warn('resetting threads to '.$this->iInstances);
+        }
+
+        // Assume we can steal all the cache memory in the box (unless told otherwise)
+        if (isset($aCMDResult['osm2pgsql-cache'])) {
+            $this->iCacheMemory = $aCMDResult['osm2pgsql-cache'];
+        } else {
+            $this->iCacheMemory = getCacheMemoryMB();
+        }
+
+        $this->sModulePath = CONST_Database_Module_Path;
+        info('module path: ' . $this->sModulePath);
+
+        // parse database string
+        $this->aDSNInfo = \Nominatim\DB::parseDSN(CONST_Database_DSN);
+        if (!isset($this->aDSNInfo['port'])) {
+            $this->aDSNInfo['port'] = 5432;
+        }
+
+        // setting member variables based on command line options stored in $aCMDResult
+        $this->bVerbose = $aCMDResult['verbose'];
+
+        //setting default values which are not set by the update.php array
+        if (isset($aCMDResult['ignore-errors'])) {
+            $this->sIgnoreErrors = $aCMDResult['ignore-errors'];
+        } else {
+            $this->sIgnoreErrors = false;
+        }
+        if (isset($aCMDResult['enable-debug-statements'])) {
+            $this->bEnableDebugStatements = $aCMDResult['enable-debug-statements'];
+        } else {
+            $this->bEnableDebugStatements = false;
+        }
+        if (isset($aCMDResult['no-partitions'])) {
+            $this->bNoPartitions = $aCMDResult['no-partitions'];
+        } else {
+            $this->bNoPartitions = false;
+        }
+        if (isset($aCMDResult['enable-diff-updates'])) {
+            $this->bEnableDiffUpdates = $aCMDResult['enable-diff-updates'];
+        } else {
+            $this->bEnableDiffUpdates = false;
+        }
+    }
+
+    public function createDB()
+    {
+        info('Create DB');
+        $oDB = new \Nominatim\DB;
+
+        if ($oDB->databaseExists()) {
+            fail('database already exists ('.CONST_Database_DSN.')');
+        }
+
+        $sCreateDBCmd = 'createdb -E UTF-8 -p '.$this->aDSNInfo['port'].' '.$this->aDSNInfo['database'];
+        if (isset($this->aDSNInfo['username'])) {
+            $sCreateDBCmd .= ' -U '.$this->aDSNInfo['username'];
+        }
+
+        if (isset($this->aDSNInfo['hostspec'])) {
+            $sCreateDBCmd .= ' -h '.$this->aDSNInfo['hostspec'];
+        }
+
+        $result = $this->runWithPgEnv($sCreateDBCmd);
+        if ($result != 0) fail('Error executing external command: '.$sCreateDBCmd);
+    }
+
+    public function connect()
+    {
+        $this->oDB = new \Nominatim\DB();
+        $this->oDB->connect();
+    }
+
+    public function setupDB()
+    {
+        info('Setup DB');
+
+        $fPostgresVersion = $this->oDB->getPostgresVersion();
+        echo 'Postgres version found: '.$fPostgresVersion."\n";
+
+        if ($fPostgresVersion < 9.01) {
+            fail('Minimum supported version of Postgresql is 9.1.');
+        }
+
+        $this->pgsqlRunScript('CREATE EXTENSION IF NOT EXISTS hstore');
+        $this->pgsqlRunScript('CREATE EXTENSION IF NOT EXISTS postgis');
+
+        // For extratags and namedetails the hstore_to_json converter is
+        // needed which is only available from Postgresql 9.3+. For older
+        // versions add a dummy function that returns nothing.
+        $iNumFunc = $this->oDB->getOne("select count(*) from pg_proc where proname = 'hstore_to_json'");
+
+        if ($iNumFunc == 0) {
+            $this->pgsqlRunScript("create function hstore_to_json(dummy hstore) returns text AS 'select null::text' language sql immutable");
+            warn('Postgresql is too old. extratags and namedetails API not available.');
+        }
+
+
+        $fPostgisVersion = $this->oDB->getPostgisVersion();
+        echo 'Postgis version found: '.$fPostgisVersion."\n";
+
+        if ($fPostgisVersion < 2.1) {
+            // Functions were renamed in 2.1 and throw an annoying deprecation warning
+            $this->pgsqlRunScript('ALTER FUNCTION st_line_interpolate_point(geometry, double precision) RENAME TO ST_LineInterpolatePoint');
+            $this->pgsqlRunScript('ALTER FUNCTION ST_Line_Locate_Point(geometry, geometry) RENAME TO ST_LineLocatePoint');
+        }
+        if ($fPostgisVersion < 2.2) {
+            $this->pgsqlRunScript('ALTER FUNCTION ST_Distance_Spheroid(geometry, geometry, spheroid) RENAME TO ST_DistanceSpheroid');
+        }
+
+        $i = $this->oDB->getOne("select count(*) from pg_user where usename = '".CONST_Database_Web_User."'");
+        if ($i == 0) {
+            echo "\nERROR: Web user '".CONST_Database_Web_User."' does not exist. Create it with:\n";
+            echo "\n          createuser ".CONST_Database_Web_User."\n\n";
+            exit(1);
+        }
+
+        // Try accessing the C module, so we know early if something is wrong
+        checkModulePresence(); // raises exception on failure
+
+        if (!file_exists(CONST_ExtraDataPath.'/country_osm_grid.sql.gz')) {
+            echo 'Error: you need to download the country_osm_grid first:';
+            echo "\n    wget -O ".CONST_ExtraDataPath."/country_osm_grid.sql.gz https://www.nominatim.org/data/country_grid.sql.gz\n";
+            exit(1);
+        }
+        $this->pgsqlRunScriptFile(CONST_BasePath.'/data/country_name.sql');
+        $this->pgsqlRunScriptFile(CONST_BasePath.'/data/country_osm_grid.sql.gz');
+        $this->pgsqlRunScriptFile(CONST_BasePath.'/data/gb_postcode_table.sql');
+
+        $sPostcodeFilename = CONST_BasePath.'/data/gb_postcode_data.sql.gz';
+        if (file_exists($sPostcodeFilename)) {
+            $this->pgsqlRunScriptFile($sPostcodeFilename);
+        } else {
+            warn('optional external UK postcode table file ('.$sPostcodeFilename.') not found. Skipping.');
+        }
+
+        if (CONST_Use_Extra_US_Postcodes) {
+            $this->pgsqlRunScriptFile(CONST_BasePath.'/data/us_postcode.sql');
+        }
+
+        if ($this->bNoPartitions) {
+            $this->pgsqlRunScript('update country_name set partition = 0');
+        }
+
+        // the following will be needed by createFunctions later but
+        // is only defined in the subsequently called createTables
+        // Create dummies here that will be overwritten by the proper
+        // versions in create-tables.
+        $this->pgsqlRunScript('CREATE TABLE IF NOT EXISTS place_boundingbox ()');
+        $this->pgsqlRunScript('CREATE TYPE wikipedia_article_match AS ()', false);
+    }
+
+    public function importData($sOSMFile)
+    {
+        info('Import data');
+
+        $osm2pgsql = CONST_Osm2pgsql_Binary;
+        if (!file_exists($osm2pgsql)) {
+            echo "Check CONST_Osm2pgsql_Binary in your local settings file.\n";
+            echo "Normally you should not need to set this manually.\n";
+            fail("osm2pgsql not found in '$osm2pgsql'");
+        }
+
+        $osm2pgsql .= ' -S '.CONST_Import_Style;
+
+        if (!is_null(CONST_Osm2pgsql_Flatnode_File) && CONST_Osm2pgsql_Flatnode_File) {
+            $osm2pgsql .= ' --flat-nodes '.CONST_Osm2pgsql_Flatnode_File;
+        }
+
+        if (CONST_Tablespace_Osm2pgsql_Data)
+            $osm2pgsql .= ' --tablespace-slim-data '.CONST_Tablespace_Osm2pgsql_Data;
+        if (CONST_Tablespace_Osm2pgsql_Index)
+            $osm2pgsql .= ' --tablespace-slim-index '.CONST_Tablespace_Osm2pgsql_Index;
+        if (CONST_Tablespace_Place_Data)
+            $osm2pgsql .= ' --tablespace-main-data '.CONST_Tablespace_Place_Data;
+        if (CONST_Tablespace_Place_Index)
+            $osm2pgsql .= ' --tablespace-main-index '.CONST_Tablespace_Place_Index;
+        $osm2pgsql .= ' -lsc -O gazetteer --hstore --number-processes 1';
+        $osm2pgsql .= ' -C '.$this->iCacheMemory;
+        $osm2pgsql .= ' -P '.$this->aDSNInfo['port'];
+        if (isset($this->aDSNInfo['username'])) {
+            $osm2pgsql .= ' -U '.$this->aDSNInfo['username'];
+        }
+        if (isset($this->aDSNInfo['hostspec'])) {
+            $osm2pgsql .= ' -H '.$this->aDSNInfo['hostspec'];
+        }
+        $osm2pgsql .= ' -d '.$this->aDSNInfo['database'].' '.$sOSMFile;
+
+        $this->runWithPgEnv($osm2pgsql);
+
+        if (!$this->sIgnoreErrors && !$this->oDB->getRow('select * from place limit 1')) {
+            fail('No Data');
+        }
+    }
+
+    public function createFunctions()
+    {
+        info('Create Functions');
+
+        // Try accessing the C module, so we know early if something is wrong
+        checkModulePresence(); // raises exception on failure
+
+        $this->createSqlFunctions();
+    }
+
+    public function createTables($bReverseOnly = false)
+    {
+        info('Create Tables');
+
+        $sTemplate = file_get_contents(CONST_BasePath.'/sql/tables.sql');
+        $sTemplate = str_replace('{www-user}', CONST_Database_Web_User, $sTemplate);
+        $sTemplate = $this->replaceTablespace(
+            '{ts:address-data}',
+            CONST_Tablespace_Address_Data,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:address-index}',
+            CONST_Tablespace_Address_Index,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:search-data}',
+            CONST_Tablespace_Search_Data,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:search-index}',
+            CONST_Tablespace_Search_Index,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-data}',
+            CONST_Tablespace_Aux_Data,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-index}',
+            CONST_Tablespace_Aux_Index,
+            $sTemplate
+        );
+
+        $this->pgsqlRunScript($sTemplate, false);
+
+        if ($bReverseOnly) {
+            $this->pgExec('DROP TABLE search_name');
+        }
+
+        $oAlParser = new AddressLevelParser(CONST_Address_Level_Config);
+        $oAlParser->createTable($this->oDB, 'address_levels');
+    }
+
+    public function createPartitionTables()
+    {
+        info('Create Partition Tables');
+
+        $sTemplate = file_get_contents(CONST_BasePath.'/sql/partition-tables.src.sql');
+        $sTemplate = $this->replaceTablespace(
+            '{ts:address-data}',
+            CONST_Tablespace_Address_Data,
+            $sTemplate
+        );
+
+        $sTemplate = $this->replaceTablespace(
+            '{ts:address-index}',
+            CONST_Tablespace_Address_Index,
+            $sTemplate
+        );
+
+        $sTemplate = $this->replaceTablespace(
+            '{ts:search-data}',
+            CONST_Tablespace_Search_Data,
+            $sTemplate
+        );
+
+        $sTemplate = $this->replaceTablespace(
+            '{ts:search-index}',
+            CONST_Tablespace_Search_Index,
+            $sTemplate
+        );
+
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-data}',
+            CONST_Tablespace_Aux_Data,
+            $sTemplate
+        );
+
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-index}',
+            CONST_Tablespace_Aux_Index,
+            $sTemplate
+        );
+
+        $this->pgsqlRunPartitionScript($sTemplate);
+    }
+
+    public function createPartitionFunctions()
+    {
+        info('Create Partition Functions');
+
+        $sTemplate = file_get_contents(CONST_BasePath.'/sql/partition-functions.src.sql');
+        $this->pgsqlRunPartitionScript($sTemplate);
+    }
+
+    public function importWikipediaArticles()
+    {
+        $sWikiArticlesFile = CONST_Wikipedia_Data_Path.'/wikipedia_article.sql.bin';
+        $sWikiRedirectsFile = CONST_Wikipedia_Data_Path.'/wikipedia_redirect.sql.bin';
+        if (file_exists($sWikiArticlesFile)) {
+            info('Importing wikipedia articles');
+            $this->pgsqlRunDropAndRestore($sWikiArticlesFile);
+        } else {
+            warn('wikipedia article dump file not found - places will have default importance');
+        }
+        if (file_exists($sWikiRedirectsFile)) {
+            info('Importing wikipedia redirects');
+            $this->pgsqlRunDropAndRestore($sWikiRedirectsFile);
+        } else {
+            warn('wikipedia redirect dump file not found - some place importance values may be missing');
+        }
+    }
+
+    public function loadData($bDisableTokenPrecalc)
+    {
+        info('Drop old Data');
+
+        $this->pgExec('TRUNCATE word');
+        echo '.';
+        $this->pgExec('TRUNCATE placex');
+        echo '.';
+        $this->pgExec('TRUNCATE location_property_osmline');
+        echo '.';
+        $this->pgExec('TRUNCATE place_addressline');
+        echo '.';
+        $this->pgExec('TRUNCATE place_boundingbox');
+        echo '.';
+        $this->pgExec('TRUNCATE location_area');
+        echo '.';
+        if (!$this->dbReverseOnly()) {
+            $this->pgExec('TRUNCATE search_name');
+            echo '.';
+        }
+        $this->pgExec('TRUNCATE search_name_blank');
+        echo '.';
+        $this->pgExec('DROP SEQUENCE seq_place');
+        echo '.';
+        $this->pgExec('CREATE SEQUENCE seq_place start 100000');
+        echo '.';
+
+        $sSQL = 'select distinct partition from country_name';
+        $aPartitions = $this->oDB->getCol($sSQL);
+
+        if (!$this->bNoPartitions) $aPartitions[] = 0;
+        foreach ($aPartitions as $sPartition) {
+            $this->pgExec('TRUNCATE location_road_'.$sPartition);
+            echo '.';
+        }
+
+        // used by getorcreate_word_id to ignore frequent partial words
+        $sSQL = 'CREATE OR REPLACE FUNCTION get_maxwordfreq() RETURNS integer AS ';
+        $sSQL .= '$$ SELECT '.CONST_Max_Word_Frequency.' as maxwordfreq; $$ LANGUAGE SQL IMMUTABLE';
+        $this->pgExec($sSQL);
+        echo ".\n";
+
+        // pre-create the word list
+        if (!$bDisableTokenPrecalc) {
+            info('Loading word list');
+            $this->pgsqlRunScriptFile(CONST_BasePath.'/data/words.sql');
+        }
+
+        info('Load Data');
+        $sColumns = 'osm_type, osm_id, class, type, name, admin_level, address, extratags, geometry';
+
+        $aDBInstances = array();
+        $iLoadThreads = max(1, $this->iInstances - 1);
+        for ($i = 0; $i < $iLoadThreads; $i++) {
+            // https://secure.php.net/manual/en/function.pg-connect.php
+            $DSN = CONST_Database_DSN;
+            $DSN = preg_replace('/^pgsql:/', '', $DSN);
+            $DSN = preg_replace('/;/', ' ', $DSN);
+            $aDBInstances[$i] = pg_connect($DSN, PGSQL_CONNECT_FORCE_NEW);
+            pg_ping($aDBInstances[$i]);
+        }
+
+        for ($i = 0; $i < $iLoadThreads; $i++) {
+            $sSQL = "INSERT INTO placex ($sColumns) SELECT $sColumns FROM place WHERE osm_id % $iLoadThreads = $i";
+            $sSQL .= " and not (class='place' and type='houses' and osm_type='W'";
+            $sSQL .= "          and ST_GeometryType(geometry) = 'ST_LineString')";
+            $sSQL .= ' and ST_IsValid(geometry)';
+            if ($this->bVerbose) echo "$sSQL\n";
+            if (!pg_send_query($aDBInstances[$i], $sSQL)) {
+                fail(pg_last_error($aDBInstances[$i]));
+            }
+        }
+
+        // last thread for interpolation lines
+        // https://secure.php.net/manual/en/function.pg-connect.php
+        $DSN = CONST_Database_DSN;
+        $DSN = preg_replace('/^pgsql:/', '', $DSN);
+        $DSN = preg_replace('/;/', ' ', $DSN);
+        $aDBInstances[$iLoadThreads] = pg_connect($DSN, PGSQL_CONNECT_FORCE_NEW);
+        pg_ping($aDBInstances[$iLoadThreads]);
+        $sSQL = 'insert into location_property_osmline';
+        $sSQL .= ' (osm_id, address, linegeo)';
+        $sSQL .= ' SELECT osm_id, address, geometry from place where ';
+        $sSQL .= "class='place' and type='houses' and osm_type='W' and ST_GeometryType(geometry) = 'ST_LineString'";
+        if ($this->bVerbose) echo "$sSQL\n";
+        if (!pg_send_query($aDBInstances[$iLoadThreads], $sSQL)) {
+            fail(pg_last_error($aDBInstances[$iLoadThreads]));
+        }
+
+        $bFailed = false;
+        for ($i = 0; $i <= $iLoadThreads; $i++) {
+            while (($hPGresult = pg_get_result($aDBInstances[$i])) !== false) {
+                $resultStatus = pg_result_status($hPGresult);
+                // PGSQL_EMPTY_QUERY, PGSQL_COMMAND_OK, PGSQL_TUPLES_OK,
+                // PGSQL_COPY_OUT, PGSQL_COPY_IN, PGSQL_BAD_RESPONSE,
+                // PGSQL_NONFATAL_ERROR and PGSQL_FATAL_ERROR
+                // echo 'Query result ' . $i . ' is: ' . $resultStatus . "\n";
+                if ($resultStatus != PGSQL_COMMAND_OK && $resultStatus != PGSQL_TUPLES_OK) {
+                    $resultError = pg_result_error($hPGresult);
+                    echo '-- error text ' . $i . ': ' . $resultError . "\n";
+                    $bFailed = true;
+                }
+            }
+        }
+        if ($bFailed) {
+            fail('SQL errors loading placex and/or location_property_osmline tables');
+        }
+
+        for ($i = 0; $i < $this->iInstances; $i++) {
+            pg_close($aDBInstances[$i]);
+        }
+
+        echo "\n";
+        info('Reanalysing database');
+        $this->pgsqlRunScript('ANALYSE');
+
+        $sDatabaseDate = getDatabaseDate($this->oDB);
+        $this->oDB->exec('TRUNCATE import_status');
+        if (!$sDatabaseDate) {
+            warn('could not determine database date.');
+        } else {
+            $sSQL = "INSERT INTO import_status (lastimportdate) VALUES('".$sDatabaseDate."')";
+            $this->oDB->exec($sSQL);
