Commit Graph

293 Commits

Author SHA1 Message Date
Al
dbf7242ea0 [fix] /cls/self/ 2017-02-04 19:12:49 -05:00
Al
0169448a4d [addresses] adding Central European city district regexes (e.g. Praha 1, Budapest IV, etc.) to country-specific cleanup 2017-02-03 20:54:23 -05:00
Al
cd1875d077 [fix] import 2017-01-27 18:35:43 -05:00
Al
82fb5c1dca [countries] moving country constants to a separate module 2017-01-27 13:15:36 -05:00
Al
287d2f4048 [fix] leading zeros on numeric phrases 2017-01-25 01:40:20 -05:00
Al
bc748b6d62 [addresses] supplying country arg when stripping name affixes both for OSM place-based data sets (ways, localities) and OpenAddresses (shouldn't affect any of the countries currently in OA though) 2017-01-23 23:30:33 -05:00
Al
c36611c060 [addresses] let containing components include all boundaries, not just those that are larger than the current boundary (affects cases like Buenos Aires where the city has a lower admin level than its districts, so would be subject to the boundary config's contained_by override) 2017-01-23 10:48:08 -05:00
Al
110665651c [fix] existing cleanup_street_name method 2017-01-19 02:40:18 -05:00
Al
a931c5ddc9 [osm] checking for valid street names in OSM street-only training data so e.g. the street name is not just a simple number like "831" 2017-01-19 02:34:29 -05:00
Al
024a6a40b1 [addresses] refactoring place dropout into its own method 2017-01-16 19:35:16 -05:00
Al
09b3aeb7d9 [fix] component 2017-01-11 16:50:54 -05:00
Al
ed5dd28023 [addresses] adding some more synonyms to Brasilia street regex 2017-01-11 16:31:30 -05:00
Al
7f851810d2 [addresses] formatting addresses in Brasilia, so e.g. "Bloco B" is never part of the street name or building name, it's the house number. place=neighbourhood maps to nothing in Brasilia as these are basically subdivisions whose streets are identically named 2017-01-11 16:18:04 -05:00
Al
d528095984 [addresses] adding random unit numbers with more digits 2017-01-11 04:24:35 -05:00
Al
86c7b7f3fe [addresses] no longer normalizing slashes in boundary names for places that have multilingual names, etc. 2017-01-08 12:41:51 -05:00
Al
a6d94f998b [addresses] stripping parentheticals in admin boundary names as sometimes cities in e.g. Switzerland are like Oberwil (ZG) in OSM 2017-01-08 03:43:22 -05:00
Al
cfdef1788c [addresses] stripping unit from street using the libpostal dictionaries in all the address data sets. Happens surprisingly often in OpenStreetMap as well as OpenAddresses 2017-01-06 10:06:23 -05:00
Al
de2dffa315 [addresses] adding Calle to purely numeric Spanish street names in OSM as well 2017-01-02 23:41:01 -05:00
Al
21a2a7419a [addresses] only add village as city component if no city can be found in the area 2016-12-29 13:41:05 -05:00
Al
f58ebbdf7f [fix] var name 2016-12-28 14:37:00 -05:00
Al
7ee44a584b [fix] genitive case for Russian/Ukrainian toponyms, not locative (#125) 2016-12-28 14:34:28 -05:00
Al
e6e4b28e43 [addresses] making the город/г. prefix apply to the Russian language rather than the country 2016-12-28 13:26:19 -05:00
Al
f995fdf9d2 [fix] default None 2016-12-28 05:09:15 -05:00
Al
91013fe296 [fix] moving checks inside the add_locatives function, fixing float cast 2016-12-28 04:59:27 -05:00
Al
6f009fb8a6 [addresses] adding pymorphy2 for converting Russian and Ukrainian place names (sticking with state and staet_district for the moment) to the locative case as mentioned in #125 2016-12-28 04:48:32 -05:00
Al
165056ccd8 [names] adding configurable prefix/suffix additions for boundary names 2016-12-27 20:32:23 -05:00
Al
80a9c1b308 [addresses] move country-specific cleanups to before reverse geocoding as those deal with the user-specified components 2016-12-27 04:19:57 -05:00
Al
76d8fc1d37 [fix] combined components 2016-12-26 21:35:27 -05:00
Al
d208397ecb [addresses] checking if component is generated in combining fields 2016-12-26 16:58:10 -05:00
Al
afe29abf6c [fix] name 2016-12-25 11:38:18 -05:00
Al
6a852f02bd [fix] var 2016-12-25 03:17:05 -05:00
Al
4edaca7d37 [fix] var name 2016-12-25 02:50:29 -05:00
Al
4cf40f8deb [addresses] sort combined Japanese suburbs by admin level 2016-12-25 02:29:06 -05:00
Al
5b5a3fe235 [fix] adding Taiwan, Hong Kong, and Macao to the CJK patterns since language affects the order 2016-12-25 01:20:59 -05:00
Al
6da092e144 [fix] update language to English when using English names in CJK countries 2016-12-25 01:18:54 -05:00
Al
51802035de [fix] var name 2016-12-25 01:08:44 -05:00
Al
57f17a5d38 [addresses] remove generated components in combined house numbers if the other components were not numeric. Add house number phrase after the units, etc. are generated so it may be applied to a combined house number as well 2016-12-25 00:44:48 -05:00
Al
dad57dc57e [fix] moving CJK check into the if block so language gets changed more often even if the street sign-based language is unk 2016-12-24 21:20:38 -05:00
Al
826cbc7f24 [addresses/JP] more checks for matching major/minor neighborhood polygons with nodes in Japan 2016-12-24 20:21:25 -05:00
Al
aea0b83619 [fix] brackets 2016-12-24 20:04:12 -05:00
Al
52a15a7c3c [addresses/JP] same deal for reverse geocoding in Japan (combine admin_level 9 and 10 into a single phrase, including points with certain conditions) 2016-12-24 20:01:11 -05:00
Al
b0062a35b8 [fix] var name 2016-12-24 18:52:34 -05:00
Al
840d6c25d8 [addresses] move suffix checks to the end of expanded(), check language_suffix rather than language for Romaji, so language can remain Japanese. For CJK languages, change language before adding generated components 2016-12-24 18:27:53 -05:00
Al
77efcb3f89 [fix] only accept language suffixes that are valid scripts or transliterations of CJK languages. Set language to language suffix so Romaji forms get used, etc. 2016-12-24 17:17:09 -05:00
Al
67d7d94eea [numbers] adding function to format full-width numbers as ASCII 2016-12-24 16:07:31 -05:00
Al
54b0af7f68 [addresses] add chome form for Japanese neighborhoods 2016-12-24 16:06:29 -05:00
Al
441ec00289 [openaddresses] using the new fuzzy equivalence comparison to check if suburb and city names are equal 2016-12-23 02:08:53 -05:00
Al
0814381d7f [fix] dehyphenate multiword names before city/suburb comparison 2016-12-23 01:53:09 -05:00
Al
70b98c877d [fix] except when None 2016-12-22 23:25:20 -05:00
Al
00f3f3f94d [fix] now that neighborhood is classified at index construction time, no longer need to assume suburb for components that might otherwise be a city, etc. 2016-12-22 23:21:08 -05:00