Merge pull request #1691 from lonvia/structured-query-via-cmdline

[nominatim.git] / utils / update.php

diff --git a/utils/update.php b/utils/update.php
old mode 100755 (executable)
new mode 100644 (file)
index 8756f2e..c1dc2ab
--- a/utils/update.php
+++ b/utils/update.php
@@ -1,10 +1,9 @@
-#!@PHP_BIN@ -Cq

 <?php
 
 <?php
 

-require_once(dirname(dirname(__FILE__)).'/settings/settings.php');

 require_once(CONST_BasePath.'/lib/init-cmd.php');
 require_once(CONST_BasePath.'/lib/setup_functions.php');
 require_once(CONST_BasePath.'/lib/setup/SetupClass.php');
 require_once(CONST_BasePath.'/lib/init-cmd.php');
 require_once(CONST_BasePath.'/lib/setup_functions.php');
 require_once(CONST_BasePath.'/lib/setup/SetupClass.php');

+require_once(CONST_BasePath.'/lib/setup/AddressLevelParser.php');

 
 ini_set('memory_limit', '800M');
 
 
 ini_set('memory_limit', '800M');
 


@@ -42,6 +41,8 @@ $aCMDOptions
 
    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('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('recompute-importance', '', 0, 1, 0, 0, 'bool', 'Recompute place importances'),

    array('no-npi', '', 0, 1, 0, 0, 'bool', '(obsolete)'),
   );
 
    array('no-npi', '', 0, 1, 0, 0, 'bool', '(obsolete)'),
   );
 


@@ -52,9 +53,10 @@ if (!isset($aResult['index-rank'])) $aResult['index-rank'] = 0;
 
 date_default_timezone_set('Etc/UTC');
 
 
 date_default_timezone_set('Etc/UTC');
 

-$oDB =& getDB();
+$oDB = new Nominatim\DB();
+$oDB->connect();

 
 

-$aDSNInfo = DB::parseDSN(CONST_Database_DSN);
+$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
 if (!isset($aDSNInfo['port']) || !$aDSNInfo['port']) $aDSNInfo['port'] = 5432;
 
 // cache memory to be used by osm2pgsql, should not be more than the available memory


@@ -63,7 +65,7 @@ if ($iCacheMemory + 500 > getTotalMemoryMB()) {
     $iCacheMemory = getCacheMemoryMB();
     echo "WARNING: resetting cache memory to $iCacheMemory\n";
 }
     $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'];
+$sOsm2pgsqlCmd = CONST_Osm2pgsql_Binary.' -klas --number-processes 1 -C '.$iCacheMemory.' -O gazetteer -S '.CONST_Import_Style.' -d '.$aDSNInfo['database'].' -P '.$aDSNInfo['port'];

 if (isset($aDSNInfo['username']) && $aDSNInfo['username']) {
     $sOsm2pgsqlCmd .= ' -U ' . $aDSNInfo['username'];
 }
 if (isset($aDSNInfo['username']) && $aDSNInfo['username']) {
     $sOsm2pgsqlCmd .= ' -U ' . $aDSNInfo['username'];
 }


@@ -79,6 +81,14 @@ if (!is_null(CONST_Osm2pgsql_Flatnode_File) && CONST_Osm2pgsql_Flatnode_File) {
     $sOsm2pgsqlCmd .= ' --flat-nodes '.CONST_Osm2pgsql_Flatnode_File;
 }
 
     $sOsm2pgsqlCmd .= ' --flat-nodes '.CONST_Osm2pgsql_Flatnode_File;
 }
 

+$sIndexCmd = CONST_BasePath.'/nominatim/nominatim.py';
+if (!$aResult['quiet']) {
+    $sIndexCmd .= ' -v';
+}
+if ($aResult['verbose']) {
+    $sIndexCmd .= ' -v';
+}
+

 if ($aResult['init-updates']) {
     // sanity check that the replication URL is correct
     $sBaseState = file_get_contents(CONST_Replication_Url.'/state.txt');
 if ($aResult['init-updates']) {
     // sanity check that the replication URL is correct
     $sBaseState = file_get_contents(CONST_Replication_Url.'/state.txt');


@@ -115,7 +125,7 @@ if ($aResult['init-updates']) {
     }
 
     $sDatabaseDate = getDatabaseDate($oDB);
     }
 
     $sDatabaseDate = getDatabaseDate($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));
         fail('Cannot determine date of database.');
     }
     $sWindBack = strftime('%Y-%m-%dT%H:%M:%SZ', strtotime($sDatabaseDate) - (3*60*60));


@@ -128,10 +138,13 @@ if ($aResult['init-updates']) {
         fail('Error running pyosmium tools');
     }
 
         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)';
     $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.');
     }
 
         fail('Could not enter sequence into database.');
     }
 


@@ -139,7 +152,7 @@ if ($aResult['init-updates']) {
 }
 
 if ($aResult['check-for-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.');
 
     if (!$aLastState['sequence_id']) {
         fail('Updates not set up. Please run ./utils/update.php --init-updates.');


@@ -219,20 +232,21 @@ if ($bHaveDiff) {
 }
 
 if ($aResult['deduplicate']) {
 }
 
 if ($aResult['deduplicate']) {

-    $oDB =& getDB();
+    $oDB = new Nominatim\DB();
+    $oDB->connect();

 
 

-    if (getPostgresVersion($oDB) < 9.3) {
+    if ($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';
         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 = $oDB->getCol($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;
     $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));
+        $nEntries = $oDB->getOne($sSQL);

         if ($nEntries == 0) {
             unset($aPartitions[$i]);
         }
         if ($nEntries == 0) {
             unset($aPartitions[$i]);
         }


