I think I would deal with the tricky parts by doing something less tricky. If your example reflects reality your database table has a primary key. Why not just use the primary key in your update()?. That would make your while loop something like this:
my $abstUP = SQL::Abstract->new(); while( my $row = $sth->fetchrow_hashref() ){ print "$0 : Processing row: ".Dumper($row)."\n"; # $subexec is a subref which takes the row as a hashref # and returns a modified clone of it to be inserted into the db # so, newrow and row have the same keys exactly my $newrow = $subexec->($row, $colnames); print "$0 : result is: ".Dumper($newrow)."\n"; my $newrow_only_colnames = { %{ $newrow }{@$colnames} }; # Update the row my($stmt, @bind) = $abstUP->update($tablename, $newrow_only_col +names, { id => $row->{id} } ); print "executing SQL:\n$stmt\nwith binds:\n@bind\n"; my $sth_upd = $dbh->prepare($stmt); if( ! $sth_upd->execute(@bind) ){ $dbh->rollback(); die "update (and rolling back): ".$dbh->errstr; } }
Note that the above uses $sth_upd for the handle to the update statement. This was done because my refactor does the update while the select is still active, so re-using $sth would clobber the select.
wyant
In reply to Re: SQL: Update column(s) value with extra WHERE
by Anonymous Monk
in thread SQL: Update column(s) value with extra WHERE
by bliako
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |