X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/ec4e3c36af238ae3c31a2f5eb727fe2a22f6d054..7f55dcef3a1e316c1c25123f49e304639cb26382:/utils/update.php diff --git a/utils/update.php b/utils/update.php index 24095ef4..7e90016c 100644 --- a/utils/update.php +++ b/utils/update.php @@ -39,10 +39,9 @@ $aCMDOptions array('index-rank', '', 0, 1, 1, 1, 'int', 'Rank to start indexing from'), array('index-instances', '', 0, 1, 1, 1, 'int', 'Number of indexing instances (threads)'), - array('deduplicate', '', 0, 1, 0, 0, 'bool', 'Deduplicate tokens'), array('recompute-word-counts', '', 0, 1, 0, 0, 'bool', 'Compute frequency of full-word search terms'), array('update-address-levels', '', 0, 1, 0, 0, 'bool', 'Reimport address level configuration (EXPERT)'), - array('no-npi', '', 0, 1, 0, 0, 'bool', '(obsolete)'), + array('recompute-importance', '', 0, 1, 0, 0, 'bool', 'Recompute place importances') ); getCmdOpt($_SERVER['argv'], $aCMDOptions, $aResult, true, true); @@ -52,9 +51,11 @@ if (!isset($aResult['index-rank'])) $aResult['index-rank'] = 0; date_default_timezone_set('Etc/UTC'); -$oDB =& getDB(); +$oDB = new Nominatim\DB(); +$oDB->connect(); +$fPostgresVersion = $oDB->getPostgresVersion(); -$aDSNInfo = DB::parseDSN(CONST_Database_DSN); +$aDSNInfo = Nominatim\DB::parseDSN(CONST_Database_DSN); if (!isset($aDSNInfo['port']) || !$aDSNInfo['port']) $aDSNInfo['port'] = 5432; // cache memory to be used by osm2pgsql, should not be more than the available memory @@ -63,21 +64,59 @@ if ($iCacheMemory + 500 > getTotalMemoryMB()) { $iCacheMemory = getCacheMemoryMB(); echo "WARNING: resetting cache memory to $iCacheMemory\n"; } -$sOsm2pgsqlCmd = CONST_Osm2pgsql_Binary.' -klas --number-processes 1 -C '.$iCacheMemory.' -O gazetteer -d '.$aDSNInfo['database'].' -P '.$aDSNInfo['port']; + +$oOsm2pgsqlCmd = (new \Nominatim\Shell(CONST_Osm2pgsql_Binary)) + ->addParams('--hstore') + ->addParams('--latlong') + ->addParams('--append') + ->addParams('--slim') + ->addParams('--number-processes', 1) + ->addParams('--cache', $iCacheMemory) + ->addParams('--output', 'gazetteer') + ->addParams('--style', CONST_Import_Style) + ->addParams('--database', $aDSNInfo['database']) + ->addParams('--port', $aDSNInfo['port']); + +if (isset($aDSNInfo['hostspec']) && $aDSNInfo['hostspec']) { + $oOsm2pgsqlCmd->addParams('--host', $aDSNInfo['hostspec']); +} if (isset($aDSNInfo['username']) && $aDSNInfo['username']) { - $sOsm2pgsqlCmd .= ' -U ' . $aDSNInfo['username']; + $oOsm2pgsqlCmd->addParams('--user', $aDSNInfo['username']); +} +if (isset($aDSNInfo['password']) && $aDSNInfo['password']) { + $oOsm2pgsqlCmd->addEnvPair('PGPASSWORD', $aDSNInfo['password']); +} +if (!is_null(CONST_Osm2pgsql_Flatnode_File) && CONST_Osm2pgsql_Flatnode_File) { + $oOsm2pgsqlCmd->addParams('--flat-nodes', CONST_Osm2pgsql_Flatnode_File); +} +if ($fPostgresVersion >= 11.0) { + $oOsm2pgsqlCmd->addEnvPair( + 'PGOPTIONS', + '-c jit=off -c max_parallel_workers_per_gather=0' + ); +} + + +$oIndexCmd = (new \Nominatim\Shell(CONST_BasePath.'/nominatim/nominatim.py')) + ->addParams('--database', $aDSNInfo['database']) + ->addParams('--port', $aDSNInfo['port']) + ->addParams('--threads', $aResult['index-instances']); +if (!