Commit | Line | Data |
---|---|---|
4c906759 SJ |
1 | <?php |
2 | /*************************************************************************** | |
12262f13 | 3 | * Copyright (C) 2003-2011 Polytechnique.org * |
4c906759 SJ |
4 | * http://opensource.polytechnique.org/ * |
5 | * * | |
6 | * This program is free software; you can redistribute it and/or modify * | |
7 | * it under the terms of the GNU General Public License as published by * | |
8 | * the Free Software Foundation; either version 2 of the License, or * | |
9 | * (at your option) any later version. * | |
10 | * * | |
11 | * This program is distributed in the hope that it will be useful, * | |
12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of * | |
13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * | |
14 | * GNU General Public License for more details. * | |
15 | * * | |
16 | * You should have received a copy of the GNU General Public License * | |
17 | * along with this program; if not, write to the Free Software * | |
18 | * Foundation, Inc., * | |
19 | * 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA * | |
20 | ***************************************************************************/ | |
21 | ||
4c906759 SJ |
22 | // Implementation of a Geocoder using the Google Maps API. Please refer to |
23 | // the following links for details: | |
24 | // http://code.google.com/apis/maps/documentation/services.html#Geocoding | |
25 | // http://code.google.com/intl/en/apis/maps/documentation/geocoding/ | |
26 | // http://code.google.com/apis/maps/documentation/reference.html#GGeoAddressAccuracy | |
27 | // | |
28 | // It requires the properties gmaps_key and gmaps_url to be defined in section | |
29 | // Geocoder in plat/al's configuration (platal.ini & platal.conf). | |
30 | class GMapsGeocoder extends Geocoder { | |
31 | ||
32 | // Maximum number of Geocoding calls to the Google Maps API. | |
33 | const MAX_GMAPS_RPC_CALLS = 5; | |
7bc2c396 SJ |
34 | // Maximum levenshtein distance authorized between input and geocoded text in a single line. |
35 | const MAX_LINE_DISTANCE = 5; | |
36 | // Maximum levenshtein distance authorized between input and geocoded text in the whole text. | |
37 | const MAX_TOTAL_DISTANCE = 6; | |
4c906759 | 38 | |
26ba053e | 39 | public function getGeocodedAddress(Address $address, $defaultLanguage = null, $forceLanguage = false) { |
4e7a3faa SJ |
40 | $this->prepareAddress($address); |
41 | $textAddress = $this->getTextToGeocode($address->text); | |
2ffc0393 SJ |
42 | if (is_null($defaultLanguage)) { |
43 | $defaultLanguage = Platal::globals()->geocoder->gmaps_hl; | |
44 | } | |
4c906759 SJ |
45 | |
46 | // Try to geocode the full address. | |
2ffc0393 SJ |
47 | if (($geocodedData = $this->getPlacemarkForAddress($textAddress, $defaultLanguage))) { |
48 | $this->getUpdatedAddress($address, $geocodedData, null, $forceLanguage); | |
4e7a3faa | 49 | return; |
4c906759 SJ |
50 | } |
51 | ||
52 | // If the full geocoding failed, try to geocode only the final part of the address. | |
53 | // We start by geocoding everything but the first line, and continue until we get | |
54 | // a result. To respect the limit of GMaps calls, we ignore the first few lines | |
55 | // if there are too many address lines. | |
56 | $addressLines = explode("\n", $textAddress); | |
57 | $linesCount = count($addressLines); | |
58 | for ($i = max(1, $linesCount - self::MAX_GMAPS_RPC_CALLS + 1); $i < $linesCount; ++$i) { | |
59 | $extraLines = implode("\n", array_slice($addressLines, 0, $i)); | |
