Merge remote-tracking branch 'upstream/master'

[nominatim.git] / utils / update.php

diff --git a/utils/update.php b/utils/update.php
index 96789ef46b845ed01bc593b10e95199c8297e072..2fa773fe3dbbb46d53cb0ccb58cc4dfb50e6384b 100644 (file)
--- a/utils/update.php
+++ b/utils/update.php
@@ -42,12 +42,14 @@ $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('update-address-levels', '', 0, 1, 0, 0, 'bool', 'Reimport address level configuration (EXPERT)'),
    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)'),
   );
 
 getCmdOpt($_SERVER['argv'], $aCMDOptions, $aResult, true, true);
 
 if (!isset($aResult['index-instances'])) $aResult['index-instances'] = 1;
    array('no-npi', '', 0, 1, 0, 0, 'bool', '(obsolete)'),
   );
 
 getCmdOpt($_SERVER['argv'], $aCMDOptions, $aResult, true, true);
 
 if (!isset($aResult['index-instances'])) $aResult['index-instances'] = 1;

+

 if (!isset($aResult['index-rank'])) $aResult['index-rank'] = 0;
 
 date_default_timezone_set('Etc/UTC');
 if (!isset($aResult['index-rank'])) $aResult['index-rank'] = 0;
 
 date_default_timezone_set('Etc/UTC');


@@ -80,6 +82,14 @@ 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');


@@ -143,7 +153,7 @@ 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.');


@@ -231,13 +241,13 @@ if ($aResult['deduplicate']) {
     }
 
     $sSQL = 'select partition from country_name order by country_code';
     }
 
     $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]);
         }


@@ -246,7 +256,7 @@ 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";


@@ -254,7 +264,7 @@ 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'];


@@ -264,32 +274,32 @@ 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($oDB->exec($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($oDB->exec($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($oDB->exec($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($oDB->exec($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($oDB->exec($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($oDB->exec($sSQL));
+            $oDB->exec($sSQL);

         }
     }
 }
         }
     }
 }


@@ -301,7 +311,11 @@ 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'];
     }


@@ -320,6 +334,23 @@ if ($aResult['update-address-levels']) {
     $oAlParser->createTable($oDB, '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'] || $aResult['import-osmosis-all']) {
     //
     if (strpos(CONST_Replication_Url, 'download.geofabrik.de') !== false && CONST_Replication_Update_Interval < 86400) {
 if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
     //
     if (strpos(CONST_Replication_Url, 'download.geofabrik.de') !== false && CONST_Replication_Update_Interval < 86400) {


@@ -330,7 +361,10 @@ 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'];
     }


@@ -340,7 +374,7 @@ 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";


@@ -352,7 +386,7 @@ 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) {


@@ -418,12 +452,12 @@ 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($oDB->exec($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($oDB->exec($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";
         }