+            echo "Latest data imported from $sDatabaseDate.\n";
+        }
+    }
+
+    public function importTigerData()
+    {
+        info('Import Tiger data');
+
+        $sTemplate = file_get_contents(CONST_BasePath.'/sql/tiger_import_start.sql');
+        $sTemplate = str_replace('{www-user}', CONST_Database_Web_User, $sTemplate);
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-data}',
+            CONST_Tablespace_Aux_Data,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-index}',
+            CONST_Tablespace_Aux_Index,
+            $sTemplate
+        );
+        $this->pgsqlRunScript($sTemplate, false);
+
+        $aDBInstances = array();
+        for ($i = 0; $i < $this->iInstances; $i++) {
+            // https://secure.php.net/manual/en/function.pg-connect.php
+            $DSN = CONST_Database_DSN;
+            $DSN = preg_replace('/^pgsql:/', '', $DSN);
+            $DSN = preg_replace('/;/', ' ', $DSN);
+            $aDBInstances[$i] = pg_connect($DSN, PGSQL_CONNECT_FORCE_NEW | PGSQL_CONNECT_ASYNC);
+            pg_ping($aDBInstances[$i]);
+        }
+
+        foreach (glob(CONST_Tiger_Data_Path.'/*.sql') as $sFile) {
+            echo $sFile.': ';
+            $hFile = fopen($sFile, 'r');
+            $sSQL = fgets($hFile, 100000);
+            $iLines = 0;
+            while (true) {
+                for ($i = 0; $i < $this->iInstances; $i++) {
+                    if (!pg_connection_busy($aDBInstances[$i])) {
+                        while (pg_get_result($aDBInstances[$i]));
+                        $sSQL = fgets($hFile, 100000);
+                        if (!$sSQL) break 2;
+                        if (!pg_send_query($aDBInstances[$i], $sSQL)) fail(pg_last_error($aDBInstances[$i]));
+                        $iLines++;
+                        if ($iLines == 1000) {
+                            echo '.';
+                            $iLines = 0;
+                        }
+                    }
+                }
+                usleep(10);
+            }
+            fclose($hFile);
+
+            $bAnyBusy = true;
+            while ($bAnyBusy) {
+                $bAnyBusy = false;
+                for ($i = 0; $i < $this->iInstances; $i++) {
+                    if (pg_connection_busy($aDBInstances[$i])) $bAnyBusy = true;
+                }
+                usleep(10);
+            }
+            echo "\n";
+        }
+
+        for ($i = 0; $i < $this->iInstances; $i++) {
+            pg_close($aDBInstances[$i]);
+        }
+
+        info('Creating indexes on Tiger data');
+        $sTemplate = file_get_contents(CONST_BasePath.'/sql/tiger_import_finish.sql');
+        $sTemplate = str_replace('{www-user}', CONST_Database_Web_User, $sTemplate);
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-data}',
+            CONST_Tablespace_Aux_Data,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-index}',
+            CONST_Tablespace_Aux_Index,
+            $sTemplate
+        );
+        $this->pgsqlRunScript($sTemplate, false);
+    }
+
+    public function calculatePostcodes($bCMDResultAll)
+    {
+        info('Calculate Postcodes');
+        $this->pgExec('TRUNCATE location_postcode');
+
+        $sSQL  = 'INSERT INTO location_postcode';
+        $sSQL .= ' (place_id, indexed_status, country_code, postcode, geometry) ';
+        $sSQL .= "SELECT nextval('seq_place'), 1, country_code,";
+        $sSQL .= "       upper(trim (both ' ' from address->'postcode')) as pc,";
+        $sSQL .= '       ST_Centroid(ST_Collect(ST_Centroid(geometry)))';
+        $sSQL .= '  FROM placex';
+        $sSQL .= " WHERE address ? 'postcode' AND address->'postcode' NOT SIMILAR TO '%(,|;)%'";
+        $sSQL .= '       AND geometry IS NOT null';
+        $sSQL .= ' GROUP BY country_code, pc';
+        $this->pgExec($sSQL);
+
+        if (CONST_Use_Extra_US_Postcodes) {
+            // only add postcodes that are not yet available in OSM
+            $sSQL  = 'INSERT INTO location_postcode';
+            $sSQL .= ' (place_id, indexed_status, country_code, postcode, geometry) ';
+            $sSQL .= "SELECT nextval('seq_place'), 1, 'us', postcode,";
+            $sSQL .= '       ST_SetSRID(ST_Point(x,y),4326)';
+            $sSQL .= '  FROM us_postcode WHERE postcode NOT IN';
+            $sSQL .= '        (SELECT postcode FROM location_postcode';
+            $sSQL .= "          WHERE country_code = 'us')";
+            $this->pgExec($sSQL);
+        }
+
+        // add missing postcodes for GB (if available)
+        $sSQL  = 'INSERT INTO location_postcode';
+        $sSQL .= ' (place_id, indexed_status, country_code, postcode, geometry) ';
+        $sSQL .= "SELECT nextval('seq_place'), 1, 'gb', postcode, geometry";
+        $sSQL .= '  FROM gb_postcode WHERE postcode NOT IN';
+        $sSQL .= '           (SELECT postcode FROM location_postcode';
+        $sSQL .= "             WHERE country_code = 'gb')";
+        $this->pgExec($sSQL);
+
+        if (!$bCMDResultAll) {
+            $sSQL = "DELETE FROM word WHERE class='place' and type='postcode'";
+            $sSQL .= 'and word NOT IN (SELECT postcode FROM location_postcode)';
+            $this->pgExec($sSQL);
+        }
+
+        $sSQL = 'SELECT count(getorcreate_postcode_id(v)) FROM ';
+        $sSQL .= '(SELECT distinct(postcode) as v FROM location_postcode) p';
+        $this->pgExec($sSQL);
+    }
+
+    public function index($bIndexNoanalyse)
+    {
+        $sOutputFile = '';
+        $sBaseCmd = CONST_InstallPath.'/nominatim/nominatim -i -d '.$this->aDSNInfo['database'].' -P '
+            .$this->aDSNInfo['port'].' -t '.$this->iInstances.$sOutputFile;
+        if (isset($this->aDSNInfo['hostspec'])) {
+            $sBaseCmd .= ' -H '.$this->aDSNInfo['hostspec'];
+        }
+        if (isset($this->aDSNInfo['username'])) {
+            $sBaseCmd .= ' -U '.$this->aDSNInfo['username'];
+        }
+
+        info('Index ranks 0 - 4');
+        $iStatus = $this->runWithPgEnv($sBaseCmd.' -R 4');
+        if ($iStatus != 0) {
+            fail('error status ' . $iStatus . ' running nominatim!');
+        }
+        if (!$bIndexNoanalyse) $this->pgsqlRunScript('ANALYSE');
+
+        info('Index ranks 5 - 25');
+        $iStatus = $this->runWithPgEnv($sBaseCmd.' -r 5 -R 25');
+        if ($iStatus != 0) {
+            fail('error status ' . $iStatus . ' running nominatim!');
+        }
+        if (!$bIndexNoanalyse) $this->pgsqlRunScript('ANALYSE');
+
+        info('Index ranks 26 - 30');
+        $iStatus = $this->runWithPgEnv($sBaseCmd.' -r 26');
+        if ($iStatus != 0) {
+            fail('error status ' . $iStatus . ' running nominatim!');
+        }
+
+        info('Index postcodes');
+        $sSQL = 'UPDATE location_postcode SET indexed_status = 0';
+        $this->pgExec($sSQL);
+    }
+
+    public function createSearchIndices()
+    {
+        info('Create Search indices');
+
+        $sTemplate = file_get_contents(CONST_BasePath.'/sql/indices.src.sql');
+        if (!$this->dbReverseOnly()) {
+            $sTemplate .= file_get_contents(CONST_BasePath.'/sql/indices_search.src.sql');
+        }
+        $sTemplate = str_replace('{www-user}', CONST_Database_Web_User, $sTemplate);
+        $sTemplate = $this->replaceTablespace(
+            '{ts:address-index}',
+            CONST_Tablespace_Address_Index,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:search-index}',
+            CONST_Tablespace_Search_Index,
+            $sTemplate
+        );
+        $sTemplate = $this->replaceTablespace(
+            '{ts:aux-index}',
+            CONST_Tablespace_Aux_Index,
+            $sTemplate
+        );
+        $this->pgsqlRunScript($sTemplate);
+    }
+
+    public function createCountryNames()
+    {
+        info('Create search index for default country names');
+
+        $this->pgsqlRunScript("select getorcreate_country(make_standard_name('uk'), 'gb')");
+        $this->pgsqlRunScript("select getorcreate_country(make_standard_name('united states'), 'us')");
+        $this->pgsqlRunScript('select count(*) from (select getorcreate_country(make_standard_name(country_code), country_code) from country_name where country_code is not null) as x');
+        $this->pgsqlRunScript("select count(*) from (select getorcreate_country(make_standard_name(name->'name'), country_code) from country_name where name ? 'name') as x");
+        $sSQL = 'select count(*) from (select getorcreate_country(make_standard_name(v),'
+            .'country_code) from (select country_code, skeys(name) as k, svals(name) as v from country_name) x where k ';
+        if (CONST_Languages) {
+            $sSQL .= 'in ';
+            $sDelim = '(';
+            foreach (explode(',', CONST_Languages) as $sLang) {
+                $sSQL .= $sDelim."'name:$sLang'";
+                $sDelim = ',';
+            }
+            $sSQL .= ')';
+        } else {
+            // all include all simple name tags
+            $sSQL .= "like 'name:%'";
+        }
+        $sSQL .= ') v';
+        $this->pgsqlRunScript($sSQL);
+    }
+
+    public function drop()
+    {
+        info('Drop tables only required for updates');
+
+        // The implementation is potentially a bit dangerous because it uses
+        // a positive selection of tables to keep, and deletes everything else.
+        // Including any tables that the unsuspecting user might have manually
+        // created. USE AT YOUR OWN PERIL.
+        // tables we want to keep. everything else goes.
+        $aKeepTables = array(
+                        '*columns',
+                        'import_polygon_*',
+                        'import_status',
+                        'place_addressline',
+                        'location_postcode',
+                        'location_property*',
+                        'placex',
+                        'search_name',
+                        'seq_*',
+                        'word',
+                        'query_log',
+                        'new_query_log',
+                        'spatial_ref_sys',
+                        'country_name',
+                        'place_classtype_*',
+                        'country_osm_grid'
+                       );
+
+        $aDropTables = array();
+        $aHaveTables = $this->oDB->getCol("SELECT tablename FROM pg_tables WHERE schemaname='public'");
+
+        foreach ($aHaveTables as $sTable) {
+            $bFound = false;
+            foreach ($aKeepTables as $sKeep) {
+                if (fnmatch($sKeep, $sTable)) {
+                    $bFound = true;
+                    break;
+                }
+            }
+            if (!$bFound) array_push($aDropTables, $sTable);
+        }
+        foreach ($aDropTables as $sDrop) {
+            if ($this->bVerbose) echo "Dropping table $sDrop\n";
+            $this->oDB->exec("DROP TABLE $sDrop CASCADE");
+            // ignore warnings/errors as they might be caused by a table having
+            // been deleted already by CASCADE
+        }
+
+        if (!is_null(CONST_Osm2pgsql_Flatnode_File) && CONST_Osm2pgsql_Flatnode_File) {
+            if (file_exists(CONST_Osm2pgsql_Flatnode_File)) {
+                if ($this->bVerbose) echo 'Deleting '.CONST_Osm2pgsql_Flatnode_File."\n";
+                unlink(CONST_Osm2pgsql_Flatnode_File);
+            }
+        }
+    }
+
+    private function pgsqlRunDropAndRestore($sDumpFile)
+    {
+        $sCMD = 'pg_restore -p '.$this->aDSNInfo['port'].' -d '.$this->aDSNInfo['database'].' --no-owner -Fc --clean '.$sDumpFile;
+        if ($this->oDB->getPostgresVersion() >= 9.04) {
+            $sCMD .= ' --if-exists';
+        }
+        if (isset($this->aDSNInfo['hostspec'])) {
+            $sCMD .= ' -h '.$this->aDSNInfo['hostspec'];
+        }
+        if (isset($this->aDSNInfo['username'])) {
+            $sCMD .= ' -U '.$this->aDSNInfo['username'];
+        }
+
+        $this->runWithPgEnv($sCMD);
+    }
+
+    private function pgsqlRunScript($sScript, $bfatal = true)
+    {
+        runSQLScript(
+            $sScript,
+            $bfatal,
+            $this->bVerbose,
+            $this->sIgnoreErrors
+        );
+    }
+
+    private function createSqlFunctions()
+    {
+        $sTemplate = file_get_contents(CONST_BasePath.'/sql/functions.sql');
+        $sTemplate = str_replace('{modulepath}', $this->sModulePath, $sTemplate);
+        if ($this->bEnableDiffUpdates) {
+            $sTemplate = str_replace('RETURN NEW; -- %DIFFUPDATES%', '--', $sTemplate);
+        }
+        if ($this->bEnableDebugStatements) {
+            $sTemplate = str_replace('--DEBUG:', '', $sTemplate);
+        }
+        if (CONST_Limit_Reindexing) {
+            $sTemplate = str_replace('--LIMIT INDEXING:', '', $sTemplate);
+        }
+        if (!CONST_Use_US_Tiger_Data) {
+            $sTemplate = str_replace('-- %NOTIGERDATA% ', '', $sTemplate);
+        }
+        if (!CONST_Use_Aux_Location_data) {
+            $sTemplate = str_replace('-- %NOAUXDATA% ', '', $sTemplate);
+        }
+
+        $sReverseOnly = $this->dbReverseOnly() ? 'true' : 'false';
+        $sTemplate = str_replace('%REVERSE-ONLY%', $sReverseOnly, $sTemplate);
+
+        $this->pgsqlRunScript($sTemplate);
+    }
+
+    private function pgsqlRunPartitionScript($sTemplate)
+    {
+        $sSQL = 'select distinct partition from country_name';
+        $aPartitions = $this->oDB->getCol($sSQL);
+        if (!$this->bNoPartitions) $aPartitions[] = 0;
+
+        preg_match_all('#^-- start(.*?)^-- end#ms', $sTemplate, $aMatches, PREG_SET_ORDER);
+        foreach ($aMatches as $aMatch) {
+            $sResult = '';
+            foreach ($aPartitions as $sPartitionName) {
+                $sResult .= str_replace('-partition-', $sPartitionName, $aMatch[1]);
+            }
+            $sTemplate = str_replace($aMatch[0], $sResult, $sTemplate);
+        }
+
+        $this->pgsqlRunScript($sTemplate);
+    }
+
+    private function pgsqlRunScriptFile($sFilename)
+    {
+        if (!file_exists($sFilename)) fail('unable to find '.$sFilename);
+
+        $sCMD = 'psql -p '.$this->aDSNInfo['port'].' -d '.$this->aDSNInfo['database'];
+        if (!$this->bVerbose) {
+            $sCMD .= ' -q';
+        }
+        if (isset($this->aDSNInfo['hostspec'])) {
+            $sCMD .= ' -h '.$this->aDSNInfo['hostspec'];
+        }
+        if (isset($this->aDSNInfo['username'])) {
+            $sCMD .= ' -U '.$this->aDSNInfo['username'];
+        }
+        $aProcEnv = null;
+        if (isset($this->aDSNInfo['password'])) {
+            $aProcEnv = array_merge(array('PGPASSWORD' => $this->aDSNInfo['password']), $_ENV);
+        }
+        $ahGzipPipes = null;
+        if (preg_match('/\\.gz$/', $sFilename)) {
+            $aDescriptors = array(
+                             0 => array('pipe', 'r'),
+                             1 => array('pipe', 'w'),
+                             2 => array('file', '/dev/null', 'a')
+                            );
+            $hGzipProcess = proc_open('zcat '.$sFilename, $aDescriptors, $ahGzipPipes);
+            if (!is_resource($hGzipProcess)) fail('unable to start zcat');
+            $aReadPipe = $ahGzipPipes[1];
+            fclose($ahGzipPipes[0]);
+        } else {
+            $sCMD .= ' -f '.$sFilename;
+            $aReadPipe = array('pipe', 'r');
+        }
+        $aDescriptors = array(
+                         0 => $aReadPipe,
+                         1 => array('pipe', 'w'),
+                         2 => array('file', '/dev/null', 'a')
+                        );
+        $ahPipes = null;
+        $hProcess = proc_open($sCMD, $aDescriptors, $ahPipes, null, $aProcEnv);
+        if (!is_resource($hProcess)) fail('unable to start pgsql');
+        // TODO: error checking
+        while (!feof($ahPipes[1])) {
+            echo fread($ahPipes[1], 4096);
+        }
+        fclose($ahPipes[1]);
+        $iReturn = proc_close($hProcess);
+        if ($iReturn > 0) {
+            fail("pgsql returned with error code ($iReturn)");
+        }
+        if ($ahGzipPipes) {
+            fclose($ahGzipPipes[1]);
+            proc_close($hGzipProcess);
+        }
+    }
+
+    private function replaceTablespace($sTemplate, $sTablespace, $sSql)
+    {
+        if ($sTablespace) {
+            $sSql = str_replace($sTemplate, 'TABLESPACE "'.$sTablespace.'"', $sSql);
+        } else {
+            $sSql = str_replace($sTemplate, '', $sSql);
+        }
+        return $sSql;
+    }
+
+    private function runWithPgEnv($sCmd)
+    {
+        if ($this->bVerbose) {
+            echo "Execute: $sCmd\n";
+        }
+
+        $aProcEnv = null;
+
+        if (isset($this->aDSNInfo['password'])) {
+            $aProcEnv = array_merge(array('PGPASSWORD' => $this->aDSNInfo['password']), $_ENV);
+        }
+
+        return runWithEnv($sCmd, $aProcEnv);
+    }
+
+    /**
+     * Execute the SQL command on the open database.
+     *
+     * @param string $sSQL SQL command to execute.
+     *
+     * @return null
+     *
+     * @pre connect() must have been called.
+     */
+    private function pgExec($sSQL)
+    {
+        $this->oDB->exec($sSQL);
+    }
+
+    /**
+     * Check if the database is in reverse-only mode.
+     *
+     * @return True if there is no search_name table and infrastructure.
+     */
+    private function dbReverseOnly()
+    {
+        return !($this->oDB->tableExists('search_name'));
+    }
+}

