From 23bb56130a44517859c1a702a7cb12fad60e06e2 Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Fri, 6 Feb 2026 19:04:18 +0900
Subject: [PATCH 01/10] Add bind_param_attributes to ::DBD::MariaDB

---
 lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm b/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm
index cec472b..03ac1db 100644
--- a/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm
+++ b/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm
@@ -7,4 +7,16 @@ use base qw( Data::ObjectDriver::Driver::DBD::mysql );
 
 sub fetch_id { $_[3]->{mariadb_insertid} || $_[3]->{insertid} }
 
+sub bind_param_attributes {
+    my ($dbd, $data_type) = @_;
+    if ($data_type) {
+        if ($data_type eq 'blob') {
+            return DBI::SQL_BINARY;
+        } elsif ($data_type eq 'binchar') {
+            return DBI::SQL_BINARY;
+        }
+    }
+    return;
+}
+
 1;

From 4e2e585e7d08d10727f7c8d52d0aaed3da62d057 Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Fri, 6 Feb 2026 19:04:30 +0900
Subject: [PATCH 02/10] Add a test

---
 t/10-resultset-blob.t | 46 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 46 insertions(+)
 create mode 100644 t/10-resultset-blob.t

diff --git a/t/10-resultset-blob.t b/t/10-resultset-blob.t
new file mode 100644
index 0000000..a0776c8
--- /dev/null
+++ b/t/10-resultset-blob.t
@@ -0,0 +1,46 @@
+# $Id: 01-col-inheritance.t 989 2005-09-23 19:58:01Z btrott $
+
+use strict;
+use warnings;
+
+use lib 't/lib';
+
+$Data::ObjectDriver::DEBUG = 0;
+use Test::More;
+use DodTestUtil;
+BEGIN { eval { require Crypt::URandom; 1 } or plan skip_all => 'requires Crypt::URandom' }
+
+BEGIN { DodTestUtil->check_driver }
+
+plan tests => 3;
+
+setup_dbs({
+    global => [ qw( wines ) ],
+});
+
+use Wine;
+use Storable;
+
+my $wine = Wine->new;
+$wine->name("Saumur Champigny, Le Grand Clos 2001");
+$wine->rating(4);
+
+## generate some binary data (SQL_BLOB / MEDIUMBLOB)
+my $binary = Crypt::URandom::urandom(300);
+$wine->content($binary);
+ok($wine->save, 'Object saved successfully');
+
+my $iter;
+
+$iter = Data::ObjectDriver::Iterator->new(sub {});
+my $wine_id = $wine->id;
+undef $wine;
+$wine = Wine->lookup($wine_id); 
+
+ok $wine;
+ok $wine->content eq $binary;
+
+# TODO: bulk_insert doesn't support blob yet. We need to change some of its API so that we can call column_def in each dbd's bulk_insert
+
+disconnect_all($wine);
+teardown_dbs(qw( global ));

From f5bc24b421601be0eeac49117193273c9b0ee3cf Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Fri, 6 Feb 2026 19:05:04 +0900
Subject: [PATCH 03/10] Add Crypt::URandom to cpanfile for testing

---
 cpanfile | 1 +
 1 file changed, 1 insertion(+)

diff --git a/cpanfile b/cpanfile
index f4639d0..18f6ed9 100644
--- a/cpanfile
+++ b/cpanfile
@@ -23,6 +23,7 @@ on develop => sub {
 on test => sub {
     requires 'version';
     requires 'Tie::IxHash';
+    requires 'Crypt::URandom';
 };
 
 feature 'test_sqlite', 'Test SQLite' => sub {

From 709edda622d1d35c8b531858cb86762836dab72b Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Tue, 10 Feb 2026 09:33:17 +0900
Subject: [PATCH 04/10] Better to use a core module