60 | $toGeocode = implode("\n", array_slice($addressLines, $i)); | |
2ffc0393 SJ |
61 | if (($geocodedData = $this->getPlacemarkForAddress($toGeocode, $defaultLanguage))) { |
62 | $this->getUpdatedAddress($address, $geocodedData, $extraLines, $forceLanguage); | |
4e7a3faa | 63 | return; |
4c906759 SJ |
64 | } |
65 | } | |
4c906759 SJ |
66 | } |
67 | ||
26ba053e | 68 | public function stripGeocodingFromAddress(Address $address) { |
4e7a3faa | 69 | $address->geocodedText = null; |
4e7a3faa SJ |
70 | $address->geoloc_choice = null; |
71 | $address->countryId = null; | |
72 | $address->country = null; | |
73 | $address->administrativeAreaName = null; | |
74 | $address->subAdministrativeAreaName = null; | |
75 | $address->localityName = null; | |
76 | $address->thoroughfareName = null; | |
77 | $address->postalCode = null; | |
78 | $address->accuracy = 0; | |
73f6c165 | 79 | } |
00e5200b | 80 | |
4c906759 SJ |
81 | // Updates the address with the geocoded information from Google Maps. Also |
82 | // cleans up the final informations. | |
26ba053e | 83 | private function getUpdatedAddress(Address $address, array $geocodedData, $extraLines, $forceLanguage) { |
803612ae | 84 | $this->fillAddressWithGeocoding($address, $geocodedData, false); |
2ffc0393 | 85 | $this->formatAddress($address, $extraLines, $forceLanguage); |
4c906759 SJ |
86 | } |
87 | ||
88 | // Retrieves the Placemark object (see #getPlacemarkFromJson()) for the @p | |
89 | // address, by querying the Google Maps API. Returns the array on success, | |
90 | // and null otherwise. | |
2ffc0393 | 91 | private function getPlacemarkForAddress($address, $defaultLanguage) { |
803612ae | 92 | $url = $this->getGeocodingUrl($address, $defaultLanguage); |
4c906759 SJ |
93 | $geoData = $this->getGeoJsonFromUrl($url); |
94 | ||
95 | return ($geoData ? $this->getPlacemarkFromJson($geoData) : null); | |
96 | } | |
97 | ||
98 | // Prepares address to be geocoded | |
26ba053e | 99 | private function prepareAddress(Address $address) { |
4e7a3faa | 100 | $address->text = preg_replace('/\s*\n\s*/m', "\n", trim($address->text)); |
4c906759 SJ |
101 | } |
102 | ||
103 | // Builds the Google Maps geocoder url to fetch information about @p address. | |
104 | // Returns the built url. | |
803612ae | 105 | private function getGeocodingUrl($address, $defaultLanguage) { |
4c906759 SJ |
106 | global $globals; |
107 | ||
108 | $parameters = array( | |
109 | 'key' => $globals->geocoder->gmaps_key, | |
110 | 'sensor' => 'false', // The queried address wasn't obtained from a GPS sensor. | |
803612ae | 111 | 'hl' => $defaultLanguage, |
4c906759 SJ |
112 | 'oe' => 'utf8', // Output encoding. |
113 | 'output' => 'json', // Output format. | |
ce3e0af3 | 114 | 'gl' => $globals->geocoder->gmaps_gl, |
4c906759 SJ |
115 | 'q' => $address, // The queries address. |
116 | ); | |
117 | ||
118 | return $globals->geocoder->gmaps_url . '?' . http_build_query($parameters); | |
119 | } | |
120 | ||
121 | // Fetches JSON-encoded data from a Google Maps API url, and decode them. | |
122 | // Returns the json array on success, and null otherwise. | |
123 | private function getGeoJsonFromUrl($url) { | |
124 | global $globals; | |
125 | ||
126 | // Prepare a backtrace object to log errors. | |
127 | $bt = null; | |
128 | if ($globals->debug & DEBUG_BT) { | |
129 | if (!isset(PlBacktrace::$bt['Geoloc'])) { | |
130 | new PlBacktrace('Geoloc'); | |
131 | } | |
132 | $bt = &PlBacktrace::$bt['Geoloc']; | |
133 | $bt->start($url); | |
134 | } | |