lib/setup_functions.php

@@ -0,0 +1,31 @@
+<?php
+
+function checkInFile($sOSMFile)
+{
+    if (!isset($sOSMFile)) {
+        fail('missing --osm-file for data import');
+    }
+
+    if (!file_exists($sOSMFile)) {
+        fail('the path supplied to --osm-file does not exist');
+    }
+
+    if (!is_readable($sOSMFile)) {
+        fail('osm-file "' . $aCMDResult['osm-file'] . '" not readable');
+    }
+}
+
+function checkModulePresence()
+{
+    // Try accessing the C module, so we know early if something is wrong.
+    // Raises Nominatim\DatabaseError on failure
+
+    $sModulePath = CONST_Database_Module_Path;
+    $sSQL = "CREATE FUNCTION nominatim_test_import_func(text) RETURNS text AS '";
+    $sSQL .= $sModulePath . "/nominatim.so', 'transliteration' LANGUAGE c IMMUTABLE STRICT";
+    $sSQL .= ';DROP FUNCTION nominatim_test_import_func(text);';
+
+    $oDB = new \Nominatim\DB();
+    $oDB->connect();
+    $oDB->exec($sSQL, null, 'Database server failed to load '.$sModulePath.'/nominatim.so module');
+}

lib/template/address-geocodejson.php

@@ -0,0 +1,81 @@
+<?php
+
+// https://github.com/geocoders/geocodejson-spec/
+
+$aFilteredPlaces = array();
+
+if (empty($aPlace)) {
+    if (isset($sError))
+        $aFilteredPlaces['error'] = $sError;
+    else $aFilteredPlaces['error'] = 'Unable to geocode';
+    javascript_renderData($aFilteredPlaces);
+} else {
+    $aFilteredPlaces = array(
+                        'type' => 'Feature',
+                        'properties' => array(
+                                         'geocoding' => array()
+                                        )
+                       );
+
+    if (isset($aPlace['place_id'])) $aFilteredPlaces['properties']['geocoding']['place_id'] = $aPlace['place_id'];
+    $sOSMType = formatOSMType($aPlace['osm_type']);
+    if ($sOSMType) {
+        $aFilteredPlaces['properties']['geocoding']['osm_type'] = $sOSMType;
+        $aFilteredPlaces['properties']['geocoding']['osm_id'] = $aPlace['osm_id'];
+    }
+
+    $aFilteredPlaces['properties']['geocoding']['type'] = $aPlace['type'];
+
+    $aFilteredPlaces['properties']['geocoding']['accuracy'] = (int) $fDistance;
+
+    $aFilteredPlaces['properties']['geocoding']['label'] = $aPlace['langaddress'];
+
+    $aFilteredPlaces['properties']['geocoding']['name'] = $aPlace['placename'];
+
+    if (isset($aPlace['address'])) {
+        $aFieldMappings = array(
+                           'house_number' => 'housenumber',
+                           'road' => 'street',
+                           'locality' => 'locality',
+                           'postcode' => 'postcode',
+                           'city' => 'city',
+                           'district' => 'district',
+                           'county' => 'county',
+                           'state' => 'state',
+                           'country' => 'country'
+                          );
+
+        $aAddressNames = $aPlace['address']->getAddressNames();
+        foreach ($aFieldMappings as $sFrom => $sTo) {
+            if (isset($aAddressNames[$sFrom])) {
+                $aFilteredPlaces['properties']['geocoding'][$sTo] = $aAddressNames[$sFrom];
+            }
+        }
+
+        $aFilteredPlaces['properties']['geocoding']['admin']
+            = $aPlace['address']->getAdminLevels();
+    }
+
+    if (isset($aPlace['asgeojson'])) {
+        $aFilteredPlaces['geometry'] = json_decode($aPlace['asgeojson']);
+    } else {
+        $aFilteredPlaces['geometry'] = array(
+                                        'type' => 'Point',
+                                        'coordinates' => array(
+                                                          (float) $aPlace['lon'],
+                                                          (float) $aPlace['lat']
+                                                         )
+                                       );
+    }
+
+    javascript_renderData(array(
+                           'type' => 'FeatureCollection',
+                           'geocoding' => array(
+                                           'version' => '0.1.0',
+                                           'attribution' => 'Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright',
+                                           'licence' => 'ODbL',
+                                           'query' => $sQuery
+                                          ),
+                           'features' => array($aFilteredPlaces)
+                          ));
+}