@@ -241,7 +255,7 @@ if ($aResult['deduplicate']) {
     $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';
     $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));
+    $aDuplicateTokens = $oDB->getAll($sSQL);

     foreach ($aDuplicateTokens as $aToken) {
         if (trim($aToken['word_token']) == '' || trim($aToken['word_token']) == '-') continue;
         echo 'Deduping '.$aToken['word_token']."\n";
     foreach ($aDuplicateTokens as $aToken) {
         if (trim($aToken['word_token']) == '' || trim($aToken['word_token']) == '-') continue;
         echo 'Deduping '.$aToken['word_token']."\n";


@@ -249,7 +263,7 @@ if ($aResult['deduplicate']) {
         $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";
         $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));
+        $aTokenSet = $oDB->getAll($sSQL);

 
         $aKeep = array_shift($aTokenSet);
         $iKeepID = $aKeep['word_id'];
 
         $aKeep = array_shift($aTokenSet);
         $iKeepID = $aKeep['word_id'];


@@ -259,32 +273,32 @@ if ($aResult['deduplicate']) {
             $sSQL .= ' name_vector = array_replace(name_vector,'.$aRemove['word_id'].','.$iKeepID.'),';
             $sSQL .= ' nameaddress_vector = array_replace(nameaddress_vector,'.$aRemove['word_id'].','.$iKeepID.')';
             $sSQL .= ' where name_vector @> ARRAY['.$aRemove['word_id'].']';
             $sSQL .= ' name_vector = array_replace(name_vector,'.$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));
+            $oDB->exec($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'].']';
 
             $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));
+            $oDB->exec($sSQL);

 
             $sSQL = 'update location_area_country set';
             $sSQL .= ' keywords = array_replace(keywords,'.$aRemove['word_id'].','.$iKeepID.')';
             $sSQL .= ' where keywords @> ARRAY['.$aRemove['word_id'].']';
 
             $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));
+            $oDB->exec($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'].']';
 
             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));
+                $oDB->exec($sSQL);

 
                 $sSQL = 'update location_area_country set';
                 $sSQL .= ' keywords = array_replace(keywords,'.$aRemove['word_id'].','.$iKeepID.')';
                 $sSQL .= ' where keywords @> ARRAY['.$aRemove['word_id'].']';
 
                 $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));
+                $oDB->exec($sSQL);

             }
 
             $sSQL = 'delete from word where word_id = '.$aRemove['word_id'];
             }
 
             $sSQL = 'delete from word where word_id = '.$aRemove['word_id'];

-            chksql($oDB->query($sSQL));
+            $oDB->exec($sSQL);

         }
     }
 }
         }
     }
 }


@@ -296,7 +310,11 @@ if ($aResult['recompute-word-counts']) {
 }
 
 if ($aResult['index']) {
 }
 
 if ($aResult['index']) {

-    $sCmd = CONST_InstallPath.'/nominatim/nominatim -i -d '.$aDSNInfo['database'].' -P '.$aDSNInfo['port'].' -t '.$aResult['index-instances'].' -r '.$aResult['index-rank'];
+    $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['hostspec']) && $aDSNInfo['hostspec']) {
         $sCmd .= ' -H ' . $aDSNInfo['hostspec'];
     }


@@ -305,6 +323,31 @@ if ($aResult['index']) {
     }
 
     runWithEnv($sCmd, $aProcEnv);
     }
 
     runWithEnv($sCmd, $aProcEnv);

+
+    $oDB->exec('update import_status set indexed = true');
+}
+
+if ($aResult['update-address-levels']) {
+    echo 'Updating address levels from '.CONST_Address_Level_Config.".\n";
+    $oAlParser = new \Nominatim\Setup\AddressLevelParser(CONST_Address_Level_Config);
+    $oAlParser->createTable($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 ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {


@@ -317,7 +360,10 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
     $sImportFile = CONST_InstallPath.'/osmosischange.osc';
     $sCMDDownload = CONST_Pyosmium_Binary.' --server '.CONST_Replication_Url.' -o '.$sImportFile.' -s '.CONST_Replication_Max_Diff_size;
     $sCMDImport = $sOsm2pgsqlCmd.' '.$sImportFile;
     $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'];
+    $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['hostspec']) && $aDSNInfo['hostspec']) {
         $sCMDIndex .= ' -H ' . $aDSNInfo['hostspec'];
     }


@@ -327,7 +373,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
 
     while (true) {
         $fStartTime = time();
 
     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";
 
         if (!$aLastState['sequence_id']) {
             echo "Updates not set up. Please run ./utils/update.php --init-updates.\n";


@@ -339,7 +385,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
         $sBatchEnd = $aLastState['lastimportdate'];
         $iEndSequence = $aLastState['sequence_id'];
 
         $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) {
             // Sleep if the update interval has not yet been reached.
             $fNextUpdate = $aLastState['unix_ts'] + CONST_Replication_Update_Interval;
             if ($fNextUpdate > $fStartTime) {


@@ -405,12 +451,12 @@ 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);
             $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);
 
             // 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";
         }
 
             echo date('Y-m-d H:i:s')." Completed download step for $sBatchEnd in ".round((time()-$fCMDStartTime)/60, 2)." minutes\n";
         }
 


@@ -428,15 +474,20 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
 
             $sSQL = 'INSERT INTO import_osmosis_log';
             $sSQL .= '(batchend, batchseq, batchsize, starttime, endtime, event)';
 
             $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);
             $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';
             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;
         }
 
         $fDuration = time() - $fStartTime;