classes/gmapsgeocoder.php

   1 <?php
   2 /***************************************************************************
   3  *  Copyright (C) 2003-2010 Polytechnique.org                              *
   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
  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;
  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;
  38
  39     public function getGeocodedAddress(Address &$address) {
  40         $this->prepareAddress($address);
  41         $textAddress = $this->getTextToGeocode($address->text);
  42
  43         // Try to geocode the full address.
  44         if (($geocodedData = $this->getPlacemarkForAddress($textAddress))) {
  45             $this->getUpdatedAddress($address, $geocodedData, null);
  46             return;
  47         }
  48
  49         // If the full geocoding failed, try to geocode only the final part of the address.
  50         // We start by geocoding everything but the first line, and continue until we get
  51         // a result. To respect the limit of GMaps calls, we ignore the first few lines
  52         // if there are too many address lines.
  53         $addressLines = explode("\n", $textAddress);
  54         $linesCount   = count($addressLines);
  55         for ($i = max(1, $linesCount - self::MAX_GMAPS_RPC_CALLS + 1); $i < $linesCount; ++$i) {
  56             $extraLines = implode("\n", array_slice($addressLines, 0, $i));
  57             $toGeocode  = implode("\n", array_slice($addressLines, $i));
  58             if (($geocodedData = $this->getPlacemarkForAddress($toGeocode))) {
  59                 $this->getUpdatedAddress($address, $geocodedData, $extraLines);
  60                 return;
  61             }
  62         }
  63     }
  64
  65     public function stripGeocodingFromAddress(Address &$address) {
  66         $address->geocodedText = null;
  67         $address->geoloc_choice = null;
  68         $address->countryId = null;
  69         $address->country = null;
  70         $address->administrativeAreaName = null;
  71         $address->subAdministrativeAreaName = null;
  72         $address->localityName = null;
  73         $address->thoroughfareName = null;
  74         $address->postalCode = null;
  75         $address->accuracy = 0;
  76     }
  77
  78     // Updates the address with the geocoded information from Google Maps. Also
  79     // cleans up the final informations.
  80     private function getUpdatedAddress(Address &$address, array $geocodedData, $extraLines) {
  81         $this->fillAddressWithGeocoding($address, $geocodedData, false);
  82         $this->formatAddress($address, $extraLines);
  83     }
  84
  85     // Retrieves the Placemark object (see #getPlacemarkFromJson()) for the @p
  86     // address, by querying the Google Maps API. Returns the array on success,
  87     // and null otherwise.
  88     private function getPlacemarkForAddress($address, $defaultLanguage = null) {
  89         if (is_null($defaultLanguage)) {
  90             $defaultLanguage = Platal::globals()->geocoder->gmaps_hl;
  91         }
  92
  93         $url     = $this->getGeocodingUrl($address, $defaultLanguage);
  94         $geoData = $this->getGeoJsonFromUrl($url);
  95
  96         return ($geoData ? $this->getPlacemarkFromJson($geoData) : null);
  97     }
  98
  99     // Prepares address to be geocoded
 100     private function prepareAddress(Address &$address) {
 101         $address->text = preg_replace('/\s*\n\s*/m', "\n", trim($address->text));
 102     }
 103
 104     // Builds the Google Maps geocoder url to fetch information about @p address.
 105     // Returns the built url.
 106     private function getGeocodingUrl($address, $defaultLanguage) {
 107         global $globals;
 108
 109         $parameters = array(
 110             'key'    => $globals->geocoder->gmaps_key,
 111             'sensor' => 'false',   // The queried address wasn't obtained from a GPS sensor.
 112             'hl'     => $defaultLanguage,
 113             'oe'     => 'utf8',    // Output encoding.
 114             'output' => 'json',    // Output format.
 115             'gl'     => $globals->geocoder->gmaps_gl,
 116             'q'      => $address,  // The queries address.
 117         );
 118
 119         return $globals->geocoder->gmaps_url . '?' . http_build_query($parameters);
 120     }
 121
 122     // Fetches JSON-encoded data from a Google Maps API url, and decode them.
 123     // Returns the json array on success, and null otherwise.
 124     private function getGeoJsonFromUrl($url) {
 125         global $globals;
 126
 127         // Prepare a backtrace object to log errors.
 128         $bt = null;
 129         if ($globals->debug & DEBUG_BT) {
 130             if (!isset(PlBacktrace::$bt['Geoloc'])) {
 131                 new PlBacktrace('Geoloc');
 132             }
 133             $bt = &PlBacktrace::$bt['Geoloc'];
 134             $bt->start($url);
 135         }
 136
 137         // Fetch the geocoding data.
 138         $rawData = file_get_contents($url);
 139         if (!$rawData) {
 140             if ($bt) {
 141                 $bt->stop(0, 'Could not retrieve geocoded address from GoogleMaps.');
 142             }
 143             return null;
 144         }
 145
 146         // Decode the JSON-encoded data, and check for their validity.
