2 * replace datrie with simple internal trie implementation
3 * add pattern-based postcode parser for queries,
4 postcodes no longer need to be present in OSM to be found
5 * take variants into account when computing token similarity
6 * add extratags output to geocodejson format
7 * fix default layer setting used for structured queries
8 * update abbreviation lists for Russian and English
9 (thanks @shoorick, @IvanShift, @mhsrn21)
10 * fix variant generation for Norwegian
11 * fix normalization around space-like characters
12 * improve postcode search and handling of postcodes in queries
13 * reorganise internal query structure and get rid of slow enums
14 * enable code linting for tests
15 * various code moderinsations in test code (thanks @eumiro)
16 * remove setting osm2pgsql location via config.lib_dir
17 * make SQL functions parallel save as far as possible (thanks @otbutz)
18 * various fixes and improvements to documentation (thanks @TuringVerified)
21 * increase required versions for PostgreSQL (12+), PostGIS (3.0+)
22 * remove installation via cmake and debundle osm2pgsql
23 * remove deprecated PHP frontend
24 * remove deprecated legacy tokenizer
25 * add configurable pre-processing of queries
26 * add query pre-processor to split up Japanese addresses
27 * rewrite of osm2pgsql style implementation
28 (also adds support for osm2pgsql-themepark)
29 * reduce the number of SQL queries needed to complete a 'lookup' call
30 * improve computation of centroid for lines with only two points
31 * improve bbox output for postcode areas
32 * improve result order by returning the largest object when other things are
34 * add fallback for reverse geocoding to default country tables
35 * exclude postcode areas from reverse geocoding
36 * disable search endpoint when database is reverse-only (regression)
37 * minor performance improvements to area split algorithm
38 * switch table and index creation to use autocommit mode to avoid deadlocks
39 * drop overly long ways during import
40 * restrict automatic migrations to versions 4.3+
41 * switch linting from pylint to flake8
42 * switch tests to use a wikimedia test file in the new CSV style
43 * various fixes and improvements to documentation
46 * allow building Nominatim as a pip package
47 * make osm2pgsql building optional
48 * switch importer to psycopg3
49 * allow output format of web search to be customized in self-installations
50 * look up potential postcode areas for postcode results
51 * add word usage statistics for address terms
52 * implement more light-weight CSV format for wiki importance tables
53 * rewrite SQL for place search to use window functions
54 * increase search radius when filtering by postcode
55 * prefer POI points over POI areas
56 * reintroduce full terms for address terms in search_name table
57 * reindex postcodes when their parent is deleted
58 * indexing: precompute counts of affected rows
59 * ensure consistent country assignments for overlapping countries
60 * make Nominatim[Async]API context manager to ensure proper calling of
62 * make usage of project dir optional for library
63 * drop interpolations when no parent can be found
64 * style tweaks to reflect OSM usage (man_made, highway and others)
65 * deprecation of: bundled osm2pgsql, legacy tokenizer, PHP frontend
66 * make documentation buildable without CMake
67 * various fixes and improvements to documentation
70 * fix geocodejson output: admin level output should only print boundaries
71 * updating: restrict invalidation of child objects on large street features
72 * restrict valid interpolation house numbers to 0-999999
73 * fix import error when SQLAlchemy 1.4 and psycopg3 are installed
74 * various typo fixes in the documentation
77 * add export to SQLite database and SQLite support for the frontend
78 * switch to Python frontend as the default frontend
79 * update to osm2pgsql 1.11.0
80 * add support for new osm2pgsql middle table format
81 * simplify geometry for large polygon objects not used in addresses
82 * various performance tweaks for search in Python frontend
83 * fix regression in search with categories where it was confused with near
85 * partially roll back use of SQLAlchemy lambda statements due to bugs
87 * fix handling of timezones for timestamps from the database
88 * fix handling of full address searches in connection with a viewbox
89 * fix postcode computation of highway areas
90 * fix handling of timeout errors for Python <= 3.10
91 * fix address computation for postcode areas
92 * fix variable shadowing in osm2pgsql flex script, causing bugs with LuaJIT
