Changeset 32538 for main/trunk/greenstone2/perllib/plugins/GreenstoneSQLPlugin.pm

Timestamp:

2018-10-24T20:41:01+13:00 (5 years ago)

Author:

ak19

Message:

Previous commit message meant to be: string names of strings shared by GS SQL Plugin and Plugout have been changed in strings.properties to indicate both modules used them. Current commit: Some tidying up the new GreenstoneSQLPlugin and moving the select statements from there into gssql.pm.

File:

• main/trunk/greenstone2/perllib/plugins/GreenstoneSQLPlugin.pm (modified) (6 diffs)

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

@@ -124 +124 @@
 }
 
-# need to store current section 
-#sub xml_start_tag {
-#    my $self = shift(@_);
-#    $self->SUPER::xml_start_tag(@_);
-#    $self->{'stored_section'} = $self->{'section'};
-#    $self->{'stored_section_level'} = $self->{'section_level'}
-#}
 
 # TODO Q: Why are there 3 passes when we're only indexing at doc and section level (2 passes)?
@@ -145 +138 @@
 
     my $oid = $self->{'doc_oid'}; # we stored current doc's OID during sub read() 
-    print $outhandle "@@@@@@ OID of document (meta|text) to be read in from DB: $oid\n"
+    print $outhandle "==== OID of document (meta|text) to be read in from DB: $oid\n"
     if $self->{'verbosity'} > 1;
 
-#    print STDERR "@@@@ section: |" . $self->{'stored_section'} . "|\n";
-#    print STDERR "@@@@ section_level: " . $self->{'stored_section_level'} . "\n";
-    
-    # for now, we have access to doc_obj (until just before super::close_document() terminates)
-    #$self->{'doc_obj'}->set_OID($oid);
+    
+    # For now, we have access to doc_obj (until just before super::close_document() terminates)
+    
+    $self->{'doc_obj'}->set_OID($oid); # complex method. Is this necessary, since we just want to write meta and txt for the docobj to index?
+    
+    # checking that complicated looking method set_OID() hasn't modified oid
+    if($oid ne $self->{'doc_obj'}->get_OID()) {
+    print STDERR "@@@@ WARNING: OID after setting on doc_obj = " . $self->{'doc_obj'}->get_OID() . " and is not the same as original OID $oid from docsqloid.xml filename\n";
+    }
+
     
     # TODO: This function is called on a per doc.xml file basis
@@ -164 +162 @@
     # read in meta for the collection (i.e. select * from <col>_metadata table
 
-    my $meta_table = $gs_sql->get_metadata_table_name();
-    
+    my $sth = $gs_sql->select_from_metatable_matching_docid($oid);  
+    print $outhandle "### stmt: ".$sth->{'Statement'}."\n" if $self->{'verbosity'} > 1;
+
+    print $outhandle "----------SQL DB contains meta-----------\n" if $self->{'verbosity'} > 1;
     # https://www.effectiveperlprogramming.com/2010/07/set-custom-dbi-error-handlers/
-    my $sth = $gs_sql->get_db_handle()->prepare(qq{SELECT * FROM $meta_table WHERE did = ?});
-    print $outhandle "### stmt: ".$sth->{'Statement'}."\n"
-        if $self->{'verbosity'} > 1;
-    $sth->execute( $oid ); # will print msg on fail
-
-    print $outhandle "----------SQL DB contains meta-----------\n" if $self->{'verbosity'} > 1;
     while( my @row = $sth->fetchrow_array() ) {     
         #print $outhandle "row: @row\n";
@@ -196 +190 @@
     
     
-    my $sth = $gs_sql->get_db_handle()->prepare(qq{SELECT * FROM $fulltxt_table WHERE did = ?});
-    print $outhandle "### stmt: ".$sth->{'Statement'}."\n"
-        if $self->{'verbosity'} > 1;
-    $sth->execute( $oid ); # will print msg on fail
+    my $sth = $gs_sql->select_from_texttable_matching_docid($oid);
+    print $outhandle "### stmt: ".$sth->{'Statement'}."\n" if $self->{'verbosity'} > 1;
 
     print $outhandle "----------\nSQL DB contains txt entries for-----------\n"
@@ -226 +218 @@
 
 
-# TODO: only want to work with sql db if buildcol.pl
+# TODO: only want to work with sql db if buildcol.pl. Unfortunately, also runs on import.pl
 # call init() not begin() because there can be multiple plugin passes
 # and init() should be called before all passes:
 # one for doc level and another for section level indexing
+# This way, we can connect to the SQL database once per buildcol run.
 sub init {
     my ($self) = shift (@_);
@@ -290 +283 @@
 }
 
-
+# This method also runs on import.pl if gs_sql has a value. But we just want to run it on buildcol
 # Call deinit() not end() because there can be multiple plugin passes:
 # one for doc level and another for section level indexing
 # and deinit() should be called before all passes
+# This way, we can close the SQL database once per buildcol run.
 sub deinit {
     my ($self) = shift (@_);