don't log IPv6 IPs for the moment

[nominatim.git] / utils / update.php

diff --git a/utils/update.php b/utils/update.php
index d729519a3ed5f3c9eb0d4f267094a702b7001eca..e35554888d4de0519571a6feca4afc1edddbaf8e 100755 (executable)
--- a/utils/update.php
+++ b/utils/update.php
@@ -7,7 +7,7 @@ ini_set('memory_limit', '800M');
 
 $aCMDOptions
 = array(
-   "Import / update / index osm data",
+   'Import / update / index osm data',
    array('help', 'h', 0, 1, 0, 0, false, 'Show Help'),
    array('quiet', 'q', 0, 1, 0, 0, 'bool', 'Quiet output'),
    array('verbose', 'v', 0, 1, 0, 0, 'bool', 'Verbose output'),
@@ -33,11 +33,13 @@ $aCMDOptions
    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('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');
@@ -64,7 +66,7 @@ if ($aResult['init-updates']) {
     if ($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.");
+        fail('replication URL not reachable.');
     }
     $sSetup = CONST_InstallPath.'/utils/setup.php';
     $iRet = -1;
@@ -75,7 +77,7 @@ if ($aResult['init-updates']) {
 
     $sDatabaseDate = getDatabaseDate($oDB);
     if ($sDatabaseDate === false) {
-        fail("Cannot determine date of database.");
+        fail('Cannot determine date of database.');
     }
     $sWindBack = strftime('%Y-%m-%dT%H:%M:%SZ', strtotime($sDatabaseDate) - (3*60*60));
 
@@ -89,9 +91,9 @@ if ($aResult['init-updates']) {
 
     pg_query($oDB->connection, 'TRUNCATE import_status');
     $sSQL = "INSERT INTO import_status (lastimportdate, sequence_id, indexed) VALUES('";
-    $sSQL .= $sDatabaseDate."',".$aOutput[0].", true)";
+    $sSQL .= $sDatabaseDate."',".$aOutput[0].', true)';
     if (!pg_query($oDB->connection, $sSQL)) {
-        fail("Could not enter sequence into database.");
+        fail('Could not enter sequence into database.');
     }
 
     echo "Done. Database updates will start at sequence $aOutput[0] ($sWindBack)\n";
@@ -118,7 +120,7 @@ if (isset($aResult['import-diff']) || isset($aResult['import-file'])) {
 }
 
 if ($aResult['calculate-postcodes']) {
-    info("Update postcodes centroids");
+    info('Update postcodes centroids');
     $sTemplate = file_get_contents(CONST_BasePath.'/sql/update-postcodes.sql');
     runSQLScript($sTemplate, true, true);
 }
@@ -129,25 +131,25 @@ $bUseOSMApi = isset($aResult['import-from-main-api']) && $aResult['import-from-m
 $sContentURL = '';
 if (isset($aResult['import-node']) && $aResult['import-node']) {
     if ($bUseOSMApi) {
-        $sContentURL = 'http://www.openstreetmap.org/api/0.6/node/'.$aResult['import-node'];
+        $sContentURL = 'https://www.openstreetmap.org/api/0.6/node/'.$aResult['import-node'];
     } else {
-        $sContentURL = 'http://overpass-api.de/api/interpreter?data=node('.$aResult['import-node'].');out%20meta;';
+        $sContentURL = 'https://overpass-api.de/api/interpreter?data=node('.$aResult['import-node'].');out%20meta;';
     }
 }
 
 if (isset($aResult['import-way']) && $aResult['import-way']) {
     if ($bUseOSMApi) {
-        $sContentURL = 'http://www.openstreetmap.org/api/0.6/way/'.$aResult['import-way'].'/full';
+        $sContentURL = 'https://www.openstreetmap.org/api/0.6/way/'.$aResult['import-way'].'/full';
     } else {
-        $sContentURL = 'http://overpass-api.de/api/interpreter?data=(way('.$aResult['import-way'].');node(w););out%20meta;';
+        $sContentURL = 'https://overpass-api.de/api/interpreter?data=(way('.$aResult['import-way'].');node(w););out%20meta;';
     }
 }
 
 if (isset($aResult['import-relation']) && $aResult['import-relation']) {
     if ($bUseOSMApi) {
-        $sContentURLsModifyXMLstr = 'http://www.openstreetmap.org/api/0.6/relation/'.$aResult['import-relation'].'/full';
+        $sContentURLsModifyXMLstr = 'https://www.openstreetmap.org/api/0.6/relation/'.$aResult['import-relation'].'/full';
     } else {
-        $sContentURL = 'http://overpass-api.de/api/interpreter?data=((rel('.$aResult['import-relation'].');way(r);node(w));node(r));out%20meta;';
+        $sContentURL = 'https://overpass-api.de/api/interpreter?data=((rel('.$aResult['import-relation'].');way(r);node(w));node(r));out%20meta;';
     }
 }
 
@@ -170,7 +172,7 @@ if ($aResult['deduplicate']) {
     $oDB =& getDB();
 
     if (getPostgresVersion($oDB) < 9.3) {
-        fail("ERROR: deduplicate is only currently supported in postgresql 9.3");
+        fail('ERROR: deduplicate is only currently supported in postgresql 9.3');
     }
 
     $sSQL = 'select partition from country_name order by country_code';
@@ -179,7 +181,7 @@ if ($aResult['deduplicate']) {
 
     // 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;
+        $sSQL = 'select count(*) from search_name_'.$sPartition;
         $nEntries = chksql($oDB->getOne($sSQL));
         if ($nEntries == 0) {
             unset($aPartitions[$i]);
@@ -187,14 +189,14 @@ 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 .= ' 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,";
-        $sSQL .= " (select count(*) from search_name where nameaddress_vector @> ARRAY[word_id]) as num";
+        echo 'Deduping '.$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($sSQL));
@@ -203,40 +205,46 @@ if ($aResult['deduplicate']) {
         $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']."]";
+            $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']."]";
+            $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']."]";
+            $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']."]";
+                $sSQL = 'update search_name_'.$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']."]";
+                $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'];
+            $sSQL = 'delete from word where word_id = '.$aRemove['word_id'];
             chksql($oDB->query($sSQL));
         }
     }
 }
 
+if ($aResult['recompute-word-counts']) {
+    info('Recompute frequency of full-word search terms');
+    $sTemplate = file_get_contents(CONST_BasePath.'/sql/words_from_search_name.sql');
+    runSQLScript($sTemplate, true, true);
+}
+
 if ($aResult['index']) {
     passthru(CONST_InstallPath.'/nominatim/nominatim -i -d '.$aDSNInfo['database'].' -P '.$aDSNInfo['port'].' -t '.$aResult['index-instances'].' -r '.$aResult['index-rank']);
 }
@@ -244,7 +252,7 @@ if ($aResult['index']) {
 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://wiki.openstreetmap.org/wiki/Nominatim/Installation#Updates)\n");
+        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");
     }
 
     $sImportFile = CONST_InstallPath.'/osmosischange.osc';
@@ -360,7 +368,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
             $oDB->query($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";
+            $sSQL = 'update import_status set indexed = true';
             $oDB->query($sSQL);
         }