93 * make sure extratags are always null when empty
94 * reduce importance of places without wikipedia reference
95 * improve performance of word count computations
96 * drop support for wikipedia tags with full URLs
97 * replace get_addressdata() SQL implementation with a Python function
98 * improve display name for non-address features
99 * fix postcode validation for postcodes with country code
100 (thanks @pawel-wroniszewski)
101 * add possibility to run imports without superuser database rights
102 (thanks @robbe-haesendonck)
103 * new CLI command for cleaning deleted relations (thanks @lujoh)
104 * add check for database version in the CLI check command
105 * updates to import styles ignoring more unused objects
106 * various typo fixes (thanks @kumarUjjawal)
109 * fix potential SQL injection issue for 'nominatim admin --collect-os-info'
110 * PHP frontend: fix on-the-fly lookup of postcode areas near boundaries
111 * Python frontend: improve handling of viewbox
112 * Python frontend: correct deployment instructions
115 * reintroduce result rematching
116 * improve search of multi-part names
117 * fix accidentally switched meaning of --reverse-only and --search-only in
121 * fix failing importance recalculation command
122 * fix merging of linked names into unnamed boundaries
123 * fix a number of corner cases with interpolation splitting resulting in
125 * fix failure in website generation when password contains curly brackets
126 * fix broken use of ST_Project in PostGIS 3.4
127 * new NOMINATIM_SEARCH_WITHIN_COUNTRIES setting to restrict reverse lookups
128 to known countries (thanks @alfmarcua)
129 * allow negative OSM IDs (thanks @alfmarcua)
130 * disallow import of Tiger data in a frozen DB
131 * avoid UPDATE to change settings to be compatible with r/o DBs (thanks @t-tomek)
132 * update bundled osm2pgsql to 1.9.2
133 * reorganise osm2pgsql flex style and make it the default
134 * exclude names ending in :wikipedia from indexing
135 * no longer accept comma as a list separator in name tags
136 * process forward dependencies on update to catch updates in geometries
137 of ways and relations
138 * fix handling of isolated silent letters during transliteration
139 * no longer assign postcodes to large linear features like rivers
140 * introduce nominatim.paths module for finding data and libraries
141 * documentation layout changed to material theme
142 * new documentation section for library
143 * various smaller fixes to existing documentation
144 (thanks @woodpeck, @bloom256, @biswajit-k)
145 * updates to vagrant install scripts, drop support for Ubuntu 18
147 * removed obsolete configuration variables from env.defaults
148 * add script for generating a taginfo description (thanks @biswajit-k)
149 * modernize Python code around BDD test and add testing of Python frontend
150 * lots of new BDD tests for API output
154 * fix deletion handling for 'nominatim add-data'
155 * adapt place_force_delete() to new deletion handling
156 * flex style: avoid dropping of postcode areas
157 * fix update errors on address interpolation handling
161 * extend flex-style library to fully support all default styles
162 * fix handling of Hebrew aleph
163 * do not assign postcodes to rivers
164 * fix string matching in PHP code
165 * update osm2pgsql (various updates to flex)
166 * fix slow query when deleting places on update
167 * fix CLI details query
168 * fix recalculation of importance values
169 * fix polygon simplification in reverse results
170 * add class/type information to reverse geocodejson result
171 * minor improvements to default tokenizer configuration
172 * various smaller fixes to documentation
176 * fix XSS vulnerability in debug view
180 * add experimental support for osm2pgsql flex style
181 * introduce secondary importance value to be retrieved from a raster data file
182 (currently still unused, to replace address importance, thanks to @tareqpi)
183 * add new report tool `nominatim admin --collect-os-info`
184 (thanks @micahcochran, @tareqpi)
185 * reorganise index to improve lookup performance and size
186 * run index creation after import in parallel
187 * run ANALYZE more selectively to speed up continuation of indexing
188 * fix crash on update when addr:interpolation receives an illegal value
189 * fix minimum number of retrieved results to be at least 10
190 * fix search for combinations of special term + name (e.g Hotel Bellevue)
191 * do not return interpolations without a parent street on reverse search
192 * improve invalidation of linked places on updates
193 * fix address parsing for interpolation lines