 147         $data = json_decode($rawData, true);
 148         if ($bt) {
 149             $bt->stop(count($data), null, $data);
 150         }
 151
 152         return $data;
 153     }
 154
 155     // Extracts the most appropriate placemark from the JSON data fetched from
 156     // Google Maps. Returns a Placemark array on success, and null otherwise. See
 157     // http://code.google.com/apis/maps/documentation/services.html#Geocoding_Structured
 158     // for details on the Placemark structure.
 159     private function getPlacemarkFromJson(array $data) {
 160         // Check for geocoding failures.
 161         if (!isset($data['Status']['code']) || $data['Status']['code'] != 200) {
 162             // TODO: handle non-200 codes in a better way, since the code might
 163             // indicate a temporary error on Google's side.
 164             return null;
 165         }
 166
 167         // Check that at least one placemark was found.
 168         if (count($data['Placemark']) == 0) {
 169             return null;
 170         }
 171
 172         // Extract the placemark with the best accuracy. This is not always the
 173         // best result (since the same address may yield two different placemarks).
 174         $result = $data['Placemark'][0];
 175         foreach ($data['Placemark'] as $place) {
 176             if ($place['AddressDetails']['Accuracy'] > $result['AddressDetails']['Accuracy']) {
 177                 $result = $place;
 178             }
 179         }
 180
 181         return $result;
 182     }
 183
 184     // Fills the address with the geocoded data
 185     private function fillAddressWithGeocoding(Address &$address, $geocodedData, $isLocal) {
 186         // The geocoded address three is
 187         // Country -> AdministrativeArea -> SubAdministrativeArea -> Locality -> Thoroughfare
 188         // with all the possible shortcuts
 189         // The address is formatted as xAL, or eXtensible Address Language, an international
 190         // standard for address formatting.
 191         // xAL documentation: http://www.oasis-open.org/committees/ciq/ciq.html#6
 192         if ($isLocal) {
 193             $ext = 'Local';
 194         } else {
 195             $ext = ucfirst(Platal::globals()->geocoder->gmaps_hl);
 196             $address->geocodedText = str_replace(', ', "\n", $geocodedData['address']);
 197         }
 198
 199         if (isset($geocodedData['AddressDetails']['Accuracy'])) {
 200             $address->accuracy = $geocodedData['AddressDetails']['Accuracy'];
 201         }
 202
 203         $currentPosition = $geocodedData['AddressDetails'];
 204         if (isset($currentPosition['Country'])) {
 205             $country = 'country' . $ext;
 206             $currentPosition    = $currentPosition['Country'];
 207             $address->countryId = $currentPosition['CountryNameCode'];
 208             $address->$country  = $currentPosition['CountryName'];
 209         }
 210         if (isset($currentPosition['AdministrativeArea'])) {
 211             $administrativeAreaName = 'administrativeAreaName' . $ext;
 212             $currentPosition                  = $currentPosition['AdministrativeArea'];
 213             $address->$administrativeAreaName = $currentPosition['AdministrativeAreaName'];
 214         }
 215         if (isset($currentPosition['SubAdministrativeArea'])) {
 216             $subAdministrativeAreaName = 'subAdministrativeAreaName' . $ext;
 217             $currentPosition                     = $currentPosition['SubAdministrativeArea'];
 218             $address->$subAdministrativeAreaName = $currentPosition['SubAdministrativeAreaName'];
 219         }
 220         if (isset($currentPosition['Locality'])) {
 221             $localityName = 'localityName' . $ext;
 222             $currentPosition        = $currentPosition['Locality'];
 223             $address->$localityName = $currentPosition['LocalityName'];
 224         }
 225         if (isset($currentPosition['PostalCode'])) {
 226             $address->postalCode = $currentPosition['PostalCode']['PostalCodeNumber'];
 227         }
 228
 229         // Gets coordinates.
 230         if (isset($geocodedData['Point']['coordinates'][0])) {
 231             $address->latitude = $geocodedData['Point']['coordinates'][0];
 232         }
 233         if (isset($geocodedData['Point']['coordinates'][1])) {
 234             $address->longitude = $geocodedData['Point']['coordinates'][1];
 235         }
 236         if (isset($geocodedData['ExtendedData']['LatLonBox']['north'])) {
 237             $address->north = $geocodedData['ExtendedData']['LatLonBox']['north'];
 238         }
 239         if (isset($geocodedData['ExtendedData']['LatLonBox']['south'])) {
 240             $address->south = $geocodedData['ExtendedData']['LatLonBox']['south'];
 241         }
 242         if (isset($geocodedData['ExtendedData']['LatLonBox']['east'])) {
 243             $address->east = $geocodedData['ExtendedData']['LatLonBox']['east'];
 244         }
 245         if (isset($geocodedData['ExtendedData']['LatLonBox']['west'])) {
 246             $address->west = $geocodedData['ExtendedData']['LatLonBox']['west'];
 247         }
 248     }
 249
 250     // Compares the geocoded address with the given address and returns true
 251     // iff their are close enough to be considered as equals or not.