135 | ||
136 | // Fetch the geocoding data. | |
137 | $rawData = file_get_contents($url); | |
138 | if (!$rawData) { | |
139 | if ($bt) { | |
4e7a3faa | 140 | $bt->stop(0, 'Could not retrieve geocoded address from GoogleMaps.'); |
4c906759 SJ |
141 | } |
142 | return null; | |
143 | } | |
144 | ||
145 | // Decode the JSON-encoded data, and check for their validity. | |
146 | $data = json_decode($rawData, true); | |
147 | if ($bt) { | |
148 | $bt->stop(count($data), null, $data); | |
149 | } | |
150 | ||
151 | return $data; | |
152 | } | |
153 | ||
154 | // Extracts the most appropriate placemark from the JSON data fetched from | |
155 | // Google Maps. Returns a Placemark array on success, and null otherwise. See | |
156 | // http://code.google.com/apis/maps/documentation/services.html#Geocoding_Structured | |
157 | // for details on the Placemark structure. | |
158 | private function getPlacemarkFromJson(array $data) { | |
159 | // Check for geocoding failures. | |
160 | if (!isset($data['Status']['code']) || $data['Status']['code'] != 200) { | |
161 | // TODO: handle non-200 codes in a better way, since the code might | |
162 | // indicate a temporary error on Google's side. | |
163 | return null; | |
164 | } | |
165 | ||
166 | // Check that at least one placemark was found. | |
167 | if (count($data['Placemark']) == 0) { | |
168 | return null; | |
169 | } | |
170 | ||
171 | // Extract the placemark with the best accuracy. This is not always the | |
172 | // best result (since the same address may yield two different placemarks). | |
173 | $result = $data['Placemark'][0]; | |
174 | foreach ($data['Placemark'] as $place) { | |
175 | if ($place['AddressDetails']['Accuracy'] > $result['AddressDetails']['Accuracy']) { | |
176 | $result = $place; | |
177 | } | |
178 | } | |
179 | ||
180 | return $result; | |
181 | } | |
182 | ||
183 | // Fills the address with the geocoded data | |
26ba053e | 184 | private function fillAddressWithGeocoding(Address $address, $geocodedData, $isLocal) { |
4c906759 SJ |
185 | // The geocoded address three is |
186 | // Country -> AdministrativeArea -> SubAdministrativeArea -> Locality -> Thoroughfare | |
187 | // with all the possible shortcuts | |
188 | // The address is formatted as xAL, or eXtensible Address Language, an international | |
189 | // standard for address formatting. | |
190 | // xAL documentation: http://www.oasis-open.org/committees/ciq/ciq.html#6 | |
803612ae SJ |
191 | if ($isLocal) { |
192 | $ext = 'Local'; | |
193 | } else { | |
194 | $ext = ucfirst(Platal::globals()->geocoder->gmaps_hl); | |
195 | $address->geocodedText = str_replace(', ', "\n", $geocodedData['address']); | |
196 | } | |
197 | ||
4c906759 | 198 | if (isset($geocodedData['AddressDetails']['Accuracy'])) { |
4e7a3faa | 199 | $address->accuracy = $geocodedData['AddressDetails']['Accuracy']; |
4c906759 SJ |
200 | } |
201 | ||
202 | $currentPosition = $geocodedData['AddressDetails']; | |
203 | if (isset($currentPosition['Country'])) { | |
803612ae | 204 | $country = 'country' . $ext; |
4e7a3faa SJ |
205 | $currentPosition = $currentPosition['Country']; |
206 | $address->countryId = $currentPosition['CountryNameCode']; | |
803612ae | 207 | $address->$country = $currentPosition['CountryName']; |
4c906759 SJ |
208 | } |
209 | if (isset($currentPosition['AdministrativeArea'])) { | |
803612ae SJ |
210 | $administrativeAreaName = 'administrativeAreaName' . $ext; |
211 | $currentPosition = $currentPosition['AdministrativeArea']; | |