194 * make sure socket timeouts are respected during replication
195 (working around a bug in some versions of pyosmium)
196 * update bundled osm2pgsql to 1.7.1
197 * add support for PostgreSQL 15
198 * typing fixes to work with latest type annotations from typeshed
199 * smaller improvements to documentation (thanks to @mausch)
203 * fix XSS vulnerability in debug view
207 * switch to ICU tokenizer as default
208 * add housenumber normalization and support optional spaces during search
209 * add postcode format checking and support optional spaces during search
210 * add function for cleaning housenumbers in word table
211 * add updates/deletion of country names imported from OSM
212 * linked places no longer overwrite names from a place permanently
213 * move default country name configuration into yaml file (thanks @tareqpi)
214 * more compact layout for interpolation and TIGER tables
215 * introduce mutations to ICU tokenizer (used for German umlauts)
216 * support reinitializing a full project directory with refresh --website
217 * fix various issues with linked places on updates
218 * add support for external sanitizers and token analyzers
219 * add CLI commands for forced indexing
220 * add CLI command for version report
221 * add offline import mode
222 * change geocodejson to return a feature class in the 'type' field
223 * add ISO3166-2 to address output (thanks @I70l0teN4ik)
224 * improve parsing and matching of addr: tags
225 * support relations as street members of associatedStreet
226 * better ranking for address results from TIGER data
227 * adapt rank classification to changed tag usage in OSM
228 * update bundled osm2pgsql to 1.6.0
229 * add typing information to Python code
230 * improve unit test coverage
231 * reorganise and speed up code for BDD tests, drop support for scenes
232 * move PHP unit tests to PHP 9.5
233 * extensive typo fixes in documentation (thanks @woodpeck,@StephanGeorg,
234 @amandasaurus, @nslxndr, @stefkiourk, @Luflosi, @kianmeng)
235 * drop official support for installation on CentOS
236 * add installation instructions for Ubuntu 22.04
237 * add support for PHP8
238 * add setup instructions for updates and systemd
239 * drop support for PostgreSQL 9.5
243 * fix XSS vulnerability in debug view
247 * fix initialisation error in replication script
248 * ICU tokenizer: avoid any special characters in word tokens
249 * better error message when API php script does not exist
250 * fix quoting of house numbers in SQL queries
251 * small fixes and improvements in search query parsing
252 * add documentation for moving the database to a different machine
256 * refactor name token computation and introduce ICU tokenizer
257 * name processing now happens in the indexer outside the DB
258 * reorganizes abbreviation handling and moves it to the indexing phases
259 * adds preprocessing of names
260 * add country-specific ranking for Spain, Slovakia
261 * partially switch to using SP-GIST indexes
262 * better updating of dependent addresses for name changes in streets
263 * remove unused/broken tables for external housenumbers
264 * move external postcodes to CSV format and no longer save them in tables
265 (adds support for postcodes for arbitrary countries)
266 * remove postcode helper entries from placex (thanks @AntoJvlt)
267 * change required format for TIGER data to CSV
268 * move configuration of default languages from wiki into config file
269 * expect customized configuration files in project directory by default
270 * disable search API for reverse-only import (thanks @darkshredder)
271 * port most of maintenance/import code to Python and remove PHP utils
272 * add catch-up mode for replication
273 * add updating of special phrases (thanks @AntoJvlt)
274 * add support for special phrases in CSV files (thanks @AntoJvlt)
275 * switch to case-independent matching between place and boundary names
276 * remove disabling of reverse query parsing
277 * minor tweaks to search algorithm to avoid more false positives
278 * major overhaul of the administrator and developer documentation
279 * add security disclosure policy
280 * add testing of installation scripts via CI
281 * drop support for Python < 3.6 and Postgresql < 9.5
285 * fix XSS vulnerability in debug view
289 * fix database check for reverse-only imports
290 * do not error out in status API result when import date is missing
291 * add array_key_last function for PHP < 7.3 (thanks to @woodpeck)
292 * fix more url when server name is unknown (thanks to @mogita)
293 * commit changes to replication log table