---
 cpanfile              | 2 +-
 t/10-resultset-blob.t | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/cpanfile b/cpanfile
index 18f6ed9..3ea7334 100644
--- a/cpanfile
+++ b/cpanfile
@@ -23,7 +23,7 @@ on develop => sub {
 on test => sub {
     requires 'version';
     requires 'Tie::IxHash';
-    requires 'Crypt::URandom';
+    requires 'Digest::SHA';
 };
 
 feature 'test_sqlite', 'Test SQLite' => sub {
diff --git a/t/10-resultset-blob.t b/t/10-resultset-blob.t
index a0776c8..ff1039e 100644
--- a/t/10-resultset-blob.t
+++ b/t/10-resultset-blob.t
@@ -8,7 +8,7 @@ use lib 't/lib';
 $Data::ObjectDriver::DEBUG = 0;
 use Test::More;
 use DodTestUtil;
-BEGIN { eval { require Crypt::URandom; 1 } or plan skip_all => 'requires Crypt::URandom' }
+BEGIN { eval { require Digest::SHA; 1 } or plan skip_all => 'requires Digest::SHA' }
 
 BEGIN { DodTestUtil->check_driver }
 
@@ -26,7 +26,7 @@ $wine->name("Saumur Champigny, Le Grand Clos 2001");
 $wine->rating(4);
 
 ## generate some binary data (SQL_BLOB / MEDIUMBLOB)
-my $binary = Crypt::URandom::urandom(300);
+my $binary = Digest::SHA::sha1("binary");
 $wine->content($binary);
 ok($wine->save, 'Object saved successfully');
 

From 90ef1082a621cba42019a1bc9f978cded2e988c8 Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Tue, 10 Feb 2026 22:43:13 +0900
Subject: [PATCH 05/10] Pass a hashref of attributes that need special care as
 the last parameter to bulk_insert of the dbd

---
 lib/Data/ObjectDriver/Driver/DBI.pm | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/lib/Data/ObjectDriver/Driver/DBI.pm b/lib/Data/ObjectDriver/Driver/DBI.pm
index 449f6eb..ed84c50 100644
--- a/lib/Data/ObjectDriver/Driver/DBI.pm
+++ b/lib/Data/ObjectDriver/Driver/DBI.pm
@@ -644,7 +644,15 @@ sub bulk_insert {
     my $tbl  = $driver->table_for($class);
     my @db_cols =  map {$dbd->db_column_name($tbl, $_) } @{$cols};
 
-    return $dbd->bulk_insert($dbh, $tbl, \@db_cols, $data);
+    my %attrs;
+    my $col_defs = $class->properties->{column_defs};
+    for my $col (@$cols) {
+        my $type = $col_defs->{$col} || 'char';
+        my $attr = $dbd->bind_param_attributes($type) or next;
+        $attrs{$col} = $attr;
+    }
+
+    return $dbd->bulk_insert($dbh, $tbl, \@db_cols, $data, \%attrs);
 }
 
 sub begin_work {

From 937504f8aaab8826c4d8ecdd51c87ddb5e9f68e4 Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Tue, 10 Feb 2026 22:43:45 +0900
Subject: [PATCH 06/10] Add customized bulk_insert to ::DBD::MariaDB

---
 lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm | 39 +++++++++++++++++++++
 1 file changed, 39 insertions(+)

diff --git a/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm b/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm
index 03ac1db..f01fac3 100644
--- a/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm
+++ b/lib/Data/ObjectDriver/Driver/DBD/MariaDB.pm
@@ -3,6 +3,7 @@
 package Data::ObjectDriver::Driver::DBD::MariaDB;
 use strict;
 use warnings;
+use Carp;
 use base qw( Data::ObjectDriver::Driver::DBD::mysql );
 
 sub fetch_id { $_[3]->{mariadb_insertid} || $_[3]->{insertid} }
@@ -19,4 +20,42 @@ sub bind_param_attributes {
     return;
 }
 
+sub bulk_insert {
+    my $dbd = shift;
+    my $dbh = shift;
+    my $table = shift;
+
+    my $cols = shift;
+    my $rows_ref = shift;
+    my $attrs = shift || {};
+
+    croak "Usage bulk_insert(dbd, dbh, table, columnref, rowsref)"
+        unless (defined $dbd && defined $dbh && defined $table && defined $cols &&
+                defined $rows_ref);
+
+    return 0e0 if (scalar(@{$rows_ref}) == 0);
+
+    my $sql = "INSERT INTO $table ("  . join(',', @{$cols}) . ") VALUES\n";
+
+    my $one_data_row = "(" . (join ',', (('?') x @$cols)) . ")";
+    my $ph = join ",", (($one_data_row) x @$rows_ref);
+    $sql .= $ph;
+
+    # For now just write all data, at some point we need to lookup the
+    # maximum packet size for SQL
+
+    if (%$attrs) {
+        my $sth = $dbh->prepare($sql);
+        my $i = 1;
+        for my $row (@$rows_ref) {
+            for (my $j = 0; $j < @$cols; $j++) {
+                $sth->bind_param($i++, $row->[$j], $attrs->{$cols->[$j]});
+            }
+        }
+        $sth->execute;
+    } else {
+        return $dbh->do($sql, undef, map { @$_ } @$rows_ref);
+    }
+}
+
 1;

