deduplicate words: do not look for duplicates in empty partitions

[nominatim.git] / utils / update.php

diff --git a/utils/update.php b/utils/update.php
index 27c538c17facda16db1aaf6860df22b33a110602..d8ac134add94d52c2894a59a1e59569f2296750f 100755 (executable)
--- a/utils/update.php
+++ b/utils/update.php
@@ -156,22 +156,36 @@ if ($bHaveDiff) {
 }
 
 if ($aResult['deduplicate']) {
-    //
-    if (getPostgresVersion() < 9.3) {
+    $oDB =& getDB();
+
+    if (getPostgresVersion($oDB) < 9.3) {
         fail("ERROR: deduplicate is only currently supported in postgresql 9.3");
     }
 
-    $oDB =& getDB();
     $sSQL = 'select partition from country_name order by country_code';
     $aPartitions = chksql($oDB->getCol($sSQL));
     $aPartitions[] = 0;
 
-    $sSQL = "select word_token,count(*) from word where substr(word_token, 1, 1) = ' ' and class is null and type is null and country_code is null group by word_token having count(*) > 1 order by word_token";
+    // we don't care about empty search_name_* artitions, they can't contain mentions of duplicates
+    foreach ($aPartitions as $i => $sPartition) {
+        $sSQL = "select count(*) from search_name_".$sPartition;
+        $nEntries = chksql($oDB->getOne($sSQL));
+        if ($nEntries == 0) {
+            unset($aPartitions[$i]);
+        }
+    }
+
+    $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";
-        $sSQL = "select word_id,(select count(*) from search_name where nameaddress_vector @> ARRAY[word_id]) as num from word where word_token = '".$aToken['word_token']."' and class is null and type is null and country_code is null order by num desc";
+        $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);
@@ -237,7 +251,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
             // First check if there are new updates published (except for minutelies - there's always new diffs to process)
             if (CONST_Replication_Update_Interval > 60) {
                 unset($aReplicationLag);
-                exec($sCMDCheckReplicationLag, $aReplicationLag, $iErrorLevel); 
+                exec($sCMDCheckReplicationLag, $aReplicationLag, $iErrorLevel);
                 while ($iErrorLevel > 0 || $aReplicationLag[0] < 1) {
                     if ($iErrorLevel) {
                         echo "Error: $iErrorLevel. ";
@@ -247,7 +261,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
                     }
                     sleep(CONST_Replication_Recheck_Interval);
                     unset($aReplicationLag);
-                    exec($sCMDCheckReplicationLag, $aReplicationLag, $iErrorLevel); 
+                    exec($sCMDCheckReplicationLag, $aReplicationLag, $iErrorLevel);
                 }
                 // There are new replication files - use osmosis to download the file
                 echo "\n".date('Y-m-d H:i:s')." Replication Delay is ".$aReplicationLag[0]."\n";
@@ -326,6 +340,7 @@ if ($aResult['import-osmosis'] || $aResult['import-osmosis-all']) {
     }
 }
 
+
 function getosmosistimestamp($sOsmosisConfigDirectory)
 {
     $sStateFile = file_get_contents($sOsmosisConfigDirectory.'/state.txt');