 252     private function compareAddress($address)
 253     {
 254         $same = true;
 255         $geoloc = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"),
 256                                           array('', "\n"), $address->geocodedText));
 257         $text   = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"),
 258                                           array('', "\n"), $address->text));
 259         $arrayGeoloc = explode("\n", $geoloc);
 260         $arrayText   = explode("\n", $text);
 261         $countGeoloc = count($arrayGeoloc);
 262         $countText   = count($arrayText);
 263
 264         $totalDistance = 0;
 265         if (($countText > $countGeoloc) || ($countText < $countGeoloc - 1)
 266             || (($countText == $countGeoloc - 1)
 267                 && ($arrayText[$countText - 1] == strtoupper($address->country)))) {
 268             $same = false;
 269         } else {
 270             for ($i = 0; $i < $countGeoloc && $i < $countText; ++$i) {
 271                 $lineDistance = levenshtein($arrayText[$i], trim($arrayGeoloc[$i]));
 272                 $totalDistance += $lineDistance;
 273                 if ($lineDistance > self::MAX_LINE_DISTANCE || $totalDistance > self::MAX_TOTAL_DISTANCE) {
 274                     $same = false;
 275                     break;
 276                 }
 277             }
 278         }
 279
 280         return $same;
 281     }
 282
 283     // Formats the text of the geocoded address using the unused data and
 284     // compares it to the given address. If they are too different, the user
 285     // will be asked to choose between them.
 286     private function formatAddress(Address &$address, $extraLines)
 287     {
 288         if ($extraLines) {
 289             $address->geocodedText = $extraLines . "\n" . $address->geocodedText;
 290         }
 291
 292         if ($this->compareAddress($address)) {
 293             $address->geocodedText = null;
 294         } else {
 295             $languages = XDB::fetchOneCell('SELECT  IF(ISNULL(gc1.belongsTo), gc1.languages, gc2.languages)
 296                                               FROM  geoloc_countries AS gc1
 297                                          LEFT JOIN  geoloc_countries AS gc2 ON (gc1.belongsTo = gc2.iso_3166_1_a2)
 298                                              WHERE  gc1.iso_3166_1_a2 = {?}',
 299                                            $address->countryId);
 300             $toGeocode = substr($address->text, strlen($extraLines));
 301             foreach (explode(',', $languages) as $language) {
 302                 if ($language != Platal::globals()->geocoder->gmaps_hl) {
 303                     $geocodedData = $this->getPlacemarkForAddress($toGeocode, $language);
 304                     $address->geocodedText = str_replace(', ', "\n", $geocodedData['address']);
 305                     if ($extraLines) {
 306                         $address->geocodedText = $extraLines . "\n" . $address->geocodedText;
 307                     }
 308                     if ($this->compareAddress($address)) {
 309                         $this->fillAddressWithGeocoding($address, $geocodedData, true);
 310                         $address->geocodedText = null;
 311                         break;
 312                     }
 313                 }
 314             }
 315             $address->geocodedText = str_replace("\n", "\r\n", $address->geocodedText);
 316         }
 317         $address->text = str_replace("\n", "\r\n", $address->text);
 318     }
 319
 320     // Trims the name of the real country if it contains an ISO 3166-1 non-country
 321     // item. For that purpose, we compare the last but one line of the address with
 322     // all non-country items of ISO 3166-1.
 323     private function getTextToGeocode($text)
 324     {
 325         $res = XDB::iterator('SELECT  countryEn, country
 326                                 FROM  geoloc_countries
 327                                WHERE  belongsTo IS NOT NULL');
 328         $countries = array();
 329         foreach ($res as $item) {
 330             $countries[] = $item[0];
 331             $countries[] = $item[1];
 332         }
 333         $textLines  = explode("\n", $text);
 334         $countLines = count($textLines);
 335         $needle     = strtoupper(trim($textLines[$countLines - 2]));
 336         $isPseudoCountry = false;
 337         foreach ($countries as $country) {
 338             if (strtoupper($country) == $needle) {
 339                 $isPseudoCountry = true;
 340                 break;
 341             }
 342         }
 343
 344         if ($isPseudoCountry) {
 345             return implode("\n", array_slice($textLines, 0, -1));
 346         }
 347         return $text;
 348     }
 349 }
 350
 351 // vim:set et sw=4 sts=4 sws=4 foldmethod=marker enc=utf-8:
 352 ?>