From 58eac4e465a177bb40e4d0bf8a5c59cec25a26ff Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Tue, 10 Feb 2026 22:44:56 +0900
Subject: [PATCH 07/10] Tweak ::DBD::Pg::bulk_insert not to use COPY statement
 if some of the attributes require binary handling

---
 lib/Data/ObjectDriver/Driver/DBD/Pg.pm | 34 ++++++++++++++++++++------
 1 file changed, 26 insertions(+), 8 deletions(-)

diff --git a/lib/Data/ObjectDriver/Driver/DBD/Pg.pm b/lib/Data/ObjectDriver/Driver/DBD/Pg.pm
index 86ad114..5dc1392 100644
--- a/lib/Data/ObjectDriver/Driver/DBD/Pg.pm
+++ b/lib/Data/ObjectDriver/Driver/DBD/Pg.pm
@@ -51,15 +51,33 @@ sub bulk_insert {
 
     my $cols = shift;
     my $rows_ref = shift;
-
-    my $sql = "COPY $table (" . join(',', @{$cols}) . ') from stdin';
-
-    $dbh->do($sql);
-    foreach my $row (@{$rows_ref}) {
-        my $line = join("\t", map {$_ || '\N'} @{$row});
-        $dbh->pg_putline("$line\n");
+    my $attrs = shift || {};
+
+    if (%$attrs) {
+        my $sql = "INSERT INTO $table ("  . join(',', @{$cols}) . ") VALUES\n";
+
+        my $one_data_row = "(" . (join ',', (('?') x @$cols)) . ")";
+        my $ph = join ",", (($one_data_row) x @$rows_ref);
+        $sql .= $ph;
+
+        my $sth = $dbh->prepare($sql);
+        my $i = 1;
+        for my $row (@$rows_ref) {
+            for (my $j = 0; $j < @$cols; $j++) {
+                $sth->bind_param($i++, $row->[$j], $attrs->{$cols->[$j]});
+            }
+        }
+        $sth->execute;
+    } else {
+        my $sql = "COPY $table (" . join(',', @{$cols}) . ') from stdin';
+
+        $dbh->do($sql);
+        foreach my $row (@{$rows_ref}) {
+            my $line = join("\t", map {$_ || '\N'} @{$row});
+            $dbh->pg_putline("$line\n");
+        }
+        return $dbh->pg_endcopy();
     }
-    return $dbh->pg_endcopy();
 }
 
 sub map_error_code {

From 8ee9c84971419bcb066bdc8ca8a72a9df19fb9e2 Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Tue, 10 Feb 2026 22:45:28 +0900
Subject: [PATCH 08/10] Tweak bulk_insert for SQLite and Oracle

---
 lib/Data/ObjectDriver/Driver/DBD/Oracle.pm | 7 ++++++-
 lib/Data/ObjectDriver/Driver/DBD/SQLite.pm | 7 ++++++-
 2 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/lib/Data/ObjectDriver/Driver/DBD/Oracle.pm b/lib/Data/ObjectDriver/Driver/DBD/Oracle.pm
index 5020193..b1f1bf5 100644
--- a/lib/Data/ObjectDriver/Driver/DBD/Oracle.pm
+++ b/lib/Data/ObjectDriver/Driver/DBD/Oracle.pm
@@ -68,6 +68,7 @@ sub bulk_insert {
     my $table    = shift;
     my $cols     = shift;
     my $rows_ref = shift;
+    my $attrs    = shift || {};
 