lib/template/address-geojson.php

@@ -0,0 +1,69 @@
+<?php
+
+$aFilteredPlaces = array();
+
+if (empty($aPlace)) {
+    if (isset($sError))
+        $aFilteredPlaces['error'] = $sError;
+    else $aFilteredPlaces['error'] = 'Unable to geocode';
+    javascript_renderData($aFilteredPlaces);
+} else {
+    $aFilteredPlaces = array(
+                        'type' => 'Feature',
+                        'properties' => array()
+                       );
+
+    if (isset($aPlace['place_id'])) $aFilteredPlaces['properties']['place_id'] = $aPlace['place_id'];
+    $sOSMType = formatOSMType($aPlace['osm_type']);
+    if ($sOSMType) {
+        $aFilteredPlaces['properties']['osm_type'] = $sOSMType;
+        $aFilteredPlaces['properties']['osm_id'] = $aPlace['osm_id'];
+    }
+
+    $aFilteredPlaces['properties']['place_rank'] = $aPlace['rank_search'];
+
+    $aFilteredPlaces['properties']['category'] = $aPlace['class'];
+    $aFilteredPlaces['properties']['type'] = $aPlace['type'];
+
+    $aFilteredPlaces['properties']['importance'] = $aPlace['importance'];
+
+    $aFilteredPlaces['properties']['addresstype'] = strtolower($aPlace['addresstype']);
+
+    $aFilteredPlaces['properties']['name'] = $aPlace['placename'];
+
+    $aFilteredPlaces['properties']['display_name'] = $aPlace['langaddress'];
+
+    if (isset($aPlace['address'])) {
+        $aFilteredPlaces['properties']['address'] = $aPlace['address']->getAddressNames();
+    }
+    if (isset($aPlace['sExtraTags'])) $aFilteredPlaces['properties']['extratags'] = $aPlace['sExtraTags'];
+    if (isset($aPlace['sNameDetails'])) $aFilteredPlaces['properties']['namedetails'] = $aPlace['sNameDetails'];
+
+    if (isset($aPlace['aBoundingBox'])) {
+        $aFilteredPlaces['bbox'] = array(
+                                    (float) $aPlace['aBoundingBox'][2], // minlon
+                                    (float) $aPlace['aBoundingBox'][0], // minlat
+                                    (float) $aPlace['aBoundingBox'][3], // maxlon
+                                    (float) $aPlace['aBoundingBox'][1]  // maxlat
+                                   );
+    }
+
+    if (isset($aPlace['asgeojson'])) {
+        $aFilteredPlaces['geometry'] = json_decode($aPlace['asgeojson']);
+    } else {
+        $aFilteredPlaces['geometry'] = array(
+                                        'type' => 'Point',
+                                        'coordinates' => array(
+                                                          (float) $aPlace['lon'],
+                                                          (float) $aPlace['lat']
+                                                         )
+                                       );
+    }
+
+
+    javascript_renderData(array(
+                           'type' => 'FeatureCollection',
+                           'licence' => 'Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright',
+                           'features' => array($aFilteredPlaces)
+                          ));
+}

lib/template/address-json.php

@@ -2,55 +2,61 @@
 
 $aFilteredPlaces = array();
 
-if (!sizeof($aPlace))
-{
+if (empty($aPlace)) {
     if (isset($sError))
         $aFilteredPlaces['error'] = $sError;
-    else
-        $aFilteredPlaces['error'] = 'Unable to geocode';
-}
-else
-{
+    else $aFilteredPlaces['error'] = 'Unable to geocode';
+} else {
     if (isset($aPlace['place_id'])) $aFilteredPlaces['place_id'] = $aPlace['place_id'];
-    $aFilteredPlaces['licence'] = "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright";
+    $aFilteredPlaces['licence'] = 'Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright';
     $sOSMType = formatOSMType($aPlace['osm_type']);
-    if ($sOSMType)
-    {
+    if ($sOSMType) {
         $aFilteredPlaces['osm_type'] = $sOSMType;
         $aFilteredPlaces['osm_id'] = $aPlace['osm_id'];
     }
     if (isset($aPlace['lat'])) $aFilteredPlaces['lat'] = $aPlace['lat'];
     if (isset($aPlace['lon'])) $aFilteredPlaces['lon'] = $aPlace['lon'];
+
+    if ($sOutputFormat == 'jsonv2' || $sOutputFormat == 'geojson') {
+        $aFilteredPlaces['place_rank'] = $aPlace['rank_search'];
+
+        $aFilteredPlaces['category'] = $aPlace['class'];
+        $aFilteredPlaces['type'] = $aPlace['type'];
+
+        $aFilteredPlaces['importance'] = $aPlace['importance'];
+
+        $aFilteredPlaces['addresstype'] = strtolower($aPlace['addresstype']);
+
+        $aFilteredPlaces['name'] = $aPlace['placename'];
+    }
+
     $aFilteredPlaces['display_name'] = $aPlace['langaddress'];
-    if (isset($aPlace['aAddress'])) $aFilteredPlaces['address'] = $aPlace['aAddress'];
+
+    if (isset($aPlace['address'])) {
+        $aFilteredPlaces['address'] = $aPlace['address']->getAddressNames();
+    }
     if (isset($aPlace['sExtraTags'])) $aFilteredPlaces['extratags'] = $aPlace['sExtraTags'];
     if (isset($aPlace['sNameDetails'])) $aFilteredPlaces['namedetails'] = $aPlace['sNameDetails'];
 
-    if (isset($aPlace['aBoundingBox']))
-    {
+    if (isset($aPlace['aBoundingBox'])) {
         $aFilteredPlaces['boundingbox'] = $aPlace['aBoundingBox'];
     }
 
-    if (isset($aPlace['asgeojson']))
-    {
+    if (isset($aPlace['asgeojson'])) {
         $aFilteredPlaces['geojson'] = json_decode($aPlace['asgeojson']);
     }
 
-    if (isset($aPlace['assvg']))
-    {
+    if (isset($aPlace['assvg'])) {
         $aFilteredPlaces['svg'] = $aPlace['assvg'];
     }
 
-    if (isset($aPlace['astext']))
-    {
+    if (isset($aPlace['astext'])) {
         $aFilteredPlaces['geotext'] = $aPlace['astext'];
     }
 
-    if (isset($aPlace['askml']))
-    {
+    if (isset($aPlace['askml'])) {
         $aFilteredPlaces['geokml'] = $aPlace['askml'];
     }
 }
 
 javascript_renderData($aFilteredPlaces);
-

lib/template/address-jsonv2.php

@@ -1,68 +0,0 @@
-<?php
-
-$aFilteredPlaces = array();
-
-if (!sizeof($aPlace))
-{
-    if (isset($sError))
-        $aFilteredPlaces['error'] = $sError;
-    else
-        $aFilteredPlaces['error'] = 'Unable to geocode';
-}
-else
-{
-    if ($aPlace['place_id']) $aFilteredPlaces['place_id'] = $aPlace['place_id'];
-    $aFilteredPlaces['licence'] = "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright";
-    $sOSMType = formatOSMType($aPlace['osm_type']);
-    if ($sOSMType)
-    {
-        $aFilteredPlaces['osm_type'] = $sOSMType;
-        $aFilteredPlaces['osm_id'] = $aPlace['osm_id'];
-    }
-    if (isset($aPlace['lat'])) $aFilteredPlaces['lat'] = $aPlace['lat'];
-    if (isset($aPlace['lon'])) $aFilteredPlaces['lon'] = $aPlace['lon'];
-
-    $aFilteredPlaces['place_rank'] = $aPlace['rank_search'];
-
-    $aFilteredPlaces['category'] = $aPlace['class'];
-    $aFilteredPlaces['type'] = $aPlace['type'];
-
-    $aFilteredPlaces['importance'] = $aPlace['importance'];
-
-    $aFilteredPlaces['addresstype'] = strtolower($aPlace['addresstype']);
-
-    $aFilteredPlaces['display_name'] = $aPlace['langaddress'];
-    $aFilteredPlaces['name'] = $aPlace['placename'];
-
-    if (isset($aPlace['aAddress'])) $aFilteredPlaces['address'] = $aPlace['aAddress'];
-    if (isset($aPlace['sExtraTags'])) $aFilteredPlaces['extratags'] = $aPlace['sExtraTags'];
-    if (isset($aPlace['sNameDetails'])) $aFilteredPlaces['namedetails'] = $aPlace['sNameDetails'];
-
-    if (isset($aPlace['aBoundingBox']))
-    {
-        $aFilteredPlaces['boundingbox'] = $aPlace['aBoundingBox'];
-    }
-
-    if (isset($aPlace['asgeojson']))
-    {
-        $aFilteredPlaces['geojson'] = json_decode($aPlace['asgeojson']);
-    }
-
-    if (isset($aPlace['assvg']))
-    {
-        $aFilteredPlaces['svg'] = $aPlace['assvg'];
-    }
-
-    if (isset($aPlace['astext']))
-    {
-        $aFilteredPlaces['geotext'] = $aPlace['astext'];
-    }
-
-    if (isset($aPlace['askml']))
-    {
-        $aFilteredPlaces['geokml'] = $aPlace['askml'];
-    }
-
-}
-
-javascript_renderData($aFilteredPlaces);

lib/template/address-xml.php

@@ -1,103 +1,87 @@
 <?php
-header("content-type: text/xml; charset=UTF-8");
+header('content-type: text/xml; charset=UTF-8');
 
-echo "<";
-echo "?xml version=\"1.0\" encoding=\"UTF-8\" ?";
+echo '<';
+echo '?xml version="1.0" encoding="UTF-8" ?';
 echo ">\n";
 
-echo "<reversegeocode";
+echo '<reversegeocode';
 echo " timestamp='".date(DATE_RFC822)."'";
 echo " attribution='Data © OpenStreetMap contributors, ODbL 1.0. http://www.openstreetmap.org/copyright'";
 echo " querystring='".htmlspecialchars($_SERVER['QUERY_STRING'], ENT_QUOTES)."'";
 echo ">\n";
 
-if (!sizeof($aPlace))
-{
+if (empty($aPlace)) {
     if (isset($sError))
         echo "<error>$sError</error>";
-    else
-        echo "<error>Unable to geocode</error>";
-}
-else
-{
-    echo "<result";
+    else echo '<error>Unable to geocode</error>';
+} else {
+    echo '<result';
     if ($aPlace['place_id']) echo ' place_id="'.$aPlace['place_id'].'"';
     $sOSMType = formatOSMType($aPlace['osm_type']);
     if ($sOSMType) echo ' osm_type="'.$sOSMType.'"'.' osm_id="'.$aPlace['osm_id'].'"';
     if ($aPlace['ref']) echo ' ref="'.htmlspecialchars($aPlace['ref']).'"';
     if (isset($aPlace['lat'])) echo ' lat="'.htmlspecialchars($aPlace['lat']).'"';
     if (isset($aPlace['lon'])) echo ' lon="'.htmlspecialchars($aPlace['lon']).'"';
-    if (isset($aPlace['aBoundingBox']))
-    {
+    if (isset($aPlace['aBoundingBox'])) {
         echo ' boundingbox="';
         echo join(',', $aPlace['aBoundingBox']);
         echo '"';
     }
 
-    if (isset($aPlace['asgeojson']))
-    {
+    if (isset($aPlace['asgeojson'])) {
         echo ' geojson=\'';
         echo $aPlace['asgeojson'];
         echo '\'';
     }
 
-    if (isset($aPlace['assvg']))
-    {
+    if (isset($aPlace['assvg'])) {
         echo ' geosvg=\'';
         echo $aPlace['assvg'];
         echo '\'';
     }
 
-    if (isset($aPlace['astext']))
-    {
+    if (isset($aPlace['astext'])) {
         echo ' geotext=\'';
         echo $aPlace['astext'];
         echo '\'';
     }
-    echo ">".htmlspecialchars($aPlace['langaddress'])."</result>";
+    echo '>'.htmlspecialchars($aPlace['langaddress']).'</result>';
 
-    if (isset($aPlace['aAddress']))
-    {
-        echo "<addressparts>";
-        foreach($aPlace['aAddress'] as $sKey => $sValue)
-        {
-            $sKey = str_replace(' ','_',$sKey);
+    if (isset($aPlace['address'])) {
+        echo '<addressparts>';
+        foreach ($aPlace['address']->getAddressNames() as $sKey => $sValue) {
+            $sKey = str_replace(' ', '_', $sKey);
             echo "<$sKey>";
             echo htmlspecialchars($sValue);
             echo "</$sKey>";
         }
-        echo "</addressparts>";
+        echo '</addressparts>';
     }
 
-    if (isset($aPlace['sExtraTags']))
-    {
-        echo "<extratags>";
-        foreach ($aPlace['sExtraTags'] as $sKey => $sValue)
-        {
+    if (isset($aPlace['sExtraTags'])) {
+        echo '<extratags>';
+        foreach ($aPlace['sExtraTags'] as $sKey => $sValue) {
             echo '<tag key="'.htmlspecialchars($sKey).'" value="'.htmlspecialchars($sValue).'"/>';
         }
-        echo "</extratags>";
+        echo '</extratags>';
     }
 
-    if (isset($aPlace['sNameDetails']))
-    {
-        echo "<namedetails>";
-        foreach ($aPlace['sNameDetails'] as $sKey => $sValue)
-        {
+    if (isset($aPlace['sNameDetails'])) {
+        echo '<namedetails>';
+        foreach ($aPlace['sNameDetails'] as $sKey => $sValue) {
             echo '<name desc="'.htmlspecialchars($sKey).'">';
             echo htmlspecialchars($sValue);
-            echo "</name>";
+            echo '</name>';
         }
-        echo "</namedetails>";
+        echo '</namedetails>';
     }
 
-    if (isset($aPlace['askml']))
-    {
+    if (isset($aPlace['askml'])) {
         echo "\n<geokml>";
         echo $aPlace['askml'];
-        echo "</geokml>";
+        echo '</geokml>';
     }
-
 }
 
