X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/a8711ab0134adb14d39dd895482bf68c43c58d29..c3238682a8867edfb3c052ca4d2d27b3926ad17a:/utils/update.php diff --git a/utils/update.php b/utils/update.php index c1dc2ab9..5bc2d73d 100644 --- a/utils/update.php +++ b/utils/update.php @@ -1,9 +1,9 @@ connect(); +$fPostgresVersion = $oDB->getPostgresVersion(); -$aDSNInfo = Nominatim\DB::parseDSN(CONST_Database_DSN); +$aDSNInfo = Nominatim\DB::parseDSN(getSetting('DATABASE_DSN')); if (!isset($aDSNInfo['port']) || !$aDSNInfo['port']) $aDSNInfo['port'] = 5432; // cache memory to be used by osm2pgsql, should not be more than the available memory @@ -65,52 +70,89 @@ if ($iCacheMemory + 500 > getTotalMemoryMB()) { $iCacheMemory = getCacheMemoryMB(); echo "WARNING: resetting cache memory to $iCacheMemory\n"; } -$sOsm2pgsqlCmd = CONST_Osm2pgsql_Binary.' -klas --number-processes 1 -C '.$iCacheMemory.' -O gazetteer -S '.CONST_Import_Style.' -d '.$aDSNInfo['database'].' -P '.$aDSNInfo['port']; -if (isset($aDSNInfo['username']) && $aDSNInfo['username']) { - $sOsm2pgsqlCmd .= ' -U ' . $aDSNInfo['username']; -} + +$oOsm2pgsqlCmd = (new \Nominatim\Shell(getOsm2pgsqlBinary())) + ->addParams('--hstore') + ->addParams('--latlong') + ->addParams('--append') + ->addParams('--slim') + ->addParams('--with-forward-dependencies', 'false') + ->addParams('--log-progress', 'true') + ->addParams('--number-processes', 1) + ->addParams('--cache', $iCacheMemory) + ->addParams('--output', 'gazetteer') + ->addParams('--style', getImportStyle()) + ->addParams('--database', $aDSNInfo['database']) + ->addParams('--port', $aDSNInfo['port']); + if (isset($aDSNInfo['hostspec']) && $aDSNInfo['hostspec']) { - $sOsm2pgsqlCmd .= ' -H ' . $aDSNInfo['hostspec']; + $oOsm2pgsqlCmd->addParams('--host', $aDSNInfo['hostspec']); +} +if (isset($aDSNInfo['username']) && $aDSNInfo['username']) { + $oOsm2pgsqlCmd->addParams('--user', $aDSNInfo['username']); } -$aProcEnv = null; if (isset($aDSNInfo['password']) && $aDSNInfo['password']) { - $aProcEnv = array_merge(array('PGPASSWORD' => $aDSNInfo['password']), $_ENV); + $oOsm2pgsqlCmd->addEnvPair('PGPASSWORD', $aDSNInfo['password']); } - -if (!is_null(CONST_Osm2pgsql_Flatnode_File) && CONST_Osm2pgsql_Flatnode_File) { - $sOsm2pgsqlCmd .= ' --flat-nodes '.CONST_Osm2pgsql_Flatnode_File; +if (getSetting('FLATNODE_FILE')) { + $oOsm2pgsqlCmd->addParams('--flat-nodes', getSetting('FLATNODE_FILE')); } +if ($fPostgresVersion >= 11.0) { + $oOsm2pgsqlCmd->addEnvPair( + 'PGOPTIONS', + '-c jit=off -c max_parallel_workers_per_gather=0' + ); +} + -$sIndexCmd = CONST_BasePath.'/nominatim/nominatim.py'; +$oIndexCmd = (new \Nominatim\Shell(CONST_DataDir.'/nominatim/nominatim.py')) + ->addParams('--database', $aDSNInfo['database']) + ->addParams('--port', $aDSNInfo['port']) + ->addParams('--threads', $aResult['index-instances']); if (!$aResult['quiet']) { - $sIndexCmd .= ' -v'; + $oIndexCmd->addParams('--verbose'); } if ($aResult['verbose']) { - $sIndexCmd .= ' -v'; + $oIndexCmd->addParams('--verbose'); +} +if (isset($aDSNInfo['hostspec']) && $aDSNInfo['hostspec']) { + $oIndexCmd->addParams('--host', $aDSNInfo['hostspec']); +} +if (isset($aDSNInfo['username']) && $aDSNInfo['username']) { + $oIndexCmd->addParams('--username', $aDSNInfo['username']); +} +if (isset($aDSNInfo['password']) && $aDSNInfo['password']) { + $oIndexCmd->addEnvPair('PGPASSWORD', $aDSNInfo['password']); } +$sPyosmiumBin = getSetting('PYOSMIUM_BINARY'); +$sBaseURL = getSetting('REPLICATION_URL'); + + if ($aResult['init-updates']) { // sanity check that the replication URL is correct - $sBaseState = file_get_contents(CONST_Replication_Url.'/state.txt'); + $sBaseState = file_get_contents($sBaseURL.'/state.txt'); if ($sBaseState === false) { echo "\nCannot find state.txt file at the configured replication URL.