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->geocodedPostalText = null;
  68         $address->geoloc_choice = null;
  69         $address->countryId = null;
  70         $address->country = null;
  71         $address->administrativeAreaName = null;
  72         $address->subAdministrativeAreaName = null;
  73         $address->localityName = null;
  74         $address->thoroughfareName = null;
  75         $address->postalCode = null;
  76         $address->accuracy = 0;
  77     }
  78
  79     // Updates the address with the geocoded information from Google Maps. Also
  80     // cleans up the final informations.
  81     private function getUpdatedAddress(Address &$address, array $geocodedData, $extraLines) {
  82         $this->fillAddressWithGeocoding($address, $geocodedData);
  83         $this->formatAddress($address, $extraLines);
  84     }
  85
  86     // Retrieves the Placemark object (see #getPlacemarkFromJson()) for the @p
  87     // address, by querying the Google Maps API. Returns the array on success,
  88     // and null otherwise.
  89     private function getPlacemarkForAddress($address) {
  90         $url     = $this->getGeocodingUrl($address);
  91         $geoData = $this->getGeoJsonFromUrl($url);
  92
  93         return ($geoData ? $this->getPlacemarkFromJson($geoData) : null);
  94     }
  95
  96     // Prepares address to be geocoded
  97     private function prepareAddress(Address &$address) {
  98         $address->text = preg_replace('/\s*\n\s*/m', "\n", trim($address->text));
  99         $address->postalText = $this->getPostalAddress($address->text);
 100     }
 101
 102     // Builds the Google Maps geocoder url to fetch information about @p address.
 103     // Returns the built url.
 104     private function getGeocodingUrl($address) {
 105         global $globals;
 106
 107         $parameters = array(
 108             'key'    => $globals->geocoder->gmaps_key,
 109             'sensor' => 'false',   // The queried address wasn't obtained from a GPS sensor.
 110             'oe'     => 'utf8',    // Output encoding.
 111             'output' => 'json',    // Output format.
 112             'gl'     => 'fr',      // Location preferences (addresses are in France by default).
 113             'q'      => $address,  // The queries address.
 114         );
 115
 116         return $globals->geocoder->gmaps_url . '?' . http_build_query($parameters);
 117     }
 118
 119     // Fetches JSON-encoded data from a Google Maps API url, and decode them.
 120     // Returns the json array on success, and null otherwise.
 121     private function getGeoJsonFromUrl($url) {
 122         global $globals;
 123
 124         // Prepare a backtrace object to log errors.
 125         $bt = null;
 126         if ($globals->debug & DEBUG_BT) {
 127             if (!isset(PlBacktrace::$bt['Geoloc'])) {
 128                 new PlBacktrace('Geoloc');
 129             }
 130             $bt = &PlBacktrace::$bt['Geoloc'];
 131             $bt->start($url);
 132         }
 133
 134         // Fetch the geocoding data.
 135         $rawData = file_get_contents($url);
 136         if (!$rawData) {
 137             if ($bt) {
 138                 $bt->stop(0, 'Could not retrieve geocoded address from GoogleMaps.');
 139             }
 140             return null;
 141         }
 142
 143         // Decode the JSON-encoded data, and check for their validity.
 144         $data = json_decode($rawData, true);
 145         if ($bt) {
 146             $bt->stop(count($data), null, $data);
 147         }
 148
 149         return $data;
 150     }
 151
 152     // Extracts the most appropriate placemark from the JSON data fetched from
 153     // Google Maps. Returns a Placemark array on success, and null otherwise. See
 154     // http://code.google.com/apis/maps/documentation/services.html#Geocoding_Structured
 155     // for details on the Placemark structure.
 156     private function getPlacemarkFromJson(array $data) {
 157         // Check for geocoding failures.
 158         if (!isset($data['Status']['code']) || $data['Status']['code'] != 200) {
 159             // TODO: handle non-200 codes in a better way, since the code might
 160             // indicate a temporary error on Google's side.
 161             return null;
 162         }
 163
 164         // Check that at least one placemark was found.
 165         if (count($data['Placemark']) == 0) {
 166             return null;
 167         }
 168
 169         // Extract the placemark with the best accuracy. This is not always the
 170         // best result (since the same address may yield two different placemarks).