-echo "</reversegeocode>";
+echo '</reversegeocode>';

lib/template/details-html.php

@@ -46,8 +46,7 @@
     function hash_to_subtable($aAssociatedList)
     {
         $sHTML = '';
-        foreach($aAssociatedList as $sKey => $sValue)
-        {
+        foreach ($aAssociatedList as $sKey => $sValue) {
             $sHTML = $sHTML.' <div class="line"><span class="name">'.$sValue.'</span> ('.$sKey.')</div>'."\n";
         }
         return $sHTML;
@@ -62,7 +61,7 @@
 
 
     function _one_row($aAddressLine){
-        $bNotUsed = (isset($aAddressLine['isaddress']) && $aAddressLine['isaddress'] == 'f');
+        $bNotUsed = isset($aAddressLine['isaddress']) && !$aAddressLine['isaddress'];
 
         echo '<tr class="' . ($bNotUsed?'notused':'') . '">'."\n";
         echo '  <td class="name">'.(trim($aAddressLine['localname'])?$aAddressLine['localname']:'<span class="noname">No Name</span>')."</td>\n";
@@ -97,7 +96,10 @@
     <div class="container">
         <div class="row">
             <div class="col-sm-10">
-                <h1><?php echo $aPointDetails['localname'] ?></h1>
+                <h1>
+                    <?php echo $aPointDetails['localname'] ?>
+                    <small><?php echo detailsPermaLink($aPointDetails, 'link to this page') ?></small>
+                </h1>
             </div>
             <div class="col-sm-2 text-right">
                 <?php map_icon($aPointDetails['icon']) ?>
@@ -111,13 +113,13 @@
 
                     kv('Name'            , hash_to_subtable($aPointDetails['aNames']) );
                     kv('Type'            , $aPointDetails['class'].':'.$aPointDetails['type'] );
-                    kv('Last Updated'    , $aPointDetails['indexed_date'] );
+                    kv('Last Updated'    , (new DateTime('@'.$aPointDetails['indexed_epoch']))->format(DateTime::RFC822) );
                     kv('Admin Level'     , $aPointDetails['admin_level'] );
                     kv('Rank'            , $aPointDetails['rank_search_label'] );
                     if ($aPointDetails['calculated_importance']) {
                         kv('Importance'    , $aPointDetails['calculated_importance'].($aPointDetails['importance']?'':' (estimated)') );
                     }
-                    kv('Coverage'        , ($aPointDetails['isarea']=='t'?'Polygon':'Point') );
+                    kv('Coverage'        , ($aPointDetails['isarea']?'Polygon':'Point') );
                     kv('Centre Point'    , $aPointDetails['lat'].','.$aPointDetails['lon'] );
                     kv('OSM'             , osmLink($aPointDetails) );
                     if ($aPointDetails['wikipedia'])
@@ -159,13 +161,10 @@
                 <tbody>
 
                 <?php
-
-                    foreach($aAddressLines as $aAddressLine)
-                    {   
+                    foreach ($aAddressLines as $aAddressLine) {
                         _one_row($aAddressLine);
                     }
                 ?>
-    
 
 
 <?php
@@ -173,39 +172,34 @@
     if ($aLinkedLines)
     {
         headline('Linked Places');
-        foreach($aLinkedLines as $aAddressLine)
-        {   
+        foreach ($aLinkedLines as $aAddressLine) {
             _one_row($aAddressLine);
         }
     }
 
-
-
-    if ($aPlaceSearchNameKeywords)
+    if ($bIncludeKeywords)
     {
         headline('Name Keywords');
-        foreach($aPlaceSearchNameKeywords as $aRow)
-        {
-            _one_keyword_row($aRow['word_token'], $aRow['word_id']);
+        if ($aPlaceSearchNameKeywords) {
+            foreach ($aPlaceSearchNameKeywords as $aRow) {
+                _one_keyword_row($aRow['word_token'], $aRow['word_id']);
+            }
+        }
+
+        headline('Address Keywords');
+        if ($aPlaceSearchAddressKeywords) {
+            foreach ($aPlaceSearchAddressKeywords as $aRow) {
+                _one_keyword_row($aRow['word_token'], $aRow['word_id']);
+            }
         }
     }
 
-    if ($aPlaceSearchAddressKeywords)
-    {
-        headline('Address Keywords');
-        foreach($aPlaceSearchAddressKeywords as $aRow)
-        {
-            _one_keyword_row($aRow['word_token'], $aRow['word_id']);
-        }
-    }
-    
-    if (sizeof($aParentOfLines))
+    if (!empty($aHierarchyLines))
     {
         headline('Parent Of');
 
         $aGroupedAddressLines = array();
-        foreach($aParentOfLines as $aAddressLine)
-        {
+        foreach ($aHierarchyLines as $aAddressLine) {
             if ($aAddressLine['type'] == 'yes') $sType = $aAddressLine['class'];
             else $sType = $aAddressLine['type'];
 
@@ -213,17 +207,15 @@
                 $aGroupedAddressLines[$sType] = array();
             $aGroupedAddressLines[$sType][] = $aAddressLine;
         }
-        foreach($aGroupedAddressLines as $sGroupHeading => $aParentOfLines)
-        {
+        foreach ($aGroupedAddressLines as $sGroupHeading => $aHierarchyLines) {
             $sGroupHeading = ucwords($sGroupHeading);
             headline3($sGroupHeading);
 
-            foreach($aParentOfLines as $aAddressLine)
-            {
+            foreach ($aHierarchyLines as $aAddressLine) {
                 _one_row($aAddressLine);
             }
         }
-        if (sizeof($aParentOfLines) >= 500) {
+        if (count($aHierarchyLines) >= 500) {
             echo '<p>There are more child objects which are not shown.</p>';
         }
     }
@@ -249,7 +241,7 @@
                 'lon' => $aPointDetails['lon'],
                 'lat' => $aPointDetails['lat'],
         );
-        echo 'var nominatim_result = ' . json_encode($aPlace, JSON_PRETTY_PRINT) . ';'; 
+        echo 'var nominatim_result = ' . json_encode($aPlace, JSON_PRETTY_PRINT) . ';';
 
 
     ?>

lib/template/details-json.php

@@ -0,0 +1,104 @@
+<?php
+
+$aPlaceDetails = array();
+
+$aPlaceDetails['place_id'] = (int) $aPointDetails['place_id'];
+$aPlaceDetails['parent_place_id'] = (int) $aPointDetails['parent_place_id'];
+
+$aPlaceDetails['osm_type'] = $aPointDetails['osm_type'];
+$aPlaceDetails['osm_id'] = (int) $aPointDetails['osm_id'];
+
+$aPlaceDetails['category'] = $aPointDetails['class'];
+$aPlaceDetails['type'] = $aPointDetails['type'];
+$aPlaceDetails['admin_level'] = $aPointDetails['admin_level'];
+
+$aPlaceDetails['localname'] = $aPointDetails['localname'];
+$aPlaceDetails['names'] = $aPointDetails['aNames'];
+
+$aPlaceDetails['addresstags'] = $aPointDetails['aAddressTags'];
+$aPlaceDetails['housenumber'] = $aPointDetails['housenumber'];
+$aPlaceDetails['calculated_postcode'] = $aPointDetails['postcode'];
+$aPlaceDetails['country_code'] = $aPointDetails['country_code'];
+
+$aPlaceDetails['indexed_date'] = (new DateTime('@'.$aPointDetails['indexed_epoch']))->format(DateTime::RFC3339);
+$aPlaceDetails['importance'] = (float) $aPointDetails['importance'];
+$aPlaceDetails['calculated_importance'] = (float) $aPointDetails['calculated_importance'];
+
+$aPlaceDetails['extratags'] = $aPointDetails['aExtraTags'];
+$aPlaceDetails['calculated_wikipedia'] = $aPointDetails['wikipedia'];
+if ($aPointDetails['icon']) {
+    $aPlaceDetails['icon'] = CONST_Website_BaseURL.'images/mapicons/'.$aPointDetails['icon'].'.n.32.png';
+}
+
+$aPlaceDetails['rank_address'] = (int) $aPointDetails['rank_address'];
+$aPlaceDetails['rank_search'] = (int) $aPointDetails['rank_search'];
+
+$aPlaceDetails['isarea'] = $aPointDetails['isarea'];
+$aPlaceDetails['centroid'] = array(
+                              'type' => 'Point',
+                              'coordinates' => array( (float) $aPointDetails['lon'], (float) $aPointDetails['lat'] )
+                             );
+
+$aPlaceDetails['geometry'] = json_decode($aPointDetails['asgeojson']);
+
+$funcMapAddressLine = function ($aFull) {
+    $aMapped = array(
+                'localname' => $aFull['localname'],
+                'place_id' => isset($aFull['place_id']) ? (int) $aFull['place_id'] : null,
+                'osm_id' => isset($aFull['osm_id']) ? (int) $aFull['osm_id'] : null,
+                'osm_type' => isset($aFull['osm_type']) ? $aFull['osm_type'] : null,
+                'class' => $aFull['class'],
+                'type' => $aFull['type'],
+                'admin_level' => isset($aFull['admin_level']) ? (int) $aFull['admin_level'] : null,
+                'rank_address' => $aFull['rank_address'] ? (int) $aFull['rank_address'] : null,
+                'distance' => (float) $aFull['distance']
+               );
+
+    return $aMapped;
+};
+
+$funcMapKeyword = function ($aFull) {
+    $aMapped = array(
+                'id' => (int) $aFull['word_id'],
+                'token' => $aFull['word_token']
+               );
+    return $aMapped;
+};
+
+if ($aAddressLines) {
+    $aPlaceDetails['address'] = array_map($funcMapAddressLine, $aAddressLines);
+}
+
+if ($aLinkedLines) {
+    $aPlaceDetails['linked_places'] = array_map($funcMapAddressLine, $aLinkedLines);
+}
+
+if ($bIncludeKeywords) {
+    $aPlaceDetails['keywords'] = array();
+
+    if ($aPlaceSearchNameKeywords) {
+        $aPlaceDetails['keywords']['name'] = array_map($funcMapKeyword, $aPlaceSearchNameKeywords);
+    }
+
+    if ($aPlaceSearchAddressKeywords) {
+        $aPlaceDetails['keywords']['address'] = array_map($funcMapKeyword, $aPlaceSearchAddressKeywords);
+    }
+}
+
+if ($bIncludeHierarchy) {
+    if ($bGroupHierarchy) {
+        $aPlaceDetails['hierarchy'] = array();
+        foreach ($aHierarchyLines as $aAddressLine) {
+            if ($aAddressLine['type'] == 'yes') $sType = $aAddressLine['class'];
+            else $sType = $aAddressLine['type'];
+
+            if (!isset($aPlaceDetails['hierarchy'][$sType]))
+                $aPlaceDetails['hierarchy'][$sType] = array();
+            $aPlaceDetails['hierarchy'][$sType][] = $funcMapAddressLine($aAddressLine);
+        }
+    } else {
+        $aPlaceDetails['hierarchy'] = array_map($funcMapAddressLine, $aHierarchyLines);
+    }
+}
+
+javascript_renderData($aPlaceDetails);

lib/template/error-html.php

@@ -0,0 +1,60 @@
+<?php
+
+    $title = 'Internal Server Error';
+    if ( $exception->getCode() == 400 ) {
+        $title = 'Bad Request';
+    }
+?>
+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <style>
+        em { font-weight: bold; font-family: monospace; color: #e00404; background-color: #ffeaea; }
+    </style>
+</head>
+<body>
+    <h1><?php echo $title ?></h1>
+    
+    <?php if (get_class($exception) == 'Nominatim\DatabaseError') { ?>
+
+        <p>Nominatim has encountered an internal error while accessing the database.
+           This may happen because the database is broken or because of a bug in
+           the software.</p>
+
+    <?php } else { ?>
+
+        <p>Nominatim has encountered an error with your request.</p>
+
+    <?php } ?>
+
+
+    <h3>Details</h3>
+
+    <?php echo $exception->getMessage() ?>
+
+    <?php if (CONST_Debug) { ?>
+        <p>
+        Exception <em><?php echo get_class($exception) ?></em> thrown in <em><?php echo $exception->getFile() . '('. $exception->getLine() . ')' ?></em>.
+
+        <?php if (get_class($exception) == 'Nominatim\DatabaseError') { ?>
+
+            <h3>SQL Error</h3>
+            <em><?php echo $exception->getSqlError() ?></em>
+
+            <pre><?php echo $exception->getSqlDebugDump() ?></pre>
+
+        <?php } ?>
+
+        <h3>Stack trace</h3>
+        <pre><?php echo $exception->getTraceAsString() ?></pre>
+
+    <?php } ?>
+
+    <p>
+        If you feel this error is incorrect feel file an issue on
+        <a href="https://github.com/openstreetmap/Nominatim/issues">Github</a>.
+
+        Please include the error message above and the URL you used.
+    </p>
+</body>
+</html>

lib/template/error-json.php

@@ -0,0 +1,11 @@
+<?php
+    $error = array(
+              'code' => $exception->getCode(),
+              'message' => $exception->getMessage()
+             );
+
+    if (CONST_Debug) {
+        $error['details'] = $exception->getFile() . '('. $exception->getLine() . ')';
+    }
+
+    echo javascript_renderData(array('error' => $error));

lib/template/error-xml.php

@@ -0,0 +1,7 @@
+<error>
+    <code><?php echo $exception->getCode() ?></code>
+    <message><?php echo $exception->getMessage() ?></message>
+    <?php if (CONST_Debug) { ?>
+    <details><?php echo $exception->getFile() . '('. $exception->getLine() . ')' ?></details>
+    <?php } ?>
+</error>

lib/template/includes/html-footer.php

@@ -11,4 +11,5 @@
 <script src="js/bootstrap.min.js"></script>
 <script src="js/leaflet.min.js"></script>
 <script src="js/Control.Minimap.min.js"></script>
+<script src="js/url-search-params.js"></script>
 <script src="js/nominatim-ui.js"></script>

lib/template/search-batch-json.php

@@ -1,42 +1,37 @@
 <?php
 
 $aOutput = array();
-$aOutput['licence'] = "Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright";
+$aOutput['licence'] = 'Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright';
 $aOutput['batch'] = array();
 
-foreach($aBatchResults as $aSearchResults)
-{
+foreach ($aBatchResults as $aSearchResults) {
     if (!$aSearchResults) $aSearchResults = array();
     $aFilteredPlaces = array();
-    foreach($aSearchResults as $iResNum => $aPointDetails)
-    {
+    foreach ($aSearchResults as $iResNum => $aPointDetails) {
         $aPlace = array(
-                    'place_id'=>$aPointDetails['place_id'],
-                );
+                   'place_id'=>$aPointDetails['place_id'],
+                  );
 