297 * fix smaller issues with special phrases import (thanks @AntoJvlt)
298 * add index to speed up continued indexing during import
299 * fix index on location_property_tiger(parent_place_id) (thanks @changpingc)
300 * make sure Python code is backward-compatible with Python 3.5
301 * various documentation fixes
305 * switch to dotenv for configuration file
306 * introduce 'make install' (reorganising most of the code)
307 * introduce nominatim tool as replacement for various php scripts
308 * introduce project directories and allow multiple installations from same build
309 * clean up BDD tests: drop nose, reorganise step code
310 * simplify test database for API BDD tests and autoinstall database
311 * port most of the code for command-line tools to Python
312 (thanks to @darkshredder and @AntoJvlt)
313 * add tests for all tooling
314 * replace pyosmium-get-changes with custom internal implementation using
316 * improve search for queries with housenumber and partial terms
317 * add database versioning
318 * use jinja2 for preprocessing SQL files
319 * introduce automatic migrations
320 * reverse fix preference of interpolations over housenumbers
321 * parallelize indexing of postcodes
322 * add non-key indexes to speed up housenumber + street searches
323 * switch housenumber field in placex to save transliterated names
327 * add full support for searching by and displaying of addr:* tags
328 * improve address output for large-area objects
329 * better use of country names from OSM data for search and display
330 * better debug output for reverse call
331 * add support for addr:place links without an place equivalent in OSM
332 * improve finding postcodes with normalisation artefacts
333 * batch object to index for rank 30, avoiding a wrap-around of transaction
335 * introduce dynamic address rank computation for administrative boundaries
336 depending on linked objects and their place in the admin level hierarchy
337 * add country-specific address ranking for Indonesia, Russia, Belgium and
338 the Netherlands (thanks @hendrikmoree)
339 * make sure wikidata/wikipedia tags are imported for all styles
340 * make POIs searchable by name and housenumber (thanks @joy-yyd)
341 * reverse geocoding now ignores places without an address rank (rivers etc.)
342 * installation of a webserver is no longer mandatory, for development
343 use the php internal webserver via 'make serve
344 * reduce the influence of place nodes in addresses
345 * drop support for the unspecific is_in tag
346 * various minor tweaks to supplied styles
347 * move HTML web frontend into its own project
348 * move scripts for processing external data sources into separate directories
349 * introduce separate configuration for website (thanks @krahulreddy)
350 * update documentation, in particular, clean up development docs
351 * update osm2pgsql to 1.4.0
355 * ensure that wikipedia tags are imported for all styles
356 * reinstate verbosity for indexing during updates
357 * make house number reappear in display name on named POIs
358 * introduce batch processing in indexer to avoid transaction ID overrun
359 * increase splitting for large geometries to improve indexing speed
360 * remove deprecated get_magic_quotes_gpc() function
361 * make sure that all postcodes have an entry in word and are thus searchable
362 * remove use of ST_Covers in conjunction with ST_Intersects,
363 causes bad query planning and slow updates in Postgis3
368 * disable jit and parallel processing in PostgreSQL for osm2pgsql
369 * update libosmium to 2.15.6 (fixes an issue with processing hanging
370 on large multipolygons)
374 * structured select on HTML search page
375 * new PHP Nominatim\Shell class to wrap shell escaping
376 * remove polygon parameter from all API calls
377 * improve handling of postcode areas
378 * reorganise place linking algorithm, now using wikidata tag as well
379 * remove linkees from search_name and larger_area tables
380 * introduce country-specific address ranks
381 * reorganise rank address computation
382 * cleanup of partition function
383 * improve parenting for large POIs
384 * add support for Postgresql 12 and Postgis 3
385 * add earlier cleanup when --drop is given, to reduce memory usage
386 * remove use of place_id in URLs
387 * replace C nominatim indexer with a simpler Python implementation
388 * split up the huge sql/functions.sql file
389 * move osm2pgsql tests to osm2pgsql
390 * add new extratags style which imports all tags from OSM
391 * add new script for checking the import after completion
392 * update osm2pgsql, reducing memory usage