$aResult['quiet']) { + $oIndexCmd->addParams('--verbose'); +} +if ($aResult['verbose']) { + $oIndexCmd->addParams('--verbose'); } if (isset($aDSNInfo['hostspec']) && $aDSNInfo['hostspec']) { - $sOsm2pgsqlCmd .= ' -H ' . $aDSNInfo['hostspec']; + $oIndexCmd->addParams('--host', $aDSNInfo['hostspec']); +} +if (isset($aDSNInfo['username']) && $aDSNInfo['username']) { + $oIndexCmd->addParams('--username', $aDSNInfo['username']); } -$aProcEnv = null; if (isset($aDSNInfo['password']) && $aDSNInfo['password']) { - $aProcEnv = array_merge(array('PGPASSWORD' => $aDSNInfo['password']), $_ENV); + $oIndexCmd->addEnvPair('PGPASSWORD', $aDSNInfo['password']); } -if (!is_null(CONST_Osm2pgsql_Flatnode_File) && CONST_Osm2pgsql_Flatnode_File) { - $sOsm2pgsqlCmd .= ' --flat-nodes '.CONST_Osm2pgsql_Flatnode_File; -} if ($aResult['init-updates']) { // sanity check that the replication URL is correct @@ -94,9 +133,11 @@ if ($aResult['init-updates']) { echo "in your local settings file.\n\n"; fail('CONST_Pyosmium_Binary not configured'); } + $aOutput = 0; - $sCmd = CONST_Pyosmium_Binary.' --help'; - exec($sCmd, $aOutput, $iRet); + $oCMD = new \Nominatim\Shell(CONST_Pyosmium_Binary, '--help'); + exec($oCMD->escapedCmd(), $aOutput, $iRet); + if ($iRet != 0) { echo "Cannot execute pyosmium-get-changes.\n"; echo "Make sure you have pyosmium installed correctly\n"; @@ -110,28 +151,33 @@ if ($aResult['init-updates']) { 'enable-diff-updates' => true, 'verbose' => $aResult['verbose'] )); - $cSetup->connect(); $cSetup->createFunctions(); } $sDatabaseDate = getDatabaseDate($oDB); - if ($sDatabaseDate === false) { + if (!$sDatabaseDate) { fail('Cannot determine date of database.'); } $sWindBack = strftime('%Y-%m-%dT%H:%M:%SZ', strtotime($sDatabaseDate) - (3*60*60)); // get the appropriate state id $aOutput = 0; - $sCmd = CONST_Pyosmium_Binary.' -D '.$sWindBack.' --server '.CONST_Replication_Url; - exec($sCmd, $aOutput, $iRet); + $oCMD = (new \Nominatim\Shell(CONST_Pyosmium_Binary)) + ->addParams('--start-date', $sWindBack) + ->addParams('--server', CONST_Replication_Url); + + exec($oCMD->escapedCmd(), $aOutput, $iRet); if ($iRet != 0 || $aOutput[0] == 'None') { fail('Error running pyosmium tools'); } - pg_query($oDB->connection, 'TRUNCATE import_status'); + $oDB->exec('TRUNCATE import_status'); $sSQL = "INSERT INTO import_status (lastimportdate, sequence_id, indexed) VALUES('"; $sSQL .= $sDatabaseDate."',".$aOutput[0].', true)'; - if (!pg_query($oDB->connection, $sSQL)) { + + try { + $oDB->exec($sSQL); + } catch (\Nominatim\DatabaseError $e) { fail('Could not enter sequence into database.'); } @@ -139,13 +185,17 @@ if ($aResult['init-updates']) { } if ($aResult['check-for-updates']) { - $aLastState = chksql($oDB->getRow('SELECT sequence_id FROM import_status')); + $aLastState = $oDB->getRow('SELECT sequence_id FROM import_status'); if (!$aLastState['sequence_id']) { fail('Updates not set up. Please run ./utils/update.php --init-updates.'); } - system(CONST_BasePath.'/utils/check_server_for_updates.py '.CONST_Replication_Url.' '.