         $sOSMType = formatOSMType($aPointDetails['osm_type']);
-        if ($sOSMType)
-        {
+        if ($sOSMType) {
             $aPlace['osm_type'] = $sOSMType;
             $aPlace['osm_id'] = $aPointDetails['osm_id'];
         }
 
-        if (isset($aPointDetails['aBoundingBox']))
-        {
+        if (isset($aPointDetails['aBoundingBox'])) {
             $aPlace['boundingbox'] = array(
-                $aPointDetails['aBoundingBox'][0],
-                $aPointDetails['aBoundingBox'][1],
-                $aPointDetails['aBoundingBox'][2],
-                $aPointDetails['aBoundingBox'][3]);
+                                      $aPointDetails['aBoundingBox'][0],
+                                      $aPointDetails['aBoundingBox'][1],
+                                      $aPointDetails['aBoundingBox'][2],
+                $aPointDetails['aBoundingBox'][3]
+                                     );
 
-            if (isset($aPointDetails['aPolyPoints']) && $bShowPolygons)
-            {
+            if (isset($aPointDetails['aPolyPoints']) && $bShowPolygons) {
                 $aPlace['polygonpoints'] = $aPointDetails['aPolyPoints'];
             }
         }
 
-        if (isset($aPointDetails['zoom']))
-        {
+        if (isset($aPointDetails['zoom'])) {
             $aPlace['zoom'] = $aPointDetails['zoom'];
         }
 
@@ -50,33 +45,27 @@ foreach($aBatchResults as $aSearchResults)
 
         $aPlace['importance'] = $aPointDetails['importance'];
 
-        if (isset($aPointDetails['icon']))
-        {
+        if (isset($aPointDetails['icon'])) {
             $aPlace['icon'] = $aPointDetails['icon'];
         }
 
-        if (isset($aPointDetails['address']) && sizeof($aPointDetails['address'])>0)
-        {
-            $aPlace['address'] = $aPointDetails['address'];
+        if (isset($aPointDetails['address'])) {
+            $aPlace['address'] = $aPointDetails['address']->getAddressNames();
         }
 
-        if (isset($aPointDetails['asgeojson']))
-        {
+        if (isset($aPointDetails['asgeojson'])) {
             $aPlace['geojson'] = json_decode($aPointDetails['asgeojson']);
         }
 
-        if (isset($aPointDetails['assvg']))
-        {
+        if (isset($aPointDetails['assvg'])) {
             $aPlace['svg'] = $aPointDetails['assvg'];
         }
 
-        if (isset($aPointDetails['astext']))
-        {
+        if (isset($aPointDetails['astext'])) {
             $aPlace['geotext'] = $aPointDetails['astext'];
         }
 
-        if (isset($aPointDetails['askml']))
-        {
+        if (isset($aPointDetails['askml'])) {
             $aPlace['geokml'] = $aPointDetails['askml'];
         }
 

lib/template/search-geocodejson.php

@@ -0,0 +1,73 @@
+<?php
+
+$aFilteredPlaces = array();
+foreach ($aSearchResults as $iResNum => $aPointDetails) {
+    $aPlace = array(
+               'type' => 'Feature',
+               'properties' => array(
+                                'geocoding' => array()
+                               )
+              );
+
+    if (isset($aPointDetails['place_id'])) $aPlace['properties']['geocoding']['place_id'] = $aPointDetails['place_id'];
+    $sOSMType = formatOSMType($aPointDetails['osm_type']);
+    if ($sOSMType) {
+        $aPlace['properties']['geocoding']['osm_type'] = $sOSMType;
+        $aPlace['properties']['geocoding']['osm_id'] = $aPointDetails['osm_id'];
+    }
+
+    $aPlace['properties']['geocoding']['type'] = $aPointDetails['type'];
+
+    $aPlace['properties']['geocoding']['label'] = $aPointDetails['langaddress'];
+
+    $aPlace['properties']['geocoding']['name'] = $aPointDetails['placename'];
+
+    if (isset($aPointDetails['address'])) {
+        $aFieldMappings = array(
+                           'house_number' => 'housenumber',
+                           'road' => 'street',
+                           'locality' => 'locality',
+                           'postcode' => 'postcode',
+                           'city' => 'city',
+                           'district' => 'district',
+                           'county' => 'county',
+                           'state' => 'state',
+                           'country' => 'country'
+                          );
+
+        $aAddrNames = $aPointDetails['address']->getAddressNames();
+        foreach ($aFieldMappings as $sFrom => $sTo) {
+            if (isset($aAddrNames[$sFrom])) {
+                $aPlace['properties']['geocoding'][$sTo] = $aAddrNames[$sFrom];
+            }
+        }
+
+        $aPlace['properties']['geocoding']['admin']
+            = $aPointDetails['address']->getAdminLevels();
+    }
+
+    if (isset($aPointDetails['asgeojson'])) {
+        $aPlace['geometry'] = json_decode($aPointDetails['asgeojson']);
+    } else {
+        $aPlace['geometry'] = array(
+                               'type' => 'Point',
+                               'coordinates' => array(
+                                                 (float) $aPointDetails['lon'],
+                                                 (float) $aPointDetails['lat']
+                                                )
+                              );
+    }
+    $aFilteredPlaces[] = $aPlace;
+}
+
+
+javascript_renderData(array(
+                       'type' => 'FeatureCollection',
+                       'geocoding' => array(
+                                       'version' => '0.1.0',
+                                       'attribution' => 'Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright',
+                                       'licence' => 'ODbL',
+                                       'query' => $sQuery
+                                      ),
+                       'features' => $aFilteredPlaces
+                      ));

lib/template/search-geojson.php

@@ -0,0 +1,71 @@
+<?php
+
+$aFilteredPlaces = array();
+foreach ($aSearchResults as $iResNum => $aPointDetails) {
+    $aPlace = array(
+               'type' => 'Feature',
+               'properties' => array(
+                                'place_id'=>$aPointDetails['place_id'],
+                               )
+              );
+    
+    $sOSMType = formatOSMType($aPointDetails['osm_type']);
+    if ($sOSMType) {
+        $aPlace['properties']['osm_type'] = $sOSMType;
+        $aPlace['properties']['osm_id'] = $aPointDetails['osm_id'];
+    }
+
+    if (isset($aPointDetails['aBoundingBox'])) {
+        $aPlace['bbox'] = array(
+                           (float) $aPointDetails['aBoundingBox'][2], // minlon
+                           (float) $aPointDetails['aBoundingBox'][0], // minlat
+                           (float) $aPointDetails['aBoundingBox'][3], // maxlon
+                           (float) $aPointDetails['aBoundingBox'][1]  // maxlat
+                          );
+    }
+
+    if (isset($aPointDetails['zoom'])) {
+        $aPlace['properties']['zoom'] = $aPointDetails['zoom'];
+    }
+
+    $aPlace['properties']['display_name'] = $aPointDetails['name'];
+
+    $aPlace['properties']['place_rank'] = $aPointDetails['rank_search'];
+    $aPlace['properties']['category'] = $aPointDetails['class'];
+
+    $aPlace['properties']['type'] = $aPointDetails['type'];
+
+    $aPlace['properties']['importance'] = $aPointDetails['importance'];
+
+    if (isset($aPointDetails['icon']) && $aPointDetails['icon']) {
+        $aPlace['properties']['icon'] = $aPointDetails['icon'];
+    }
+
+    if (isset($aPointDetails['address'])) {
+        $aPlace['properties']['address'] = $aPointDetails['address']->getAddressNames();
+    }
+
+    if (isset($aPointDetails['asgeojson'])) {
+        $aPlace['geometry'] = json_decode($aPointDetails['asgeojson']);
+    } else {
+        $aPlace['geometry'] = array(
+                               'type' => 'Point',
+                               'coordinates' => array(
+                                                 (float) $aPointDetails['lon'],
+                                                 (float) $aPointDetails['lat']
+                                                )
+                              );
+    }
+
+
+    if (isset($aPointDetails['sExtraTags'])) $aPlace['properties']['extratags'] = $aPointDetails['sExtraTags'];
+    if (isset($aPointDetails['sNameDetails'])) $aPlace['properties']['namedetails'] = $aPointDetails['sNameDetails'];
+
+    $aFilteredPlaces[] = $aPlace;
+}
+
+javascript_renderData(array(
+                       'type' => 'FeatureCollection',
+                       'licence' => 'Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright',
+                       'features' => $aFilteredPlaces
+                      ));

lib/template/search-html.php

@@ -57,7 +57,7 @@
                 echo '</div>';
                 $i = $i+1;
             }
-            if (sizeof($aSearchResults) && $sMoreURL)
+            if (!empty($aSearchResults) && $sMoreURL)
             {
                 echo '<div class="more"><a class="btn btn-primary" href="'.htmlentities($sMoreURL).'">Search for more results</a></div>';
             }

lib/template/search-json.php

@@ -1,72 +1,66 @@
 <?php
-header("content-type: application/json; charset=UTF-8");
 
 $aFilteredPlaces = array();
-foreach($aSearchResults as $iResNum => $aPointDetails)
-{
+foreach ($aSearchResults as $iResNum => $aPointDetails) {
     $aPlace = array(
-            'place_id'=>$aPointDetails['place_id'],
-            'licence'=>"Data © OpenStreetMap contributors, ODbL 1.0. https://www.openstreetmap.org/copyright",
-        );
+               'place_id'=>$aPointDetails['place_id'],
+               'licence'=>'Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright',
+              );
     
     $sOSMType = formatOSMType($aPointDetails['osm_type']);
-    if ($sOSMType)
-    {
+    if ($sOSMType) {
         $aPlace['osm_type'] = $sOSMType;
         $aPlace['osm_id'] = $aPointDetails['osm_id'];
     }
 
-    if (isset($aPointDetails['aBoundingBox']))
-    {
+    if (isset($aPointDetails['aBoundingBox'])) {
         $aPlace['boundingbox'] = $aPointDetails['aBoundingBox'];
 
-        if (isset($aPointDetails['aPolyPoints']))
-        {
+        if (isset($aPointDetails['aPolyPoints'])) {
             $aPlace['polygonpoints'] = $aPointDetails['aPolyPoints'];
         }
     }
 
-    if (isset($aPointDetails['zoom']))
-    {
+    if (isset($aPointDetails['zoom'])) {
         $aPlace['zoom'] = $aPointDetails['zoom'];
     }
 
     $aPlace['lat'] = $aPointDetails['lat'];
     $aPlace['lon'] = $aPointDetails['lon'];
+
     $aPlace['display_name'] = $aPointDetails['name'];
 
-    $aPlace['class'] = $aPointDetails['class'];
+    if ($sOutputFormat == 'jsonv2' || $sOutputFormat == 'geojson') {
+        $aPlace['place_rank'] = $aPointDetails['rank_search'];
+        $aPlace['category'] = $aPointDetails['class'];
+    } else {
+        $aPlace['class'] = $aPointDetails['class'];
+    }
     $aPlace['type'] = $aPointDetails['type'];
 
     $aPlace['importance'] = $aPointDetails['importance'];
 
-    if (isset($aPointDetails['icon']) && $aPointDetails['icon'])
-    {
+    if (isset($aPointDetails['icon']) && $aPointDetails['icon']) {
         $aPlace['icon'] = $aPointDetails['icon'];
     }
 
-    if (isset($aPointDetails['address']))
-    {
-        $aPlace['address'] = $aPointDetails['address'];
+    if (isset($aPointDetails['address'])) {
+        $aPlace['address'] = $aPointDetails['address']->getAddressNames();
     }
 
-    if (isset($aPointDetails['asgeojson']))
-    {
+    if (isset($aPointDetails['asgeojson'])) {
         $aPlace['geojson'] = json_decode($aPointDetails['asgeojson']);
     }
 
-    if (isset($aPointDetails['assvg']))
-    {
+    if (isset($aPointDetails['assvg'])) {
         $aPlace['svg'] = $aPointDetails['assvg'];
     }
 
-    if (isset($aPointDetails['astext']))
-    {
+    if (isset($aPointDetails['astext'])) {
         $aPlace['geotext'] = $aPointDetails['astext'];
     }
 
-    if (isset($aPointDetails['askml']))
-    {
+    if (isset($aPointDetails['askml'])) {
         $aPlace['geokml'] = $aPointDetails['askml'];
     }
 

lib/template/search-jsonv2.php

@@ -1,79 +0,0 @@
-<?php
-
-$aFilteredPlaces = array();
-foreach($aSearchResults as $iResNum => $aPointDetails)
-{
-    $aPlace = array(
-            'place_id'=>$aPointDetails['place_id'],
-            'licence'=>"Data © OpenStreetMap contributors, ODbL 1.0. https://osm.org/copyright",
-        );
-
-    $sOSMType = formatOSMType($aPointDetails['osm_type']);
-    if ($sOSMType)
-    {
-        $aPlace['osm_type'] = $sOSMType;
-        $aPlace['osm_id'] = $aPointDetails['osm_id'];
-    }
-
-    if (isset($aPointDetails['aBoundingBox']))
-    {
-        $aPlace['boundingbox'] = $aPointDetails['aBoundingBox'];
-
-        if (isset($aPointDetails['aPolyPoints']))
-        {
-            $aPlace['polygonpoints'] = $aPointDetails['aPolyPoints'];
-        }
-    }
-
-    if (isset($aPointDetails['zoom']))
-    {
-        $aPlace['zoom'] = $aPointDetails['zoom'];
-    }
-
-    $aPlace['lat'] = $aPointDetails['lat'];
-    $aPlace['lon'] = $aPointDetails['lon'];
-    $aPlace['display_name'] = $aPointDetails['name'];
-    $aPlace['place_rank'] = $aPointDetails['rank_search'];
-
-    $aPlace['category'] = $aPointDetails['class'];
-    $aPlace['type'] = $aPointDetails['type'];
-
-    $aPlace['importance'] = $aPointDetails['importance'];
-
-    if (isset($aPointDetails['icon']))
-    {
-        $aPlace['icon'] = $aPointDetails['icon'];
-    }
-
-    if (isset($aPointDetails['address']) && sizeof($aPointDetails['address'])>0)
-    {
-        $aPlace['address'] = $aPointDetails['address'];
-    }
-
-    if (isset($aPointDetails['asgeojson']))
-    {
-        $aPlace['geojson'] = json_decode($aPointDetails['asgeojson']);
-    }
-
-    if (isset($aPointDetails['assvg']))
-    {
-        $aPlace['svg'] = $aPointDetails['assvg'];
-    }
-
-    if (isset($aPointDetails['astext']))
-    {
-        $aPlace['geotext'] = $aPointDetails['astext'];
-    }
-
-    if (isset($aPointDetails['askml']))
-    {
-        $aPlace['geokml'] = $aPointDetails['askml'];
-    }
-
-    if (isset($aPointDetails['sExtraTags'])) $aPlace['extratags'] = $aPointDetails['sExtraTags'];
-    if (isset($aPointDetails['sNameDetails'])) $aPlace['namedetails'] = $aPointDetails['sNameDetails'];
-
-    $aFilteredPlaces[] = $aPlace;
-}
-
-javascript_renderData($aFilteredPlaces);

lib/template/search-xml.php

@@ -1,72 +1,63 @@
 <?php
-header("content-type: text/xml; charset=UTF-8");
+header('content-type: text/xml; charset=UTF-8');
 