393 * use new wikipedia importance and add processing of wikidata tags
394 * add search form for details page
395 * use ExtraDataPath for country_grid table
396 * remove short_name from list of names to be displayed
397 * split up CMakeFile, so that all parts can be built separately
398 * update installation instructions for CentOS and Ubuntu
399 * add script for importing/updating multiple country extracts
400 * various documentation improvements
404 * fix security bug in /details endpoint where user input was not
409 * update osm2pgsql to fix hans during updates and lost address numbers
414 * increase required version for PostgreSQL(9.3), PostGIS(2.2) and PHP(7.0)
415 * better error reporting for out-of-memory errors
416 * exclude postcode ranges separated by colon from centre point calculation
417 * update osm2pgsql, better handling of imports without flatnode file
418 * switch to more efficient algorithm for word set computation
419 * use only boundaries for country and state parts of addresses
420 * improve updates of addresses with housenumbers and interpolations
421 * remove country from place_addressline table and use country_code instead
422 * optimise indexes on search_name partition tables
423 * improve searching of attached streets for large objects like airports
424 * drop support for python 2
425 * new scripts for importing Wikidata for importance
426 * create and drop indexes concurrently to not clash with auto vacuum
427 * various documentation improvements
432 * zoom 17 in reverse now zooms in on minor streets
433 * fix use of postcode relations in address
434 * support for housenumber 0 on interpolations
435 * replace database abstraction DB with PDO and switch to using exceptions
436 * exclude line features at rank 30 from reverse geocoding
437 * remove self-reference and country from place_addressline
438 * make json output more readable (less escaping)
439 * update conversion scripts for postcodes
440 * scripts in utils/ are no longer executable (always use scripts in build dir)
441 * remove Natural Earth country fallback (OSM is complete enough)
442 * make rank assignments configurable
443 * allow accept languages with underscore
444 * new reverse-only import mode (without search index table)
445 * rely on boundaries only for states and countries
446 * update osm2pgsql, now using a configurable style
447 * provide multiple import styles
448 * improve search when house number and postcodes are dropped
449 * overhaul of setup code
450 * add support for PHPUnit 6
451 * update test database
452 * various documentation improvements
456 * complete rewrite of reverse search algorithm
457 * add new geojson and geocodejson output formats
458 * add simple export script to export addresses to CSV
459 * remove is_in terms from address computation
460 * remove unused search_name_country tables
461 * various smaller fixes to query parsing
462 * convert Tokens and token types to class types
463 * correctly handle update when boundary object type is changed
464 * improve debug output for /search endpoint
465 * update to latest osm2pgsql and leaflet.js
466 * overhaul of /details endpoint:
467 * new class parameter when using osmtype/osmid parameters
468 * permalink to instance-independent osmtype/osmid parameter format
469 * new json output format
470 * update CentOS vagrant machine to use SELinux
471 * add vagrant scripts for Ubuntu 18.04
472 * fix build process for BSD
473 * enable running the database on a different host than the setup scripts
474 * allow to configure use of custom PHP binaries (PHP_BIN)
475 * extensive coding style improvements to PHP code
476 * more PHP unit tests for new classes
477 * increase coverage for API tests
478 * add documentation for API
482 * rework postcode handling and introduce location_postcode table
483 * make setup less verbose and print a summary in the end
484 * setup: error out when web site user does not exist
485 * add more API tests to complete code coverage
486 * reinstate key-value amenity search (special term [key=value])
487 * fix detection of coordinates in query
488 * various smaller tweaks to ranking of search interpretations
489 * complete overhaul of PHP frontend code using OOP
490 * add address rank to details page
491 * update Tiger scripts for 2017 data and clean up unused code
492 * various bug fixes and improvements to UI
493 * improve reverse geocoding performance close to coasts
494 * more PHP style cleanup (quoting)
495 * allow unnamed road in reverse geocoding to avoid too far off results
496 * add function to recalculate counts for full-word search term