\n"; echo "Does the URL point to a directory containing OSM update data?\n\n"; fail('replication URL not reachable.'); } // sanity check for pyosmium-get-changes - if (!CONST_Pyosmium_Binary) { - echo "\nCONST_Pyosmium_Binary not configured.\n"; + if (!$sPyosmiumBin) { + echo "\nNOMINATIM_PYOSMIUM_BINARY not configured.\n"; echo "You need to install pyosmium and set up the path to pyosmium-get-changes\n"; - echo "in your local settings file.\n\n"; - fail('CONST_Pyosmium_Binary not configured'); + echo "in your local .env file.\n\n"; + fail('NOMINATIM_PYOSMIUM_BINARY not configured'); } + $aOutput = 0; - $sCmd = CONST_Pyosmium_Binary.' --help'; - exec($sCmd, $aOutput, $iRet); + $oCMD = new \Nominatim\Shell($sPyosmiumBin, '--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"; - echo "and have set up CONST_Pyosmium_Binary to point to pyosmium-get-changes.\n"; + echo "and have set up NOMINATIM_PYOSMIUM_BINARY to point to pyosmium-get-changes.\n"; fail('pyosmium-get-changes not found or not usable'); } @@ -120,7 +162,6 @@ if ($aResult['init-updates']) { 'enable-diff-updates' => true, 'verbose' => $aResult['verbose'] )); - $cSetup->connect(); $cSetup->createFunctions(); } @@ -132,8 +173,11 @@ if ($aResult['init-updates']) { // get the appropriate state id $aOutput = 0; - $sCmd = CONST_Pyosmium_Binary.' -D '.$sWindBack.' --server '.CONST_Replication_Url; - exec($sCmd, $aOutput, $iRet); + $oCMD = (new \Nominatim\Shell($sPyosmiumBin)) + ->addParams('--start-date', $sWindBack) + ->addParams('--server', $sBaseURL); + + exec($oCMD->escapedCmd(), $aOutput, $iRet); if ($iRet != 0 || $aOutput[0] == 'None') { fail('Error running pyosmium tools'); } @@ -158,7 +202,11 @@ if ($aResult['check-for-updates']) { 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); + $oCmd = (new \Nominatim\Shell(CONST_BinDir.'/check_server_for_updates.py')) + ->addParams($sBaseURL) + ->addParams($aLastState['sequence_id']); + $iRet = $oCmd->run(); + exit($iRet); } @@ -171,12 +219,12 @@ 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, $iRet\n"); } // Don't update the import status - we don't know what this file contains @@ -184,11 +232,11 @@ if (isset($aResult['import-diff']) || isset($aResult['import-file'])) { if ($aResult['calculate-postcodes']) { info('Update postcodes centroids'); - $sTemplate = file_get_contents(CONST_BasePath.'/sql/update-postcodes.sql'); + $sTemplate = file_get_contents(CONST_DataDir.'/sql/update-postcodes.sql'); runSQLScript($sTemplate, true, true); } -$sTemporaryFile = CONST_BasePath.'/data/osmosischange.osc'; +$sTemporaryFile = CONST_InstallDir.'/osmosischange.osc'; $bHaveDiff = false; $bUseOSMApi = isset($aResult['import-from-main-api']) && $aResult['import-from-main-api']; $sContentURL = ''; @@ -204,15 +252,15 @@ if (isset($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;'; } } @@ -223,113 +271,38 @@ 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 = new Nominatim\DB(); - $oDB->connect(); + $oCMD = (clone $oOsm2pgsqlCmd)->addParams($sTemporaryFile); + echo $oCMD->escapedCmd()."\n"; - if ($oDB->getPostgresVersion() < 9.3) { - fail('ERROR: deduplicate is only currently supported in postgresql 9.3'); - } - - $sSQL = 'select partition from country_name order by country_code'; - $aPartitions = $oDB->getCol($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 = $oDB->getOne($sSQL); - if ($nEntries == 0) { - unset($aPartitions[$i]); - } - } - - $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 = $oDB->getAll($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 = $oDB->getAll($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'].']'; - $oDB->exec($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'].']'; - $oDB->exec($sSQL); - - $sSQL = 'update location_area_country set'; - $sSQL .= ' keywords = array_replace(keywords,'.$aRemove['word_id'].','.$iKeepID.')'; - $sSQL .= ' where keywords @> ARRAY['.$aRemove['word_id'].']'; - $oDB->exec($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'].']'; - $oDB->exec($sSQL); - - $sSQL = 'update location_area_country set'; - $sSQL .= ' keywords = array_replace(keywords,'.$aRemove['word_id'].','.$iKeepID.')'; - $sSQL .= ' where keywords @> ARRAY['.$aRemove['word_id'].']'; - $oDB->exec($sSQL); - } - - $sSQL = 'delete from word where word_id = '.$aRemove['word_id']; - $oDB->exec($sSQL); - } + $iRet = $oCMD->run(); + if ($iRet) { + fail("osm2pgsql exited with error level $iRet\n"); } } if ($aResult['recompute-word-counts']) { info('Recompute frequency of full-word search terms'); - $sTemplate = file_get_contents(CONST_BasePath.'/sql/words_from_search_name.sql'); + $sTemplate = file_get_contents(CONST_DataDir.'/sql/words_from_search_name.sql'); runSQLScript($sTemplate, true, true); } if ($aResult['index']) { - $sCmd = $sIndexCmd - .' -d '.$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'); + $oCmd->run(); - runWithEnv($sCmd, $aProcEnv); + $oCmd = (clone $oIndexCmd) + ->addParams('--minrank', $aResult['index-rank']); + $oCmd->run(); $oDB->exec('update import_status set indexed = true'); } if ($aResult['update-address-levels']) { - echo 'Updating address levels from '.CONST_Address_Level_Config.".\n"; - $oAlParser = new \Nominatim\Setup\AddressLevelParser(CONST_Address_Level_Config); + $sAddressLevelConfig = getSettingConfig('ADDRESS_LEVEL_CONFIG', 'address-levels.json'); + echo 'Updating address levels from '.$sAddressLevelConfig.".\n"; + $oAlParser = new \Nominatim\Setup\AddressLevelParser($sAddressLevelConfig); $oAlParser->createTable($oDB, 'address_levels'); } @@ -352,24 +325,19 @@ if ($aResult['recompute-importance']) { if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { // - if (strpos(CONST_Replication_Url, 'download.geofabrik.de') !== false && CONST_Replication_Update_Interval < 86400) { + if (strpos($sBaseURL, 'download.geofabrik.de') !== false && getSetting('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 = $sIndexCmd - .' -d '.$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']; - } + $sImportFile = CONST_InstallDir.'/osmosischange.osc'; + + $oCMDDownload = (new \Nominatim\Shell($sPyosmiumBin)) + ->addParams('--server', $sBaseURL) + ->addParams('--outfile', $sImportFile) + ->addParams('--size', getSetting('REPLICATION_MAX_DIFF')); + + $oCMDImport = (clone $oOsm2pgsqlCmd)->addParams($sImportFile); while (true) { $fStartTime = time(); @@ -387,7 +355,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { if ($aLastState['indexed']) { // Sleep if the update interval has not yet been reached. - $fNextUpdate = $aLastState['unix_ts'] + CONST_Replication_Update_Interval; + $fNextUpdate = $aLastState['unix_ts'] + getSetting('REPLICATION_UPDATE_INTERVAL'); if ($fNextUpdate > $fStartTime) { $iSleepTime = $fNextUpdate - $fStartTime; echo "Waiting for next update for $iSleepTime sec."; @@ -399,15 +367,18 @@ if ($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"; - sleep(CONST_Replication_Recheck_Interval); + $sSleep = getSetting('REPLICATION_RECHECK_INTERVAL'); + echo 'No new updates. Sleeping for '.$sSleep." sec.\n"; + sleep($sSleep); } elseif ($iResult != 0) { echo 'ERROR: updates failed.'; exit($iResult); @@ -419,7 +390,8 @@ if ($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_BinDir.'/osm_file_date.py', $sImportFile); + exec($oCMD->escapedCmd(), $sBatchEnd, $iRet); if ($iRet == 5) { echo "Diff file is empty. skipping import.\n"; if (!$aResult['import-osmosis-all']) { @@ -435,9 +407,11 @@ if ($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); @@ -462,11 +436,20 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) { // Index file if (!$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);