-echo "<";
-echo "?xml version=\"1.0\" encoding=\"UTF-8\" ?";
+echo '<';
+echo '?xml version="1.0" encoding="UTF-8" ?';
 echo ">\n";
 
-echo "<";
+echo '<';
 echo (isset($sXmlRootTag)?$sXmlRootTag:'searchresults');
 echo " timestamp='".date(DATE_RFC822)."'";
 echo " attribution='Data © OpenStreetMap contributors, ODbL 1.0. http://www.openstreetmap.org/copyright'";
 echo " querystring='".htmlspecialchars($sQuery, ENT_QUOTES)."'";
 if (isset($aMoreParams['viewbox'])) echo " viewbox='".htmlspecialchars($aMoreParams['viewbox'], ENT_QUOTES)."'";
 echo " polygon='".(isset($aMoreParams['polygon'])?'true':'false')."'";
-if (isset($aMoreParams['exclude_place_ids']))
-{
+if (isset($aMoreParams['exclude_place_ids'])) {
     echo " exclude_place_ids='".htmlspecialchars($aMoreParams['exclude_place_ids'])."'";
 }
 echo " more_url='".htmlspecialchars($sMoreURL)."'";
 echo ">\n";
 
-foreach($aSearchResults as $iResNum => $aResult)
-{
+foreach ($aSearchResults as $iResNum => $aResult) {
     echo "<place place_id='".$aResult['place_id']."'";
     $sOSMType = formatOSMType($aResult['osm_type']);
-    if ($sOSMType)
-    {
+    if ($sOSMType) {
         echo " osm_type='$sOSMType'";
         echo " osm_id='".$aResult['osm_id']."'";
     }
     echo " place_rank='".$aResult['rank_search']."'";
 
-    if (isset($aResult['aBoundingBox']))
-    {
+    if (isset($aResult['aBoundingBox'])) {
         echo ' boundingbox="';
-        echo join(',',$aResult['aBoundingBox']);
+        echo join(',', $aResult['aBoundingBox']);
         echo '"';
 
-        if (isset($aResult['aPolyPoints']))
-        {
+        if (isset($aResult['aPolyPoints'])) {
             echo ' polygonpoints=\'';
             echo json_encode($aResult['aPolyPoints']);
             echo '\'';
         }
     }
 
-    if (isset($aResult['asgeojson']))
-    {
+    if (isset($aResult['asgeojson'])) {
         echo ' geojson=\'';
         echo $aResult['asgeojson'];
         echo '\'';
     }
 
-    if (isset($aResult['assvg']))
-    {
+    if (isset($aResult['assvg'])) {
         echo ' geosvg=\'';
         echo $aResult['assvg'];
         echo '\'';
     }
 
-    if (isset($aResult['astext']))
-    {
+    if (isset($aResult['astext'])) {
         echo ' geotext=\'';
         echo $aResult['astext'];
         echo '\'';
     }
 
-    if (isset($aResult['zoom']))
-    {
+    if (isset($aResult['zoom'])) {
         echo " zoom='".$aResult['zoom']."'";
     }
 
@@ -77,82 +68,67 @@ foreach($aSearchResults as $iResNum => $aResult)
     echo " class='".htmlspecialchars($aResult['class'])."'";
     echo " type='".htmlspecialchars($aResult['type'], ENT_QUOTES)."'";
     echo " importance='".htmlspecialchars($aResult['importance'])."'";
-    if (isset($aResult['icon']) && $aResult['icon'])
-    {
+    if (isset($aResult['icon']) && $aResult['icon']) {
         echo " icon='".htmlspecialchars($aResult['icon'], ENT_QUOTES)."'";
     }
 
     $bHasDelim = false;
 
-    if (isset($aResult['askml']))
-    {
-        if (!$bHasDelim)
-        {
+    if (isset($aResult['askml'])) {
+        if (!$bHasDelim) {
             $bHasDelim = true;
-            echo ">";
+            echo '>';
         }
         echo "\n<geokml>";
         echo $aResult['askml'];
-        echo "</geokml>";
+        echo '</geokml>';
     }
 
-    if (isset($aResult['sExtraTags']))
-    {
-        if (!$bHasDelim)
-        {
+    if (isset($aResult['sExtraTags'])) {
+        if (!$bHasDelim) {
             $bHasDelim = true;
-            echo ">";
+            echo '>';
         }
         echo "\n<extratags>";
-        foreach ($aResult['sExtraTags'] as $sKey => $sValue)
-        {
+        foreach ($aResult['sExtraTags'] as $sKey => $sValue) {
             echo '<tag key="'.htmlspecialchars($sKey).'" value="'.htmlspecialchars($sValue).'"/>';
         }
-        echo "</extratags>";
+        echo '</extratags>';
     }
 
-    if (isset($aResult['sNameDetails']))
-    {
-        if (!$bHasDelim)
-        {
+    if (isset($aResult['sNameDetails'])) {
+        if (!$bHasDelim) {
             $bHasDelim = true;
-            echo ">";
+            echo '>';
         }
         echo "\n<namedetails>";
-        foreach ($aResult['sNameDetails'] as $sKey => $sValue)
-        {
+        foreach ($aResult['sNameDetails'] as $sKey => $sValue) {
             echo '<name desc="'.htmlspecialchars($sKey).'">';
             echo htmlspecialchars($sValue);
-            echo "</name>";
+            echo '</name>';
         }
-        echo "</namedetails>";
+        echo '</namedetails>';
     }
 
-    if (isset($aResult['address']))
-    {
-        if (!$bHasDelim)
-        {
+    if (isset($aResult['address'])) {
+        if (!$bHasDelim) {
             $bHasDelim = true;
-            echo ">";
+            echo '>';
         }
         echo "\n";
-        foreach($aResult['address'] as $sKey => $sValue)
-        {
-            $sKey = str_replace(' ','_',$sKey);
+        foreach ($aResult['address']->getAddressNames() as $sKey => $sValue) {
+            $sKey = str_replace(' ', '_', $sKey);
             echo "<$sKey>";
             echo htmlspecialchars($sValue);
             echo "</$sKey>";
         }
     }
 
-    if ($bHasDelim)
-    {
-        echo "</place>";
-    }
-    else
-    {
-        echo "/>";
+    if ($bHasDelim) {
+        echo '</place>';
+    } else {
+        echo '/>';
     }
 }
 
-echo "</" . (isset($sXmlRootTag)?$sXmlRootTag:'searchresults') . ">";
+echo '</' . (isset($sXmlRootTag)?$sXmlRootTag:'searchresults') . '>';

module/nominatim.c

@@ -157,17 +157,18 @@ transliteration( PG_FUNCTION_ARGS )
 	PG_RETURN_TEXT_P(result);
 }
 
+// Set isspace=1 if the replacement _only_ adds a space before the search string.  I.e. to == " " + from
 void str_replace(char* buffer, int* len, int* changes, char* from, int fromlen, char* to, int tolen, int isspace)
 {
         char *p;
 
-        // Search string is too long to be pressent
+        // Search string is too long to be present
         if (fromlen > *len) return;
 
         p = strstr(buffer, from);
         while(p)
         {
-                if (!isspace || *(p-1) != ' ')
+                if (!isspace || (p > buffer && *(p-1) != ' '))
                 {
                         (*changes)++;
                         if (tolen != fromlen) memmove(p+tolen, p+fromlen, *len-(p-buffer)+1);
@@ -230,7 +231,7 @@ gettokenstring( PG_FUNCTION_ARGS )
 	sourcedata = (unsigned char *)VARDATA(source);
 	sourcedatalength = VARSIZE(source) - VARHDRSZ;
 
-	// Buffer for doing the replace in - string could get slightly longer (double is mastive overkill)
+	// Buffer for doing the replace in - string could get slightly longer (double is massive overkill)
 	buffer = (char *)palloc((sourcedatalength*2)*sizeof(char));
 	memcpy(buffer+1, sourcedata, sourcedatalength);
 	buffer[0] = 32;

nominatim/CMakeLists.txt

@@ -1,12 +1,12 @@
 add_executable(nominatim export.c geometry.cpp import.c index.c input.c nominatim.c postgresql.c sprompt.c)
-include(CheckIncludeFile)
-CHECK_INCLUDE_FILE(byteswap.h HAVE_BYTESWAP_H)
-CHECK_INCLUDE_FILE(sys/endian.h HAVE_SYS_ENDIAN_H)
-if(HAVE_BYTESWAP_H)
-  target_compile_definitions(nominatim PRIVATE HAVE_BYTESWAP_H)
-endif(HAVE_BYTESWAP_H)
-if(HAVE_SYS_ENDIAN_H)
-  target_compile_definitions(nominatim PRIVATE HAVE_SYS_ENDIAN_H)
-endif(HAVE_SYS_ENDIAN_H)
+
+CHECK_SYMBOL_EXISTS(bswap_32 "byteswap.h" HAVE_BYTESWAP)
+CHECK_SYMBOL_EXISTS(bswap32 "sys/endian.h" HAVE_SYS_ENDIAN)
+
+target_compile_definitions(nominatim
+    PRIVATE HAVE_BYTESWAP=$<BOOL:${HAVE_BYTESWAP}>
+    PRIVATE HAVE_SYS_ENDIAN=$<BOOL:${HAVE_SYS_ENDIAN}>
+)
+
 target_link_libraries(nominatim ${LIBXML2_LIBRARIES} ${ZLIB_LIBRARIES} ${BZIP2_LIBRARIES} ${PostgreSQL_LIBRARIES} ${CMAKE_THREAD_LIBS_INIT})
 

nominatim/index.c

@@ -262,13 +262,16 @@ struct index_thread_data * thread_data, const char *structuredoutputfile)
 
 void nominatim_index(int rank_min, int rank_max, int num_threads, const char *conninfo, const char *structuredoutputfile)
 {
-    struct index_thread_data * thread_data;
+    struct index_thread_data *thread_data;
 
     PGconn *conn;
-    PGresult * res;
+    PGresult *res;
+    int num_rows = 0, status_code = 0;
+    int db_has_locale = 0;
+    char *result_string = NULL;
 
     int rank;
-    
+
     int i;
 
     xmlTextWriterPtr writer;
@@ -283,6 +286,23 @@ void nominatim_index(int rank_min, int rank_max, int num_threads, const char *co
         exit(EXIT_FAILURE);
     }
 
+    res = PQexec(conn, "SHOW lc_messages");
+    status_code = PQresultStatus(res);
+    if (status_code != PGRES_TUPLES_OK && status_code != PGRES_SINGLE_TUPLE) {
+        fprintf(stderr, "Failed determining database locale: %s\n", PQerrorMessage(conn));
+        exit(EXIT_FAILURE);
+    }
+    num_rows = PQntuples(res);
+    if (num_rows > 0)
+    {
+        result_string = PQgetvalue(res, 0, 0);
+        if (result_string && (strlen(result_string) > 0) && (strcasecmp(result_string, "C") != 0))
+        {
+            // non-default locale if the result exists, is non-empty, and is not "C"
+            db_has_locale = 1;
+        }
+    }
+
     pg_prepare_params[0] = PG_OID_INT4;
     res = PQprepare(conn, "index_sectors",
                     "select geometry_sector,count(*) from placex where rank_search = $1 and indexed_status > 0 group by geometry_sector order by geometry_sector",
@@ -392,19 +412,20 @@ void nominatim_index(int rank_min, int rank_max, int num_threads, const char *co
         }
         PQclear(res);
 
-        // Make sure the error message is not localized as we parse it later.
-        res = PQexec(thread_data[i].conn, "SET lc_messages TO 'C'");
-        if (PQresultStatus(res) != PGRES_COMMAND_OK)
+        if (db_has_locale)
         {
-            fprintf(stderr, "Failed to set langauge: %s\n", PQerrorMessage(thread_data[i].conn));
-            exit(EXIT_FAILURE);
+            // Make sure the error message is not localized as we parse it later.
+            res = PQexec(thread_data[i].conn, "SET lc_messages TO 'C'");
+            if (PQresultStatus(res) != PGRES_COMMAND_OK)
+            {
+                fprintf(stderr, "Failed to set langauge: %s\n", PQerrorMessage(thread_data[i].conn));
+                exit(EXIT_FAILURE);
+            }
+            PQclear(res);
         }
-        PQclear(res);
-
         nominatim_exportCreatePreparedQueries(thread_data[i].conn);
     }
 