$aLastState['sequence_id'], $iRet); + $oCmd = (new \Nominatim\Shell(CONST_BasePath.'/utils/check_server_for_updates.py')) + ->addParams(CONST_Replication_Url) + ->addParams($aLastState['sequence_id']); + $iRet = $oCmd->run(); + exit($iRet); } @@ -158,12 +208,12 @@ if (isset($aResult['import-diff']) || isset($aResult['import-file'])) { } // Import the file - $sCMD = $sOsm2pgsqlCmd.' '.$sNextFile; - echo $sCMD."\n"; - $iErrorLevel = runWithEnv($sCMD, $aProcEnv); + $oCMD = (clone $oOsm2pgsqlCmd)->addParams($sNextFile); + echo $oCMD->escapedCmd()."\n"; + $iRet = $oCMD->run(); - if ($iErrorLevel) { - fail("Error from osm2pgsql, $iErrorLevel\n"); + if ($iRet) { + fail("Error from osm2pgsql, $iRet\n"); } // Don't update the import status - we don't know what this file contains @@ -191,15 +241,15 @@ if (isset($aResult['import-way']) && $aResult['import-way']) { if ($bUseOSMApi) { $sContentURL = 'https://www.openstreetmap.org/api/0.6/way/'.$aResult['import-way'].'/full'; } else { - $sContentURL = 'https://overpass-api.de/api/interpreter?data=(way('.$aResult['import-way'].');node(w););out%20meta;'; + $sContentURL = 'https://overpass-api.de/api/interpreter?data=(way('.$aResult['import-way'].');%3E;);out%20meta;'; } } if (isset($aResult['import-relation']) && $aResult['import-relation']) { if ($bUseOSMApi) { - $sContentURLsModifyXMLstr = 'https://www.openstreetmap.org/api/0.6/relation/'.$aResult['import-relation'].'/full'; + $sContentURL = 'https://www.openstreetmap.org/api/0.6/relation/'.$aResult['import-relation'].'/full'; } else { - $sContentURL = 'https://overpass-api.de/api/interpreter?data=((rel('.$aResult['import-relation'].');way(r);node(w));node(r));out%20meta;'; + $sContentURL = 'https://overpass-api.de/api/interpreter?data=(rel(id:'.$aResult['import-relation'].');%3E;);out%20meta;'; } } @@ -210,82 +260,13 @@ if ($sContentURL) { if ($bHaveDiff) { // import generated change file - $sCMD = $sOsm2pgsqlCmd.' '.$sTemporaryFile; - echo $sCMD."\n"; - $iErrorLevel = runWithEnv($sCMD, $aProcEnv); - if ($iErrorLevel) { - fail("osm2pgsql exited with error level $iErrorLevel\n"); - } -} - -if ($aResult['deduplicate']) { - $oDB =& getDB(); - - if (getPostgresVersion($oDB) < 9.3) { - fail('ERROR: deduplicate is only currently supported in postgresql 9.3'); - } - - $sSQL = 'select partition from country_name order by country_code'; - $aPartitions = chksql($oDB->getCol($sSQL)); - $aPartitions[] = 0; - // we don't care about empty search_name_* partitions, they can't contain mentions of duplicates - foreach ($aPartitions as $i => $sPartition) { - $sSQL = 'select count(*) from search_name_'.$sPartition; - $nEntries = chksql($oDB->getOne($sSQL)); - if ($nEntries == 0) { - unset($aPartitions[$i]); - } - } + $oCMD = (clone $oOsm2pgsqlCmd)->addParams($sTemporaryFile); + echo $oCMD->escapedCmd()."\n"; - $sSQL = "select word_token,count(*) from word where substr(word_token, 1, 1) = ' '"; - $sSQL .= ' and class is null and type is null and country_code is null'; - $sSQL .= ' group by word_token having count(*) > 1 order by word_token'; - $aDuplicateTokens = chksql($oDB->getAll($sSQL)); - foreach ($aDuplicateTokens as $aToken) { - if (trim($aToken['word_token']) == '' || trim($aToken['word_token']) == '-') continue; - echo 'Deduping '.