212 | $address->$administrativeAreaName = $currentPosition['AdministrativeAreaName']; | |
4c906759 SJ |
213 | } |
214 | if (isset($currentPosition['SubAdministrativeArea'])) { | |
803612ae SJ |
215 | $subAdministrativeAreaName = 'subAdministrativeAreaName' . $ext; |
216 | $currentPosition = $currentPosition['SubAdministrativeArea']; | |
217 | $address->$subAdministrativeAreaName = $currentPosition['SubAdministrativeAreaName']; | |
4c906759 SJ |
218 | } |
219 | if (isset($currentPosition['Locality'])) { | |
803612ae SJ |
220 | $localityName = 'localityName' . $ext; |
221 | $currentPosition = $currentPosition['Locality']; | |
222 | $address->$localityName = $currentPosition['LocalityName']; | |
4c906759 SJ |
223 | } |
224 | if (isset($currentPosition['PostalCode'])) { | |
4e7a3faa | 225 | $address->postalCode = $currentPosition['PostalCode']['PostalCodeNumber']; |
4c906759 SJ |
226 | } |
227 | ||
228 | // Gets coordinates. | |
229 | if (isset($geocodedData['Point']['coordinates'][0])) { | |
4e7a3faa | 230 | $address->latitude = $geocodedData['Point']['coordinates'][0]; |
4c906759 SJ |
231 | } |
232 | if (isset($geocodedData['Point']['coordinates'][1])) { | |
4e7a3faa | 233 | $address->longitude = $geocodedData['Point']['coordinates'][1]; |
4c906759 SJ |
234 | } |
235 | if (isset($geocodedData['ExtendedData']['LatLonBox']['north'])) { | |
4e7a3faa | 236 | $address->north = $geocodedData['ExtendedData']['LatLonBox']['north']; |
4c906759 SJ |
237 | } |
238 | if (isset($geocodedData['ExtendedData']['LatLonBox']['south'])) { | |
4e7a3faa | 239 | $address->south = $geocodedData['ExtendedData']['LatLonBox']['south']; |
4c906759 SJ |
240 | } |
241 | if (isset($geocodedData['ExtendedData']['LatLonBox']['east'])) { | |
4e7a3faa | 242 | $address->east = $geocodedData['ExtendedData']['LatLonBox']['east']; |
4c906759 SJ |
243 | } |
244 | if (isset($geocodedData['ExtendedData']['LatLonBox']['west'])) { | |
4e7a3faa | 245 | $address->west = $geocodedData['ExtendedData']['LatLonBox']['west']; |
4c906759 SJ |
246 | } |
247 | } | |
248 | ||
803612ae SJ |
249 | // Compares the geocoded address with the given address and returns true |
250 | // iff their are close enough to be considered as equals or not. | |
251 | private function compareAddress($address) | |
252 | { | |
4c906759 | 253 | $same = true; |
4c906759 | 254 | $geoloc = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"), |
4e7a3faa | 255 | array('', "\n"), $address->geocodedText)); |
4c906759 | 256 | $text = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"), |
4e7a3faa | 257 | array('', "\n"), $address->text)); |
4c906759 SJ |
258 | $arrayGeoloc = explode("\n", $geoloc); |
259 | $arrayText = explode("\n", $text); | |
260 | $countGeoloc = count($arrayGeoloc); | |
261 | $countText = count($arrayText); | |
262 | ||
7bc2c396 | 263 | $totalDistance = 0; |
4c906759 SJ |
264 | if (($countText > $countGeoloc) || ($countText < $countGeoloc - 1) |
265 | || (($countText == $countGeoloc - 1) | |
4e7a3faa | 266 | && ($arrayText[$countText - 1] == strtoupper($address->country)))) { |
4c906759 SJ |
267 | $same = false; |
268 | } else { | |
269 | for ($i = 0; $i < $countGeoloc && $i < $countText; ++$i) { | |
7bc2c396 SJ |
270 | $lineDistance = levenshtein($arrayText[$i], trim($arrayGeoloc[$i])); |
271 | $totalDistance += $lineDistance; | |
272 | if ($lineDistance > self::MAX_LINE_DISTANCE || $totalDistance > self::MAX_TOTAL_DISTANCE) { | |
4c906759 | 273 | $same = false; |
7bc2c396 | 274 | break; |
4c906759 SJ |
275 | } |
276 | } | |
277 | } | |
