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);
  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) {
  89         $url     = $this->getGeocodingUrl($address);
  90         $geoData = $this->getGeoJsonFromUrl($url);
  91
  92         return ($geoData ? $this->getPlacemarkFromJson($geoData) : null);
  93     }
  94
  95     // Prepares address to be geocoded
  96     private function prepareAddress(Address &$address) {
  97         $address->text = preg_replace('/\s*\n\s*/m', "\n", trim($address->text));
  98     }
  99
 100     // Builds the Google Maps geocoder url to fetch information about @p address.
 101     // Returns the built url.
 102     private function getGeocodingUrl($address) {
 103         global $globals;
 104
 105         $parameters = array(
 106             'key'    => $globals->geocoder->gmaps_key,
 107             'sensor' => 'false',   // The queried address wasn't obtained from a GPS sensor.
 108             'hl'     => $globals->geocoder->gmaps_hl,
 109             'oe'     => 'utf8',    // Output encoding.
 110             'output' => 'json',    // Output format.
 111             'gl'     => $globals->geocoder->gmaps_gl,
 112             'q'      => $address,  // The queries address.
 113         );
 114
 115         return $globals->geocoder->gmaps_url . '?' . http_build_query($parameters);
 116     }
 117
 118     // Fetches JSON-encoded data from a Google Maps API url, and decode them.
 119     // Returns the json array on success, and null otherwise.
 120     private function getGeoJsonFromUrl($url) {
 121         global $globals;
 122
 123         // Prepare a backtrace object to log errors.
 124         $bt = null;
 125         if ($globals->debug & DEBUG_BT) {
 126             if (!isset(PlBacktrace::$bt['Geoloc'])) {
 127                 new PlBacktrace('Geoloc');
 128             }
 129             $bt = &PlBacktrace::$bt['Geoloc'];
 130             $bt->start($url);
 131         }
 132
 133         // Fetch the geocoding data.
 134         $rawData = file_get_contents($url);
 135         if (!$rawData) {
 136             if ($bt) {
 137                 $bt->stop(0, 'Could not retrieve geocoded address from GoogleMaps.');
 138             }
 139             return null;
 140         }
 141
 142         // Decode the JSON-encoded data, and check for their validity.
 143         $data = json_decode($rawData, true);
 144         if ($bt) {
 145             $bt->stop(count($data), null, $data);
 146         }
 147
 148         return $data;
 149     }
 150
 151     // Extracts the most appropriate placemark from the JSON data fetched from
 152     // Google Maps. Returns a Placemark array on success, and null otherwise. See
 153     // http://code.google.com/apis/maps/documentation/services.html#Geocoding_Structured
 154     // for details on the Placemark structure.
 155     private function getPlacemarkFromJson(array $data) {
 156         // Check for geocoding failures.
 157         if (!isset($data['Status']['code']) || $data['Status']['code'] != 200) {
 158             // TODO: handle non-200 codes in a better way, since the code might
 159             // indicate a temporary error on Google's side.
 160             return null;
 161         }
 162
 163         // Check that at least one placemark was found.
 164         if (count($data['Placemark']) == 0) {
 165             return null;
 166         }
 167
 168         // Extract the placemark with the best accuracy. This is not always the
 169         // best result (since the same address may yield two different placemarks).
 170         $result = $data['Placemark'][0];
 171         foreach ($data['Placemark'] as $place) {
 172             if ($place['AddressDetails']['Accuracy'] > $result['AddressDetails']['Accuracy']) {
 173                 $result = $place;
 174             }
 175         }
 176
 177         return $result;
 178     }
 179
 180     // Fills the address with the geocoded data
 181     private function fillAddressWithGeocoding(Address &$address, $geocodedData) {
 182         // The geocoded address three is
 183         // Country -> AdministrativeArea -> SubAdministrativeArea -> Locality -> Thoroughfare
 184         // with all the possible shortcuts
 185         // The address is formatted as xAL, or eXtensible Address Language, an international
 186         // standard for address formatting.
 187         // xAL documentation: http://www.oasis-open.org/committees/ciq/ciq.html#6
 188         $address->geocodedText = str_replace(', ', "\n", $geocodedData['address']);
 189         if (isset($geocodedData['AddressDetails']['Accuracy'])) {
 190             $address->accuracy = $geocodedData['AddressDetails']['Accuracy'];
 191         }
 192
 193         $currentPosition = $geocodedData['AddressDetails'];
 194         if (isset($currentPosition['Country'])) {
 195             $currentPosition    = $currentPosition['Country'];
 196             $address->countryId = $currentPosition['CountryNameCode'];
 197             $address->country   = $currentPosition['CountryName'];
 198         }
 199         if (isset($currentPosition['AdministrativeArea'])) {
 200             $currentPosition                 = $currentPosition['AdministrativeArea'];
 201             $address->administrativeAreaName = $currentPosition['AdministrativeAreaName'];
 202         }
 203         if (isset($currentPosition['SubAdministrativeArea'])) {
 204             $currentPosition                    = $currentPosition['SubAdministrativeArea'];
 205             $address->subAdministrativeAreaName = $currentPosition['SubAdministrativeAreaName'];
 206         }
 207         if (isset($currentPosition['Locality'])) {
 208             $currentPosition       = $currentPosition['Locality'];
 209             $address->localityName = $currentPosition['LocalityName'];
 210         }
 211         if (isset($currentPosition['Thoroughfare'])) {
 212             $address->thoroughfareName = $currentPosition['Thoroughfare']['ThoroughfareName'];
 213         }
 214         if (isset($currentPosition['PostalCode'])) {
 215             $address->postalCode = $currentPosition['PostalCode']['PostalCodeNumber'];
 216         }
 217
 218         // Gets coordinates.
