[Bps-public-commit] dbix-searchbuilder branch, master, updated. 1.65_01-1-g8e843ec

[Alex Vandiver]

The branch, master has been updated
       via  8e843ec1b0592b24c51667ce407c325f7974e199 (commit)
      from  c7a9b718589f6907e43e2095b6b76b45d9b259bc (commit)

Summary of changes:
 lib/DBIx/SearchBuilder/Handle.pm | 3 ---
 1 file changed, 3 deletions(-)

- Log -----------------------------------------------------------------
commit 8e843ec1b0592b24c51667ce407c325f7974e199
Author: Alex Vandiver <alexmv at bestpractical.com>
Date:   Tue Aug 19 16:58:33 2014 -0400

    Stop manually disabling the "UTF8" flag before executing queries
    
    f9471ffb added this code to work around the newly added UTF-8 support in
    DBD::Oracle 1.16.  It is not known what the specifics of the failure it
    was attempting to avoid are -- the best guess is that Oracle failed when
    handed invalid UTF8 codepoints.
    
    However, blindly turning the "UTF8" flag off is no better, as it may or
    may not result in UTF-8 encoded data.  The correct fix is to be
    consistent about the data (bytes vs characters) that is passed to the
    database.

diff --git a/lib/DBIx/SearchBuilder/Handle.pm b/lib/DBIx/SearchBuilder/Handle.pm
index d974356..35b0b68 100755
--- a/lib/DBIx/SearchBuilder/Handle.pm
+++ b/lib/DBIx/SearchBuilder/Handle.pm
@@ -574,9 +574,6 @@ sub SimpleQuery {
             delete $bhash->{'value'};
             $sth->bind_param( $bind_idx + 1, undef, $bhash );
         }
-        # Some databases, such as Oracle fail to cope if it's a perl utf8
-        # string. they desperately want bytes.
-         Encode::_utf8_off($bind_values[$bind_idx]);
     }
 
     my $basetime;

-----------------------------------------------------------------------