Changeset 32570 for main

Timestamp:

2018-11-05T16:43:30+13:00 (5 years ago)

Author:

ak19

Message:

1. Bugfix for when renaming an imported doc and incrementally-rebuilding the collection: need to do set_OID() during GreenstoneSQLPlugin::close_doc() in order for the old doc oid to be removed from the index, else the old doc filenames/titles are still present when browsing rebuilt collection. 2. To be on the safe side, overridden remove_one() in GS SQL Plugin now remembers to call super version whose implementation is ultimately inherited from BaseImporter, just in case this (ever) does something relevant.

File:

• main/trunk/greenstone2/perllib/plugins/GreenstoneSQLPlugin.pm (modified) (3 diffs)

main/trunk/greenstone2/perllib/plugins/GreenstoneSQLPlugin.pm

@@ -218 +218 @@
     my ($file, $oids, $archivedir) = @_;
 
+    my $rv = $self->SUPER::remove_one(@_);
+    
     print STDERR "@@@ IN SQLPLUG::REMOVE_ONE: $file\n";
     
-    #return undef unless $self->can_process_this_file($file); # NO, DON'T DO THIS: 
+    #return undef unless $self->can_process_this_file($file); # NO, DON'T DO THIS (inherited remove_one behaviour) HERE: 
            # WE DON'T CARE IF IT'S AN IMAGE FILE THAT WAS DELETED.
-           # WE CARE ABOUT REMOVING THE DOCOID OF THAT IMAGE FILE FROM THE DB
-
+           # WE CARE ABOUT REMOVING THE DOCOID OF THAT IMAGE FILE FROM THE SQL DB
+           # SO DON'T RETURN IF CAN'T_PROCESS_THIS_FILE
+    
     print STDERR "*****************************\nAsked to remove_one oid\n***********************\n";
     
     my $gs_sql = $self->{'gs_sql'} || return 0; # couldn't make the connection or no db etc
-    if(scalar @$oids > 1) {
-    print STDERR "TODO: We now have reason to optimise GreenstoneSQLPlugin::remove_one() by using dbi::prepare() + dbi::execute() instead of dbi::do().\n";
+    if(scalar (@$oids) > 1) {
+    print STDERR "TODO: As there's more than 1 docid, we now have reason to optimise GreenstoneSQLPlugin::remove_one() by using dbi::prepare() + dbi::execute() instead of dbi::do().\n";
     }
     
@@ -242 +245 @@
     }
     }
-    return 1;
+    return $rv;
 }
 
@@ -292 +295 @@
     
     # For now, we have access to doc_obj (until just before super::close_document() terminates)
-    
-    # no need to call $self->{'doc_obj'}->set_OID($oid);
-    # because either the OID is stored in the SQL db as meta 'Identifier' alongside other metadata
-    # or it's stored in the doc.xml as metadata 'Identifier' alongside other metadata
-    # Either way, Identifier meta will be read into the docobj automatically with other meta.
-
+
+    # OID parsed of docsql.xml file does need to be set on $doc_obj, as noticed in this case:
+    # when a doc in import is renamed, and you do incremental import, it is marked for reindexing 
+    # (reindexing is implemented by this plugin as a delete followed by add into the sql db).
+    # In that case, UNLESS you set the OID at this stage, the old deleted doc id (for the old doc
+    # name) continues to exist in the index at the end of incremental rebuilding if you were to
+    # browse the rebuilt collection by files/titles. So unless you set the OID here, the deleted
+    # doc oids will still be listed in the index.
+    $self->{'doc_obj'}->set_OID($oid);
+    
     print STDERR "   GreenstoneSQLPlugin processing doc $oid (reading into docobj from SQL db)\n"
     if $self->{'verbosity'} > 0;