 219         if (isset($geocodedData['Point']['coordinates'][0])) {
 220             $address->latitude = $geocodedData['Point']['coordinates'][0];
 221         }
 222         if (isset($geocodedData['Point']['coordinates'][1])) {
 223             $address->longitude = $geocodedData['Point']['coordinates'][1];
 224         }
 225         if (isset($geocodedData['ExtendedData']['LatLonBox']['north'])) {
 226             $address->north = $geocodedData['ExtendedData']['LatLonBox']['north'];
 227         }
 228         if (isset($geocodedData['ExtendedData']['LatLonBox']['south'])) {
 229             $address->south = $geocodedData['ExtendedData']['LatLonBox']['south'];
 230         }
 231         if (isset($geocodedData['ExtendedData']['LatLonBox']['east'])) {
 232             $address->east = $geocodedData['ExtendedData']['LatLonBox']['east'];
 233         }
 234         if (isset($geocodedData['ExtendedData']['LatLonBox']['west'])) {
 235             $address->west = $geocodedData['ExtendedData']['LatLonBox']['west'];
 236         }
 237     }
 238
 239     // Formats the text of the geocoded address using the unused data and
 240     // compares it to the given address. If they are too different, the user
 241     // will be asked to choose between them.
 242     private function formatAddress(Address &$address, $extraLines) {
 243         $same = true;
 244         if ($extraLines) {
 245             $address->geocodedText = $extraLines . "\n" . $address->geocodedText;
 246         }
 247         $geoloc = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"),
 248                                           array('', "\n"), $address->geocodedText));
 249         $text   = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"),
 250                                           array('', "\n"), $address->text));
 251         $arrayGeoloc = explode("\n", $geoloc);
 252         $arrayText   = explode("\n", $text);
 253         $countGeoloc = count($arrayGeoloc);
 254         $countText   = count($arrayText);
 255
 256         $totalDistance = 0;
 257         if (($countText > $countGeoloc) || ($countText < $countGeoloc - 1)
 258             || (($countText == $countGeoloc - 1)
 259                 && ($arrayText[$countText - 1] == strtoupper($address->country)))) {
 260             $same = false;
 261         } else {
 262             for ($i = 0; $i < $countGeoloc && $i < $countText; ++$i) {
 263                 $lineDistance = levenshtein($arrayText[$i], trim($arrayGeoloc[$i]));
 264                 $totalDistance += $lineDistance;
 265                 if ($lineDistance > self::MAX_LINE_DISTANCE || $totalDistance > self::MAX_TOTAL_DISTANCE) {
 266                     $same = false;
 267                     break;
 268                 }
 269             }
 270         }
 271
 272         if ($same) {
 273             $address->geocodedText = null;
 274         } else {
 275             $address->geocodedText = str_replace("\n", "\r\n", $address->geocodedText);
 276         }
 277         $address->text = str_replace("\n", "\r\n", $address->text);
 278     }
 279
 280     // Trims the name of the real country if it contains an ISO 3166-1 non-country
 281     // item. For that purpose, we compare the last but one line of the address with
 282     // all non-country items of ISO 3166-1.
 283     private function getTextToGeocode($text)
 284     {
 285         $res = XDB::iterator('SELECT  countryEn, country
 286                                 FROM  geoloc_countries
 287                                WHERE  belongsTo IS NOT NULL');
 288         $countries = array();
 289         foreach ($res as $item) {
 290             $countries[] = $item[0];
 291             $countries[] = $item[1];
 292         }
 293         $textLines  = explode("\n", $text);
 294         $countLines = count($textLines);
 295         $needle     = strtoupper(trim($textLines[$countLines - 2]));
 296         $isPseudoCountry = false;
 297         foreach ($countries as $country) {
 298             if (strtoupper($country) == $needle) {
 299                 $isPseudoCountry = true;
 300                 break;
 301             }
 302         }
 303
 304         if ($isPseudoCountry) {
 305             return implode("\n", array_slice($textLines, 0, -1));
 306         }
 307         return $text;
 308     }
 309 }
 310
 311 // vim:set et sw=4 sts=4 sws=4 foldmethod=marker enc=utf-8:
 312 ?>