     my $sql = "INSERT INTO $table("
               . join(',', @$cols)
@@ -76,7 +77,11 @@ sub bulk_insert {
               .  ")";
     my $sth = $dbh->prepare($sql);
     foreach my $row (@{ $rows_ref || []}) {
-        $sth->execute(@$row);
+        my $i = 1;
+        for (my $j = 0; $j < @$cols; $j++) {
+            $sth->bind_param($i++, $row->[$j], $attrs->{$cols->[$j]});
+        }
+        $sth->execute;
     }
     return 1;
 }
diff --git a/lib/Data/ObjectDriver/Driver/DBD/SQLite.pm b/lib/Data/ObjectDriver/Driver/DBD/SQLite.pm
index de65575..18840f2 100644
--- a/lib/Data/ObjectDriver/Driver/DBD/SQLite.pm
+++ b/lib/Data/ObjectDriver/Driver/DBD/SQLite.pm
@@ -46,13 +46,18 @@ sub bulk_insert {
 
     my $cols = shift;
     my $rows_ref = shift;
+    my $attrs = shift || {};
 
     my $sql = "INSERT INTO $table("  . join(',', @{$cols}) . ") VALUES (" . join(',',  map {'?'} @{$cols}) .  ")\n";
 
     my $sth = $dbh->prepare($sql);
 
     foreach my $row (@{$rows_ref}) {
-	$sth->execute(@{$row});
+        my $i = 1;
+        for (my $j = 0; $j < @$cols; $j++) {
+            $sth->bind_param($i++, $row->[$j], $attrs->{$cols->[$j]});
+        }
+        $sth->execute;
     }
 
     # For now just write all data, at some point we need to lookup the

From 60da6c6faa47a6642105992adaf0e2893c08caf4 Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Tue, 10 Feb 2026 22:46:01 +0900
Subject: [PATCH 09/10] Add tests for bulk insertion of binary

---
 t/10-resultset-blob.t | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/t/10-resultset-blob.t b/t/10-resultset-blob.t
index ff1039e..1ced70f 100644
--- a/t/10-resultset-blob.t
+++ b/t/10-resultset-blob.t
@@ -12,7 +12,7 @@ BEGIN { eval { require Digest::SHA; 1 } or plan skip_all => 'requires Digest::SH
 
 BEGIN { DodTestUtil->check_driver }
 
-plan tests => 3;
+plan tests => 5;
 
 setup_dbs({
     global => [ qw( wines ) ],
@@ -40,7 +40,13 @@ $wine = Wine->lookup($wine_id);
 ok $wine;
 ok $wine->content eq $binary;
 
-# TODO: bulk_insert doesn't support blob yet. We need to change some of its API so that we can call column_def in each dbd's bulk_insert
+my @names = qw(Margaux Latour);
+Wine->bulk_insert([qw(name content)], [ map {[$_, Digest::SHA::sha1($_)]} @names ]);
+
+for my $name (@names) {
+    my ($found) = Wine->search({name => $name});
+    ok $found->content eq Digest::SHA::sha1($name);
+}
 
 disconnect_all($wine);
 teardown_dbs(qw( global ));

From e0b218443715ac4740facad0b80c191a8247a9ed Mon Sep 17 00:00:00 2001
From: Kenichi Ishigaki <ishigaki@cpan.org>
Date: Wed, 11 Feb 2026 00:36:45 +0900
Subject: [PATCH 10/10] Better to use db_col

---
 lib/Data/ObjectDriver/Driver/DBI.pm | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/lib/Data/ObjectDriver/Driver/DBI.pm b/lib/Data/ObjectDriver/Driver/DBI.pm
index ed84c50..c722f17 100644
--- a/lib/Data/ObjectDriver/Driver/DBI.pm
+++ b/lib/Data/ObjectDriver/Driver/DBI.pm
@@ -642,14 +642,16 @@ sub bulk_insert {
     # pass this directly to the backend DBD
     my $dbh = $driver->rw_handle($class->properties->{db});
     my $tbl  = $driver->table_for($class);
-    my @db_cols =  map {$dbd->db_column_name($tbl, $_) } @{$cols};
 
+    my @db_cols;
     my %attrs;
     my $col_defs = $class->properties->{column_defs};
     for my $col (@$cols) {
+        my $db_col = $dbd->db_column_name($tbl, $col);
+        push @db_cols, $db_col;
         my $type = $col_defs->{$col} || 'char';
         my $attr = $dbd->bind_param_attributes($type) or next;
-        $attrs{$col} = $attr;
+        $attrs{$db_col} = $attr;
     }
 
     return $dbd->bulk_insert($dbh, $tbl, \@db_cols, $data, \%attrs);