-
     fprintf(stderr, "Starting indexing rank (%i to %i) using %i threads\n", rank_min, rank_max, num_threads);
 
     for (rank = rank_min; rank <= rank_max; rank++)

nominatim/postgresql.h

@@ -7,20 +7,32 @@
 #define PG_OID_INT8			20
 #define PG_OID_INT4			23
 
-#if defined(HAVE_BYTESWAP_H)
+#if HAVE_BYTESWAP
 #include <byteswap.h>
-#elif defined(HAVE_SYS_ENDIAN_H)
+#define PG_BSWAP32(x) bswap_32(x)
+#define PG_BSWAP64(x) bswap_64(x)
+#elif HAVE_SYS_ENDIAN
 #include <sys/endian.h>
+#define PG_BSWAP32(x) bswap32(x)
+#define PG_BSWAP64(x) bswap64(x)
+#else
+#error "No appropriate byteswap found for your system."
 #endif
 
-#if __BYTE_ORDER == __BIG_ENDIAN
-#define PGint16(x)	(x)
+#if defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)
 #define PGint32(x)	(x)
 #define PGint64(x)	(x)
+#elif defined(__BYTE_ORDER__) &&  (__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)
+#define PGint32(x)	PG_BSWAP32(x)
+#define PGint64(x)	PG_BSWAP64(x)
+#elif defined(_BYTE_ORDER) && (_BYTE_ORDER == _BIG_ENDIAN)
+#define PGint32(x)	(x)
+#define PGint64(x)	(x)
+#elif defined(_BYTE_ORDER) &&  (_BYTE_ORDER == _LITTLE_ENDIAN)
+#define PGint32(x)	PG_BSWAP32(x)
+#define PGint64(x)	PG_BSWAP64(x)
 #else
-#define PGint16(x)	__bswap_16 (x)
-#define PGint32(x)	__bswap_32 (x)
-#define PGint64(x)	__bswap_64 (x)
+#error "Cannot determine byte order."
 #endif
 
 const char *build_conninfo(const char *db, const char *username, const char *password, const char *host, const char *port);

phpcs.xml

@@ -8,6 +8,12 @@
   <!-- https://github.com/squizlabs/PHP_CodeSniffer/blob/master/CodeSniffer/Standards/PSR2/ruleset.xml -->
   <rule ref="PSR2"/>
 
+  <exclude-pattern>./lib/template/*html*</exclude-pattern>
+  <exclude-pattern>./lib/template/includes/</exclude-pattern>
+  <exclude-pattern>./module/</exclude-pattern>
+  <exclude-pattern>./website/css</exclude-pattern>
+  <exclude-pattern>./website/js</exclude-pattern>
+
   <rule ref="Generic.Files.LineLength">
     <properties>
      <property name="lineLimit" value="194"/>
@@ -26,6 +32,13 @@
 
 
 
+  <!-- eval, system, etc -->
+  <rule ref="Generic.PHP.ForbiddenFunctions">
+    <properties>
+      <property name="forbiddenFunctions" type="array" value="sizeof=>count,delete=>unset,print=>echo,create_function=>null,eval=>null"/>
+    </properties>
+  </rule>
+
   <!-- **************************************************************
        DOCUMENTATION
        ************************************************************** -->
@@ -80,18 +93,7 @@
        INDENTATION, SPACING
        ************************************************************** -->
 
-  <!-- We don't need 2 blank lines after function -->
-  <rule ref="Squiz.WhiteSpace.FunctionSpacing.After">
-    <severity>0</severity>
-  </rule>
-
-  <!-- Aligned looks nicer, but causes too many warnings currently -->
-  <rule ref="Generic.Formatting.MultipleStatementAlignment.NotSame">
-    <severity>0</severity>
-  </rule>
-  <rule ref="Generic.Formatting.MultipleStatementAlignment.NotSameWarning">
-    <severity>0</severity>
-  </rule>
+  <rule ref="Squiz.Arrays.ArrayDeclaration.KeyNotAligned" />
 
   <!-- Aligned looks nicer, but causes too many warnings currently -->
   <rule ref="Squiz.Arrays.ArrayDeclaration.DoubleArrowNotAligned">
@@ -103,7 +105,6 @@
 
 
 
-
   <!-- **************************************************************
        VARIABLES
        ************************************************************** -->
@@ -126,6 +127,8 @@
     <severity>0</severity>
   </rule>
 
+  <!-- array() instead of [] for initialisation -->
+  <rule ref="Generic.Arrays.DisallowShortArraySyntax.Found" />
 
 
 

settings/address-levels.json

@@ -0,0 +1,85 @@
+[
+{ "tags" : {
+      "place" : {
+          "sea" : [2, 0],
+          "continent" : [2, 0],
+          "country" : [4, 4],
+          "state" : [8, 8],
+          "region" : [18, 0],
+          "county" : 12,
+          "city" : 16,
+          "island" : [17, 0],
+          "town" : [18, 16],
+          "village" : [19, 16],
+          "hamlet" : [19, 16],
+          "municipality" : [19, 16],
+          "district" : [19, 16],
+          "unincorporated_area" : [19, 16],
+          "borough" : [19, 16],
+          "suburb" : 20,
+          "croft" : 20,
+          "subdivision" : 20,
+          "isolated_dwelling" : 20,
+          "farm" : [20, 0],
+          "locality" : [20, 0],
+          "islet" : [20, 0],
+          "mountain_pass" : [20, 0],
+          "neighbourhood" : 22,
+          "houses" : [28, 0]
+      },
+      "boundary" : {
+          "administrative2" : 4,
+          "administrative3" : 6,
+          "administrative4" : 8,
+          "administrative5" : 10,
+          "administrative6" : 12,
+          "administrative7" : 14,
+          "administrative8" : 16,
+          "administrative9" : 18,
+          "administrative10" : 20,
+          "administrative11" : 22,
+          "administrative12" : 24
+      },
+      "landuse" : {
+          "residential" : 22,
+          "farm" : 22,
+          "farmyard" : 22,
+          "industrial" : 22,
+          "commercial" : 22,
+          "allotments" : 22,
+          "retail" : 22,
+          "" : [22, 0]
+      },
+      "leisure" : {
+          "park" : [24, 0]
+      },
+      "natural" : {
+          "peak" : [18, 0],
+          "volcano" : [18, 0],
+          "mountain_range" : [18, 0],
+          "sea" : [4, 0]
+      },
+      "waterway" : {
+          "" : [17, 0]
+      },
+      "highway" : {
+          "" : 26,
+          "service" : 27,
+          "cycleway" : 27,
+          "path" : 27,
+          "footway" : 27,
+          "steps" : 27,
+          "bridleway" : 27,
+          "motorway_link" : 27,
+          "primary_link" : 27,
+          "trunk_link" : 27,
+          "secondary_link" : 27,
+          "tertiary_link" : 27
+      },
+      "mountain_pass" : {
+          "" : [20, 0]
+      }
+  }
+}
+]
+

settings/defaults.php

@@ -7,8 +7,9 @@ if (isset($_GET['debug']) && $_GET['debug']) @define('CONST_Debug', true);
 
 // General settings
 @define('CONST_Debug', false);
-@define('CONST_Database_DSN', 'pgsql://@/nominatim'); // <driver>://<username>:<password>@<host>:<port>/<database>
+@define('CONST_Database_DSN', 'pgsql:dbname=nominatim'); // or add ;host=...;port=...;user=...;password=...
 @define('CONST_Database_Web_User', 'www-data');
+@define('CONST_Database_Module_Path', CONST_InstallPath.'/module');
 @define('CONST_Max_Word_Frequency', '50000');
 @define('CONST_Limit_Reindexing', true);
 // Restrict search languages.
@@ -20,7 +21,7 @@ if (isset($_GET['debug']) && $_GET['debug']) @define('CONST_Debug', true);
 // Rules for normalizing terms for comparison before doing comparisons.
 // The default is to remove accents and punctuation and to lower-case the
 // term. Spaces are kept but collapsed to one standard space.
-@define('CONST_Term_Normalization_Rules', ":: NFD (); [:Nonspacing Mark:] >;  :: lower (); [[:Punctuation:][:Space:]]+ > ' '; :: NFC ();");
+@define('CONST_Term_Normalization_Rules', ":: NFD (); [[:Nonspacing Mark:] [:Cf:]] >;  :: lower (); [[:Punctuation:][:Space:]]+ > ' '; :: NFC ();");
 
 // Set to false to avoid importing extra postcodes for the US.
 @define('CONST_Use_Extra_US_Postcodes', true);
@@ -48,6 +49,9 @@ if (isset($_GET['debug']) && $_GET['debug']) @define('CONST_Debug', true);
 @define('CONST_Pyosmium_Binary', '@PYOSMIUM_PATH@');
 @define('CONST_Tiger_Data_Path', CONST_ExtraDataPath.'/tiger');
 @define('CONST_Wikipedia_Data_Path', CONST_ExtraDataPath);
+@define('CONST_Phrase_Config', CONST_BasePath.'/settings/phrase_settings.php');
+@define('CONST_Address_Level_Config', CONST_BasePath.'/settings/address-levels.json');
+@define('CONST_Import_Style', CONST_BasePath.'/settings/import-full.style');
 
 // osm2pgsql settings
 @define('CONST_Osm2pgsql_Flatnode_File', null);

settings/import-address.style

@@ -0,0 +1,116 @@
+[
+{
+    "keys" : [ "" ],
+    "values" : {
+        "no" : "skip"
+    }
+},
+{
+    "keys" : ["name:prefix", "name:suffix", "name:botanical", "*wikidata"],
+    "values" : {
+        "" : "skip"
+    }
+},
+{
+    "keys" : ["ref", "int_ref", "nat_ref", "reg_ref", "loc_ref", "old_ref",
+              "iata", "icao", "pcode"],
+    "values" : {
+        "" : "ref"
+    }
+},
+{
+    "keys" : ["name", "name:*", "int_name", "int_name:*", "nat_name", "nat_name:*",
+              "reg_name", "reg_name:*", "loc_name", "loc_name:*",
+              "old_name", "old_name:*", "alt_name", "alt_name:*", "alt_name_*",
+              "official_name", "official_name:*", "place_name", "place_name:*",
+              "short_name", "short_name:*", "brand"],
+    "values" : {
+        "" : "name"
+    }
+},
+{
+    "keys" : ["landuse"],
+    "values" : {
+        "cemetry" : "skip",
+        "" : "fallback,with_name"
+    }
+},
+{
+    "keys" : ["boundary"],
+    "values" : {
+        "administrative" : "main"
+    }
+},
+{
+    "keys" : ["place"],
+    "values" : {
+        "" : "main"
+    }
+},
+{
+    "keys" : ["addr:housename"],
+    "values" : {
+        "" : "name,house"
+    }
+},
+{
+    "keys" : ["addr:housenumber", "addr:conscriptionnumber", "addr:streetnumber"],
+    "values" : {
+        "" : "address,house"
+    }
+},
+{
+    "keys" : ["addr:interpolation"],
+    "values" : {
+        "" : "interpolation,address"
+    }
+},
+{
+    "keys" : ["postal_code", "postcode", "addr:postcode",
+              "tiger:zip_left", "tiger:zip_right"],
+    "values" : {
+        "" : "postcode,fallback"
+    }
+},
+{
+    "keys" : ["country_code", "ISO3166-1", "is_in:country_code", "is_in_country",
+              "addr:country", "addr:country", "addr:country_code"],
+    "values" : {
+        "" : "country"
+    }
+},
+{
+    "keys" : ["addr:*", "is_in:*", "tiger:county"],
+    "values" : {
+        "" : "address"
+    }
+},
+{
+    "keys" : ["highway"],
+    "values" : {
+        "motorway" : "main",
+        "trunk" : "main",
+        "primary" : "main",
+        "secondary" : "main",
+        "tertiary" : "main",
+        "unclassified" : "main",
+        "residential" : "main",
+        "living_street" : "main",
+        "pedestrian" : "main",
+        "road" : "main",
+        "service" : "main,with_name",
+        "cycleway" : "main,with_name",
+        "path" : "main,with_name",
+        "footway" : "main,with_name",
+        "steps" : "main,with_name",
+        "bridleway" : "main,with_name",
+        "track" : "main,with_name",
+        "byway": "main,with_name",
+        "motorway_link" : "main,with_name",
+        "trunk_link" : "main,with_name",
+        "primary_link" : "main,with_name",
+        "secondary_link" : "main,with_name",
+        "tertiary_link" : "main,with_name"
+    }
+}
+]

settings/import-admin.style

@@ -0,0 +1,70 @@
+[
+{
+    "keys" : ["name:prefix", "name:suffix", "name:botanical", "*wikidata"],
+    "values" : {
+        "" : "skip"
+    }
+},
+{
+    "keys" : ["ref", "int_ref", "nat_ref", "reg_ref", "loc_ref", "old_ref",
+              "iata", "icao", "pcode"],
+    "values" : {
+        "" : "ref"
+    }
+},
+{
+    "keys" : ["name", "name:*", "int_name", "int_name:*", "nat_name", "nat_name:*",
+              "reg_name", "reg_name:*", "loc_name", "loc_name:*",
+              "old_name", "old_name:*", "alt_name", "alt_name:*", "alt_name_*",
+              "official_name", "official_name:*", "place_name", "place_name:*",
+              "short_name", "short_name:*", "brand"],
+    "values" : {
+        "" : "name"
+    }
+},
+{
+    "keys" : ["landuse"],
+    "values" : {
+        "cemetry" : "skip",
+        "" : "fallback,with_name"
+    }
+},
+{
+    "keys" : ["boundary"],
+    "values" : {
+        "administrative" : "main"
+    }
+},
+{
+    "keys" : ["place"],
+    "values" : {
+        "" : "main"
+    }
+},
+{
+    "keys" : ["country_code", "ISO3166-1", "is_in:country_code", "is_in_country",
+              "addr:country", "addr:country", "addr:country_code"],
+    "values" : {
+        "" : "country"
+    }
+},
+{
+    "keys" : ["addr:*", "is_in:*", "tiger:county"],
+    "values" : {
+        "" : "address"
+    }
+},
+{
+    "keys" : ["postal_code", "postcode", "addr:postcode",
+              "tiger:zip_left", "tiger:zip_right"],
+    "values" : {
+        "" : "postcode"
+    }
+},
+{
+    "keys" : ["capital"],
+    "values" : {
+        "" : "extra"
+    }
+}
+]