7bc2c396 | 278 | |
803612ae SJ |
279 | return $same; |
280 | } | |
281 | ||
282 | // Formats the text of the geocoded address using the unused data and | |
283 | // compares it to the given address. If they are too different, the user | |
284 | // will be asked to choose between them. | |
26ba053e | 285 | private function formatAddress(Address $address, $extraLines, $forceLanguage) |
803612ae SJ |
286 | { |
287 | if ($extraLines) { | |
288 | $address->geocodedText = $extraLines . "\n" . $address->geocodedText; | |
289 | } | |
290 | ||
291 | if ($this->compareAddress($address)) { | |
4e7a3faa | 292 | $address->geocodedText = null; |
2ffc0393 SJ |
293 | } elseif (!$forceLanguage) { |
294 | $languages = XDB::fetchOneCell('SELECT IF(ISNULL(gc1.belongsTo), gl1.language, gl2.language) | |
803612ae | 295 | FROM geoloc_countries AS gc1 |
2ffc0393 | 296 | INNER JOIN geoloc_languages AS gl1 ON (gc1.iso_3166_1_a2 = gl1.iso_3166_1_a2) |
803612ae | 297 | LEFT JOIN geoloc_countries AS gc2 ON (gc1.belongsTo = gc2.iso_3166_1_a2) |
2ffc0393 | 298 | LEFT JOIN geoloc_languages AS gl2 ON (gc2.iso_3166_1_a2 = gl2.iso_3166_1_a2) |
803612ae SJ |
299 | WHERE gc1.iso_3166_1_a2 = {?}', |
300 | $address->countryId); | |
301 | $toGeocode = substr($address->text, strlen($extraLines)); | |
302 | foreach (explode(',', $languages) as $language) { | |
303 | if ($language != Platal::globals()->geocoder->gmaps_hl) { | |
304 | $geocodedData = $this->getPlacemarkForAddress($toGeocode, $language); | |
305 | $address->geocodedText = str_replace(', ', "\n", $geocodedData['address']); | |
306 | if ($extraLines) { | |
307 | $address->geocodedText = $extraLines . "\n" . $address->geocodedText; | |
308 | } | |
309 | if ($this->compareAddress($address)) { | |
310 | $this->fillAddressWithGeocoding($address, $geocodedData, true); | |
311 | $address->geocodedText = null; | |
312 | break; | |
313 | } | |
314 | } | |
315 | } | |
4e7a3faa | 316 | $address->geocodedText = str_replace("\n", "\r\n", $address->geocodedText); |
4c906759 | 317 | } |
4e7a3faa | 318 | $address->text = str_replace("\n", "\r\n", $address->text); |
5a10ab14 SJ |
319 | } |
320 | ||
00e5200b SJ |
321 | // Trims the name of the real country if it contains an ISO 3166-1 non-country |
322 | // item. For that purpose, we compare the last but one line of the address with | |
323 | // all non-country items of ISO 3166-1. | |
4e7a3faa | 324 | private function getTextToGeocode($text) |
00e5200b | 325 | { |
1c305d4c | 326 | $res = XDB::iterator('SELECT countryEn, country |
00e5200b SJ |
327 | FROM geoloc_countries |
328 | WHERE belongsTo IS NOT NULL'); | |
329 | $countries = array(); | |
330 | foreach ($res as $item) { | |
331 | $countries[] = $item[0]; | |
332 | $countries[] = $item[1]; | |
333 | } | |
4e7a3faa | 334 | $textLines = explode("\n", $text); |
00e5200b SJ |
335 | $countLines = count($textLines); |
336 | $needle = strtoupper(trim($textLines[$countLines - 2])); | |
337 | $isPseudoCountry = false; | |
96c7ea54 SJ |
338 | if ($needle) { |
339 | foreach ($countries as $country) { | |
340 | if (strtoupper($country) === $needle) { | |
341 | $isPseudoCountry = true; | |
342 | break; | |
343 | } | |
00e5200b SJ |
344 | } |
345 | } | |
346 | ||
347 | if ($isPseudoCountry) { | |
02c4b93a | 348 | return implode("\n", array_slice($textLines, 0, -1)); |
00e5200b | 349 | } |
4e7a3faa | 350 | return $text; |
00e5200b | 351 | } |
4c906759 SJ |
352 | } |
353 | ||
354 | // vim:set et sw=4 sts=4 sws=4 foldmethod=marker enc=utf-8: | |
355 | ?> |