From 1a98b6480fb4e8beb9091326112d8ebc6ca1e7cd Mon Sep 17 00:00:00 2001 From: Jacek Kowalski <Jacek@jacekk.info> Date: Mon, 18 Feb 2019 23:46:55 +0000 Subject: [PATCH] Don't overwrite non-conflicting vehicle mappings from previous runs --- parse.php | 78 ++++++++++++++++++++++---- lib/database.php | 47 +++++++++++++++ 2 files changed, 113 insertions(+), 12 deletions(-) diff --git a/lib/database.php b/lib/database.php new file mode 100644 index 0000000..48ad047 --- /dev/null +++ b/lib/database.php @@ -0,0 +1,47 @@ +<?php +class Database { + private $pdo; + private $getByIdStatement; + private $getByNumStatement; + private $addStatement; + + public function __construct($file) { + $this->pdo = new PDO('sqlite:'.$file); + $this->pdo->setAttribute(PDO::ATTR_ERRMODE, PDO::ERRMODE_EXCEPTION); + + $this->pdo->query('CREATE TABLE IF NOT EXISTS vehicles ( + id INT PRIMARY KEY, + num INT UNIQUE, + weight INT + )'); + + $this->getByIdStatement = $this->pdo->prepare('SELECT num, weight FROM vehicles WHERE id=? LIMIT 1'); + $this->getByNumStatement = $this->pdo->prepare('SELECT id, weight FROM vehicles WHERE num=? LIMIT 1'); + $this->addStatement = $this->pdo->prepare('INSERT OR REPLACE INTO vehicles (id, num, weight) VALUES (?, ?, ?)'); + } + + public function getById($id) { + $this->getByIdStatement->execute([$id]); + return $this->getByIdStatement->fetch(); + } + + public function getByNum($num) { + $st = $this->getByNumStatement->execute([(int)substr($num, 2)]); + return $this->getByNumStatement->fetch(); + } + + public function clear() { + $this->pdo->query('DELETE FROM vehicles'); + } + + public function add($id, $num, $weight) { + $this->addStatement->execute([$id, $num, $weight]); + } + + public function addMapping($mapping) { + $weight = count($mapping); + foreach($mapping as $id => $vehicle) { + $this->add($id, (int)substr($vehicle['num'], 2), $weight); + } + } +} diff --git a/parse.php b/parse.php index 02e5f28..432b3ef 100644 --- a/parse.php +++ b/parse.php @@ -1,4 +1,5 @@ <?php +require_once(__DIR__.'/lib/database.php'); require_once(__DIR__.'/lib/fetch.php'); require_once(__DIR__.'/lib/mapper.php'); @@ -10,6 +11,7 @@ 'gtfs_file' => 'VehiclePositions_A.pb', 'ttss' => 'http://91.223.13.70/internetservice/geoserviceDispatcher/services/vehicleinfo/vehicles', 'ttss_file' => 'vehicles_A.json', + 'database' => 'mapping_A.sqlite3', 'result' => 'mapping_A.json', ], ]; @@ -17,32 +19,84 @@ foreach($sources as $name => $source) { $logger = new Monolog\Logger('fetch_'.$name); try { + foreach(['gtfs_file', 'ttss_file', 'database', 'result'] as $field) { + $source[$field] = __DIR__.'/data/'.$source[$field]; + } + $source['result_temp'] = $source['result'].'.tmp'; + $logger->info('Fetching '.$name.' position data from FTP...'); - $updated = ftp_fetch_if_newer($source['gtfs'], __DIR__.'/data/'.$source['gtfs_file']); + $updated = ftp_fetch_if_newer($source['gtfs'], $source['gtfs_file']); if(!$updated) { $logger->info('Nothing to do, remote file not newer than local one'); continue; } $logger->info('Fetching '.$name.' positions from TTSS...'); - fetch($source['ttss'], __DIR__.'/data/'.$source['ttss_file']); + fetch($source['ttss'],$source['ttss_file']); $logger->info('Loading data...'); $mapper = new Mapper(); - $mapper->loadTTSS(__DIR__.'/data/'.$source['ttss_file']); - $mapper->loadGTFS(__DIR__.'/data/'.$source['gtfs_file']); + $mapper->loadTTSS($source['ttss_file']); + $mapper->loadGTFS($source['gtfs_file']); + + $db = new Database($source['database']); $logger->info('Finding correct offset...'); $offset = $mapper->findOffset(); - if($offset) { - $logger->info('Got offset '.$offset.', creating mapping...'); - $mapping = $mapper->getMapping($offset); - $json = json_encode($mapping); - if(!file_put_contents(__DIR__.'/data/'.$source['result'].'.tmp', $json)) { - throw new Exception('Result save failed'); - } - rename(__DIR__.'/data/'.$source['result'].'.tmp', __DIR__.'/data/'.$source['result']); + if(!$offset) { + throw new Exception('Offset not found'); } + + $logger->info('Got offset '.$offset.', creating mapping...'); + $mapping = $mapper->getMapping($offset); + + $logger->info('Checking the data for correctness...'); + $weight = count($mapping); + $replace = 0; + $ignore = 0; + foreach($mapping as $id => $vehicle) { + $dbVehicle = $db->getById($id); + if($dbVehicle) { + if((int)substr($vehicle['num'], 2) != (int)$dbVehicle['num']) { + if($weight > $dbVehicle['weight']) { + $replace += 1; + } else { + $ignore += 1; + } + } + continue; + } + + $dbVehicle = $db->getByNum($vehicle['num']); + if($dbVehicle && $dbVehicle['id'] != $id) { + $replace += 1; + } + } + $logger->info('Weight: '.$weight.', ignore: '.$ignore.', replace: '.$replace); + + $previousMapping = NULL; + if($ignore > 0 && $ignore >= $replace) { + throw new Exception('Ignoring result due to better data already present'); + } elseif($replace > 0) { + $logger->warn('Replacing DB data with the mapping'); + $db->clear(); + } else { + $previousMapping = @json_decode(@file_get_contents($source['result']), TRUE); + } + + $db->addMapping($mapping); + + if(is_array($previousMapping)) { + $logger->info('Merging previous data with current mapping'); + $mapping = $previousMapping + $mapping; + ksort($mapping); + } + + $json = json_encode($mapping); + if(!file_put_contents($source['result_temp'], $json)) { + throw new Exception('Result save failed'); + } + rename($source['result_temp'], $source['result']); $logger->info('Finished'); } catch(Throwable $e) { $logger->error($e->getMessage(), ['exception' => $e, 'exception_string' => (string)$e]); -- Gitblit v1.9.1