$aToken['word_token']."\n"; - $sSQL = 'select word_id,'; - $sSQL .= ' (select count(*) from search_name where nameaddress_vector @> ARRAY[word_id]) as num'; - $sSQL .= " from word where word_token = '".$aToken['word_token']; - $sSQL .= "' and class is null and type is null and country_code is null order by num desc"; - $aTokenSet = chksql($oDB->getAll($sSQL)); - - $aKeep = array_shift($aTokenSet); - $iKeepID = $aKeep['word_id']; - - foreach ($aTokenSet as $aRemove) { - $sSQL = 'update search_name set'; - $sSQL .= ' name_vector = array_replace(name_vector,'.$aRemove['word_id'].','.$iKeepID.'),'; - $sSQL .= ' nameaddress_vector = array_replace(nameaddress_vector,'.$aRemove['word_id'].','.$iKeepID.')'; - $sSQL .= ' where name_vector @> ARRAY['.$aRemove['word_id'].']'; - chksql($oDB->query($sSQL)); - - $sSQL = 'update search_name set'; - $sSQL .= ' nameaddress_vector = array_replace(nameaddress_vector,'.$aRemove['word_id'].','.$iKeepID.')'; - $sSQL .= ' where nameaddress_vector @> ARRAY['.$aRemove['word_id'].']'; - chksql($oDB->query($sSQL)); - - $sSQL = 'update location_area_country set'; - $sSQL .= ' keywords = array_replace(keywords,'.$aRemove['word_id'].','.$iKeepID.')'; - $sSQL .= ' where keywords @> ARRAY['.$aRemove['word_id'].']'; - chksql($oDB->query($sSQL)); - - foreach ($aPartitions as $sPartition) { - $sSQL = 'update search_name_'.$sPartition.' set'; - $sSQL .= ' name_vector = array_replace(name_vector,'.$aRemove['word_id'].','.$iKeepID.')'; - $sSQL .= ' where name_vector @> ARRAY['.$aRemove['word_id'].']'; - chksql($oDB->query($sSQL)); - - $sSQL = 'update location_area_country set'; - $sSQL .= ' keywords = array_replace(keywords,'.$aRemove['word_id'].','.$iKeepID.')'; - $sSQL .= ' where keywords @> ARRAY['.$aRemove['word_id'].']'; - chksql($oDB->query($sSQL)); - } - - $sSQL = 'delete from word where word_id = '.$aRemove['word_id']; - chksql($oDB->query($sSQL)); - } + $iRet = $oCMD->run(); + if ($iRet) { + fail("osm2pgsql exited with error level $iRet\n"); } } @@ -296,15 +277,15 @@ if ($aResult['recompute-word-counts']) { } if ($aResult['index']) { - $sCmd = CONST_InstallPath.'/nominatim/nominatim -i -d '.$aDSNInfo['database'].' -P '.$aDSNInfo['port'].' -t '.$aResult['index-instances'].' -r '.$aResult['index-rank']; - if (isset($aDSNInfo['hostspec']) && $aDSNInfo['hostspec']) { - $sCmd .= ' -H ' . $aDSNInfo['hostspec']; - } - if (isset($aDSNInfo['username']) && $aDSNInfo['username']) { - $sCmd .= ' -U ' . $aDSNInfo['username']; - } + $oCmd = (clone $oIndexCmd) + ->addParams('--minrank', $aResult['index-rank'], '-b'); + $oCmd->run(); + + $oCmd = (clone $oIndexCmd) + ->addParams('--minrank', $aResult['index-rank']); + $oCmd->run(); - runWithEnv($sCmd, $aProcEnv); + $oDB->exec('update import_status set indexed = true'); } if ($aResult['update-address-levels']) { @@ -313,27 +294,42 @@ if ($aResult['update-address-levels']) { $oAlParser->createTable($oDB, 'address_levels'); } +if ($aResult['recompute-importance']) { + echo "Updating importance values for database.