 171         $result = $data['Placemark'][0];
 172         foreach ($data['Placemark'] as $place) {
 173             if ($place['AddressDetails']['Accuracy'] > $result['AddressDetails']['Accuracy']) {
 174                 $result = $place;
 175             }
 176         }
 177
 178         return $result;
 179     }
 180
 181     // Fills the address with the geocoded data
 182     private function fillAddressWithGeocoding(Address &$address, $geocodedData) {
 183         // The geocoded address three is
 184         // Country -> AdministrativeArea -> SubAdministrativeArea -> Locality -> Thoroughfare
 185         // with all the possible shortcuts
 186         // The address is formatted as xAL, or eXtensible Address Language, an international
 187         // standard for address formatting.
 188         // xAL documentation: http://www.oasis-open.org/committees/ciq/ciq.html#6
 189         $address->geocodedText = str_replace(', ', "\n", $geocodedData['address']);
 190         if (isset($geocodedData['AddressDetails']['Accuracy'])) {
 191             $address->accuracy = $geocodedData['AddressDetails']['Accuracy'];
 192         }
 193
 194         $currentPosition = $geocodedData['AddressDetails'];
 195         if (isset($currentPosition['Country'])) {
 196             $currentPosition    = $currentPosition['Country'];
 197             $address->countryId = $currentPosition['CountryNameCode'];
 198             $address->country   = $currentPosition['CountryName'];
 199         }
 200         if (isset($currentPosition['AdministrativeArea'])) {
 201             $currentPosition                 = $currentPosition['AdministrativeArea'];
 202             $address->administrativeAreaName = $currentPosition['AdministrativeAreaName'];
 203         }
 204         if (isset($currentPosition['SubAdministrativeArea'])) {
 205             $currentPosition                    = $currentPosition['SubAdministrativeArea'];
 206             $address->subAdministrativeAreaName = $currentPosition['SubAdministrativeAreaName'];
 207         }
 208         if (isset($currentPosition['Locality'])) {
 209             $currentPosition       = $currentPosition['Locality'];
 210             $address->localityName = $currentPosition['LocalityName'];
 211         }
 212         if (isset($currentPosition['Thoroughfare'])) {
 213             $address->thoroughfareName = $currentPosition['Thoroughfare']['ThoroughfareName'];
 214         }
 215         if (isset($currentPosition['PostalCode'])) {
 216             $address->postalCode = $currentPosition['PostalCode']['PostalCodeNumber'];
 217         }
 218
 219         // Gets coordinates.
 220         if (isset($geocodedData['Point']['coordinates'][0])) {
 221             $address->latitude = $geocodedData['Point']['coordinates'][0];
 222         }
 223         if (isset($geocodedData['Point']['coordinates'][1])) {
 224             $address->longitude = $geocodedData['Point']['coordinates'][1];
 225         }
 226         if (isset($geocodedData['ExtendedData']['LatLonBox']['north'])) {
 227             $address->north = $geocodedData['ExtendedData']['LatLonBox']['north'];
 228         }
 229         if (isset($geocodedData['ExtendedData']['LatLonBox']['south'])) {
 230             $address->south = $geocodedData['ExtendedData']['LatLonBox']['south'];
 231         }
 232         if (isset($geocodedData['ExtendedData']['LatLonBox']['east'])) {
 233             $address->east = $geocodedData['ExtendedData']['LatLonBox']['east'];
 234         }
 235         if (isset($geocodedData['ExtendedData']['LatLonBox']['west'])) {
 236             $address->west = $geocodedData['ExtendedData']['LatLonBox']['west'];
 237         }
 238     }
 239
 240     // Formats the text of the geocoded address using the unused data and
 241     // compares it to the given address. If they are too different, the user
 242     // will be asked to choose between them.