497 * add function to check if new updates are available
498 * update documentation and switch to mkdocs for generating HTML
502 * fix bug in geometry building algorithm in osm2pgsql
503 * fix typos in build instructions
507 * move to cmake build system
508 * various fixes to HTML output
509 * reverse endpoint now can return geometries
510 * add support for PHP7
511 * move to on-the-fly computation of interpolations
512 * improve handling of linked places (updating)
513 * improve test framework:
514 * replace lettuce with behave
515 * use smaller database for API tests
516 * drop support for postgres < 9.1, postgis < 2.0 and PHP < 5.4
517 * make external data use optional (useful for imports without US)
518 * detect postgres and postgis versions automatically
519 * clean up query logging and remove unused tables
520 * move installation documentation into this repo
521 * add self-documenting vagrant scripts
522 * remove --create-website, recommend to use website directory in build
523 * add accessor functions for URL parameters and improve error checking
524 * remove IP blocking and rate-limiting code
525 * enable CI via travis
526 * reformatting for more consistent coding style
527 * make country search term creation part of setup
528 * update country names and country grid
529 * handle roads that cross boundaries better
530 * keep full information on address tags
531 * update to refactored osm2pgsql which use libosmium based types
532 * switch from osmosis to pyosmium for updates
533 * be more strict when matching against special search terms
534 * handle postcode entries with multiple values correctly
538 * reverse geocoding includes looking up housenumbers from Tiger data
539 * added parameter to return simplified geometries
540 * new lookup call for getting address information for OSM objects
541 * new namedetails and extratags parameters that expose the name and extratags
542 fields of the placex table
548 * drop support for postgres 8.4
549 * rewrite address interpolation
550 * switch to C++ version of osm2pgsql and rewrite tag filtering
551 * support for bridge:name and tunnel:name, man_made, junction
552 * drop way-node index (reduces database size by about 15%)
553 * add support for configuring tablespaces and webserver user
554 * better evaluation of search queries in right-to-left notation
555 * improve house number search for streets with many duplicate entries
556 * code cleanup (remove unused functions and tables)
560 * further improve ordering of results
561 * support for more lat/lon formats in search-as-reverse
562 * fix handling of GB postcodes
563 * new functional test suite
564 * support for waterway relations
565 * inherit postcodes from street to poi
566 * fix housenumber normalisation to find non-latin house numbers
567 * take viewbox into account for ordering of results
568 * pois may now inherit address tags from surrounding buildings
569 * improve what objects may participate in an address
570 * clean up handled class/type combinations to current OSM usage
575 * correct database rights for www-data
576 * add timestamps for update output
577 * load postgis via extension for postgis >= 2.0
578 * remove non-admin boundaries from addresses
579 * further improve ordering of results with same importance
580 * merge addr:postcode tags into object addresses
581 * include rank and importance in reverse geocode output
582 * replace ST_Line_Interpolate_Point with ST_LineInterpolatePoint
584 * update osm2pgsql to latest version
585 * properly detect changes of admin_level
586 * remove landuses when name is removed
591 * update to openlayers 2.12 (now custom built)
592 * update fallback OSM boundaries
593 * add support for postgresql 9.2/9.3 and postgis 2.x
594 * add structured queries
595 * add options for polygon output in various formats
596 (geojson, svg, kml, postgis text)
597 * maintenance functions for deleting objects and updating regions
598 (plcae_force_update/place_force_delete)
599 * web view for polygons that need deleting
600 * rate limiting using memcache
601 * improve layout of details page
602 * add support for boundary:postal_code
604 * improve parenting of POIs
605 * support for extract daily diffs from Geofabrik
606 * support for addresses without a street
607 (addr:place and conscription number house numbers)
608 * improve layout of word and search_name_* tables
609 * support for US ZIP+4 codes
610 * refactoring of front-end PHP code
611 * lots of smaller bug fixes
615 * delete outdated entries from location_area_country
616 * remove remaining uses of INTEGER, to allow node ids larger than 2^31