\n"; + $oDB = new Nominatim\DB(); + $oDB->connect(); + + $sSQL = 'ALTER TABLE placex DISABLE TRIGGER ALL;'; + $sSQL .= 'UPDATE placex SET (wikipedia, importance) ='; + $sSQL .= ' (SELECT wikipedia, importance'; + $sSQL .= ' FROM compute_importance(extratags, country_code, osm_type, osm_id));'; + $sSQL .= 'UPDATE placex s SET wikipedia = d.wikipedia, importance = d.importance'; + $sSQL .= ' FROM placex d'; + $sSQL .= ' WHERE s.place_id = d.linked_place_id and d.wikipedia is not null'; + $sSQL .= ' and (s.wikipedia is null or s.importance < d.importance);'; + $sSQL .= 'ALTER TABLE placex ENABLE TRIGGER ALL;'; + $oDB->exec($sSQL); +} + if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { // if (strpos(CONST_Replication_Url, 'download.geofabrik.de') !== false && CONST_Replication_Update_Interval < 86400) { fail('Error: Update interval too low for download.geofabrik.de. ' . - "Please check install documentation (http://nominatim.org/release-docs/latest/Import-and-Update#setting-up-the-update-process)\n"); + "Please check install documentation (https://nominatim.org/release-docs/latest/admin/Import-and-Update#setting-up-the-update-process)\n"); } $sImportFile = CONST_InstallPath.'/osmosischange.osc'; - $sCMDDownload = CONST_Pyosmium_Binary.' --server '.CONST_Replication_Url.' -o '.$sImportFile.' -s '.CONST_Replication_Max_Diff_size; - $sCMDImport = $sOsm2pgsqlCmd.' '.$sImportFile; - $sCMDIndex = CONST_InstallPath.'/nominatim/nominatim -i -d '.$aDSNInfo['database'].' -P '.$aDSNInfo['port'].' -t '.$aResult['index-instances']; - if (isset($aDSNInfo['hostspec']) && $aDSNInfo['hostspec']) { - $sCMDIndex .= ' -H ' . $aDSNInfo['hostspec']; - } - if (isset($aDSNInfo['username']) && $aDSNInfo['username']) { - $sCMDIndex .= ' -U ' . $aDSNInfo['username']; - } + + $oCMDDownload = (new \Nominatim\Shell(CONST_Pyosmium_Binary)) + ->addParams('--server', CONST_Replication_Url) + ->addParams('--outfile', $sImportFile) + ->addParams('--size', CONST_Replication_Max_Diff_size); + + $oCMDImport = (clone $oOsm2pgsqlCmd)->addParams($sImportFile); while (true) { $fStartTime = time(); - $aLastState = chksql($oDB->getRow('SELECT *, EXTRACT (EPOCH FROM lastimportdate) as unix_ts FROM import_status')); + $aLastState = $oDB->getRow('SELECT *, EXTRACT (EPOCH FROM lastimportdate) as unix_ts FROM import_status'); if (!$aLastState['sequence_id']) { echo "Updates not set up. Please run ./utils/update.php --init-updates.\n"; @@ -345,7 +341,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { $sBatchEnd = $aLastState['lastimportdate']; $iEndSequence = $aLastState['sequence_id']; - if ($aLastState['indexed'] == 't') { + if ($aLastState['indexed']) { // Sleep if the update interval has not yet been reached. $fNextUpdate = $aLastState['unix_ts'] + CONST_Replication_Update_Interval; if ($fNextUpdate > $fStartTime) { @@ -359,11 +355,13 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { $fCMDStartTime = time(); $iNextSeq = (int) $aLastState['sequence_id']; unset($aOutput); - echo "$sCMDDownload -I $iNextSeq\n"; + + $oCMD = (clone $oCMDDownload)->addParams('--start-id', $iNextSeq); + echo $oCMD->escapedCmd()."