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