 243     private function formatAddress(Address &$address, $extraLines) {
 244         $same = true;
 245         if ($extraLines) {
 246             $address->geocodedText = $extraLines . "\n" . $address->geocodedText;
 247         }
 248         $address->geocodedPostalText = $this->getPostalAddress($address->geocodedText);
 249         $geoloc = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"),
 250                                           array('', "\n"), $address->geocodedText));
 251         $text   = strtoupper(preg_replace(array("/[0-9,\"'#~:;_\- ]/", "/\r\n/"),
 252                                           array('', "\n"), $address->text));
 253         $arrayGeoloc = explode("\n", $geoloc);
 254         $arrayText   = explode("\n", $text);
 255         $countGeoloc = count($arrayGeoloc);
 256         $countText   = count($arrayText);
 257
 258         $totalDistance = 0;
 259         if (($countText > $countGeoloc) || ($countText < $countGeoloc - 1)
 260             || (($countText == $countGeoloc - 1)
 261                 && ($arrayText[$countText - 1] == strtoupper($address->country)))) {
 262             $same = false;
 263         } else {
 264             for ($i = 0; $i < $countGeoloc && $i < $countText; ++$i) {
 265                 $lineDistance = levenshtein($arrayText[$i], trim($arrayGeoloc[$i]));
 266                 $totalDistance += $lineDistance;
 267                 if ($lineDistance > self::MAX_LINE_DISTANCE || $totalDistance > self::MAX_TOTAL_DISTANCE) {
 268                     $same = false;
 269                     break;
 270                 }
 271             }
 272         }
 273
 274         if ($same) {
 275             $address->geocodedText = null;
 276             $address->geocodedPostalText = null;
 277         } else {
 278             $address->geocodedText = str_replace("\n", "\r\n", $address->geocodedText);
 279             $address->geocodedPostalText = str_replace("\n", "\r\n", $address->geocodedPostalText);
 280         }
 281         $address->text = str_replace("\n", "\r\n", $address->text);
 282         $address->postalText = str_replace("\n", "\r\n", $address->postalText);
 283     }
 284
 285     // Returns the address formated for postal use.
 286     // The main rules are (cf AFNOR XPZ 10-011):
 287     // -everything in upper case;
 288     // -if there are more then than 38 characters in a line, split it;
 289     // -if there are more then than 32 characters in the description of the "street", use abbreviations.
 290     private function getPostalAddress($text) {
 291          static $abbreviations = array(
 292              'IMPASSE'   => 'IMP',
 293              'RUE'       => 'R',
 294              'AVENUE'    => 'AV',
 295              'BOULEVARD' => 'BVD',
 296              'ROUTE'     => 'R',
 297              'STREET'    => 'ST',
 298              'ROAD'      => 'RD',
 299              );
 300
 301         $text = strtoupper($text);
 302         $arrayText = explode("\n", $text);
 303         $postalText = '';
 304
 305         foreach ($arrayText as $i => $line) {
 306             $postalText .= (($i == 0) ? '' : "\n");
 307             if (($length = strlen($line)) > 32) {
 308                 $words = explode(' ', $line);
 309                 $count = 0;
 310                 foreach ($words as $word) {
 311                     if (isset($abbreviations[$word])) {
 312                         $word = $abbreviations[$word];
 313                     }
 314                     if ($count + ($wordLength = strlen($word)) <= 38) {
 315                         $postalText .= (($count == 0) ? '' : ' ') . $word;
 316                         $count += (($count == 0) ? 0 : 1) + $wordLength;
 317                     } else {
 318                         $postalText .= "\n" . $word;
 319                         $count = strlen($word);
 320                     }
 321                 }
 322             } else {
 323                 $postalText .= $line;
 324             }
 325         }
 326         return $postalText;
 327     }
 328
 329     // Trims the name of the real country if it contains an ISO 3166-1 non-country
 330     // item. For that purpose, we compare the last but one line of the address with
 331     // all non-country items of ISO 3166-1.
 332     private function getTextToGeocode($text)
 333     {
 334         $res = XDB::iterator('SELECT  country, countryFR
 335                                 FROM  geoloc_countries
 336                                WHERE  belongsTo IS NOT NULL');
 337         $countries = array();
 338         foreach ($res as $item) {
 339             $countries[] = $item[0];
 340             $countries[] = $item[1];
 341         }
 342         $textLines  = explode("\n", $text);
 343         $countLines = count($textLines);
 344         $needle     = strtoupper(trim($textLines[$countLines - 2]));
 345         $isPseudoCountry = false;
 346         foreach ($countries as $country) {
 347             if (strtoupper($country) == $needle) {
 348                 $isPseudoCountry = true;
 349                 break;
 350             }
 351         }
 352
 353         if ($isPseudoCountry) {
 354             return implode("\n", array_slice($textLines, 0, -1));
 355         }
 356         return $text;
 357     }
 358 }
 359
 360 // vim:set et sw=4 sts=4 sws=4 foldmethod=marker enc=utf-8:
 361 ?>