\n"; if (file_exists($sImportFile)) { unlink($sImportFile); } - exec($sCMDDownload.' -I '.$iNextSeq, $aOutput, $iResult); + exec($oCMD->escapedCmd(), $aOutput, $iResult); if ($iResult == 3) { echo 'No new updates. Sleeping for '.CONST_Replication_Recheck_Interval." sec.\n"; @@ -379,7 +377,8 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { // get the newest object from the diff file $sBatchEnd = 0; $iRet = 0; - exec(CONST_BasePath.'/utils/osm_file_date.py '.$sImportFile, $sBatchEnd, $iRet); + $oCMD = new \Nominatim\Shell(CONST_BasePath.'/utils/osm_file_date.py', $sImportFile); + exec($oCMD->escapedCmd(), $sBatchEnd, $iRet); if ($iRet == 5) { echo "Diff file is empty. skipping import.\n"; if (!$aResult['import-osmosis-all']) { @@ -395,9 +394,11 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { // Import the file $fCMDStartTime = time(); - echo $sCMDImport."\n"; + + + echo $oCMDImport->escapedCmd()."\n"; unset($sJunk); - $iErrorLevel = runWithEnv($sCMDImport, $aProcEnv); + $iErrorLevel = $oCMDImport->run(); if ($iErrorLevel) { echo "Error executing osm2pgsql: $iErrorLevel\n"; exit($iErrorLevel); @@ -411,22 +412,31 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { $sSQL .= date('Y-m-d H:i:s', $fCMDStartTime)."','"; $sSQL .= date('Y-m-d H:i:s')."','import')"; var_Dump($sSQL); - chksql($oDB->query($sSQL)); + $oDB->exec($sSQL); // update the status $sSQL = "UPDATE import_status SET lastimportdate = '$sBatchEnd', indexed=false, sequence_id = $iEndSequence"; var_Dump($sSQL); - chksql($oDB->query($sSQL)); + $oDB->exec($sSQL); echo date('Y-m-d H:i:s')." Completed download step for $sBatchEnd in ".round((time()-$fCMDStartTime)/60, 2)." minutes\n"; } // Index file if (!$aResult['no-index']) { - $sThisIndexCmd = $sCMDIndex; $fCMDStartTime = time(); - echo "$sThisIndexCmd\n"; - $iErrorLevel = runWithEnv($sThisIndexCmd, $aProcEnv); + $oThisIndexCmd = clone($oIndexCmd); + $oThisIndexCmd->addParams('-b'); + echo $oThisIndexCmd->escapedCmd()."\n"; + $iErrorLevel = $oThisIndexCmd->run(); + if ($iErrorLevel) { + echo "Error: $iErrorLevel\n"; + exit($iErrorLevel); + } + + $oThisIndexCmd = clone($oIndexCmd); + echo $oThisIndexCmd->escapedCmd()."\n"; + $iErrorLevel = $oThisIndexCmd->run(); if ($iErrorLevel) { echo "Error: $iErrorLevel\n"; exit($iErrorLevel); @@ -434,15 +444,20 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { $sSQL = 'INSERT INTO import_osmosis_log'; $sSQL .= '(batchend, batchseq, batchsize, starttime, endtime, event)'; - $sSQL .= " values ('$sBatchEnd',$iEndSequence,$iFileSize,'"; + $sSQL .= " values ('$sBatchEnd',$iEndSequence,NULL,'"; $sSQL .= date('Y-m-d H:i:s', $fCMDStartTime)."','"; $sSQL .= date('Y-m-d H:i:s')."','index')"; var_Dump($sSQL); - $oDB->query($sSQL); + $oDB->exec($sSQL); echo date('Y-m-d H:i:s')." Completed index step for $sBatchEnd in ".round((time()-$fCMDStartTime)/60, 2)." minutes\n"; $sSQL = 'update import_status set indexed = true'; - $oDB->query($sSQL); + $oDB->exec($sSQL); + } else { + if ($aResult['import-osmosis-all']) { + echo "Error: --no-index cannot be used with continuous imports (--import-osmosis-all).\n"; + exit(1); + } } $fDuration = time() - $fStartTime;