do_sqlserver 0.10.1-java

data/CONNECTING.markdown

@@ -0,0 +1,41 @@
+CONNECTING
+==========
+
+Various notes (needs to be cleaned-up).
+
+See:
+ * http://blogs.msdn.com/sqlexpress/archive/2004/07/23/192044.aspx
+ * http://blogs.msdn.com/bethmassi/archive/2008/09/17/enabling-remote-sql-express-2008-network-connections-on-vista.aspx
+
+
+Example Setup
+-------------
+
+* In Visual Studio, click Server Explorer
+* Right click Data Connections
+* Create New SQL Server Database
+* Server Name: YOURPCNAME\SQLEXPRESS
+* Use Windows Authentication
+* #Use SQL Server Authentication
+* #User name: do_test
+* #Password:  do_test
+* New database name: do_test
+
+See:
+http://social.msdn.microsoft.com/Forums/en-US/Vsexpressinstall/thread/aaf2f68c-4a40-44c8-b7ee-b2f5d94e23c3
+
+
+---
+
+
+Tips
+----
+
+* Check the password is not required to be set on first connect.
+* Test you can connect locally. Either through Visual Studio SQL Server tools,
+  SQL Server Management Studio, or simply using telnet: `telnet localhost 4322`.
+* Configure Firewall correctly: http://support.microsoft.com/kb/287932
+* Specify an instance: in the DO URL append `INSTANCE=SQLEXPRESS`.
+* You can not add the instance to the hostname, i.e. `192.168.2.110\SQLEXPRESS`
+  as the underlying jTDS driver needs a URL that looks like this:
+  `jdbc:jtds:sqlserver://192.168.2.110:1433/do_test;instance=SQLEXPRESS`.

data/ChangeLog.markdown

@@ -0,0 +1,25 @@
+## 0.10.1 (unreleased, in git)
+
+Initial release as part of mainline DataObjects project.
+
+* Switch to Jeweler for Gem building tasks (this change may be temporary).
+* Switch to using Bacon for running specs: This should make specs friendlier to
+  new Ruby implementations that are not yet 100% MRI-compatible, and in turn,
+  prepared the road for our own IronRuby and MacRuby support.
+* Switch to the newly added rake-compiler `JavaExtensionTask` for compiling
+  JRuby extensions, instead of our (broken) home-grown solution.
+
+* Known Issues:
+  * Writing Extlib::ByteArray is not currently supported.
+
+## 0.10.0 2009-09-15
+
+(NOT RELEASED)
+
+* Improvements
+  * JRuby Support (using *do_jdbc*)
+
+## 0.0.1 2009-05-11
+
+* 1 major enhancement:
+  * Initial release

data/FAQS.markdown

@@ -0,0 +1,8 @@
+FAQS
+====
+
+* Can I use Microsoft's SQL Server driver instead of jTDS?
+
+No, not currently. Currently the DataObjects Driver implementation is tied not
+only to a relational database, but to its JDBC driver implementation. You could
+create an alternate database implementation.

data/INSTALL.markdown

@@ -0,0 +1,76 @@
+INSTALLING
+==========
+
+JRuby variant driver
+--------------------
+
+  * Install the jTDS JDBC Driver. For your convenience, DO packages it as a Ruby
+    Gem. If installing from source, `cd` to `../jdbc_drivers/sqlserver/` and run
+    `rake install`.
+  * There are currently no other prerequisites for installation.
+
+1.8.6/7 (MRI) and 1.9.x (YARV) variant
+--------------------------------------
+
+  1. Install the DBI and DBD::ODBC dependencies:
+
+        sudo gem install dbi
+        sudo gem install dbd-odbc
+
+  2. Install [ODBC Binding for Ruby][rubyodbc]. It it *not* currently available
+     via RubyGems, so you'll have to download the source tarball:
+
+        wget http://www.ch-werner.de/rubyodbc/ruby-odbc-0.9997.tar.gz
+        tar xvfz ruby-odbc-0.9997.tar.gz
+        cd ruby-odbc-0.9997
+
+
+  3. Ensure you read the accompanying `COPYING` file, as the ODBC Binding for Ruby
+     is licensed under the GPL, unlike DataObjects. To install read the
+     accompanying `INSTALL` file. On a recent variant of OS X, installation
+     should look like this:
+
+        ruby extconf.rb
+        make
+        sudo make install
+
+  4. You must also have FreeTDS or [unixODBC][unixodbc] (SQL Server license for
+     unixODBC is commercially licensed though) installed.
+
+  5. To install FreeTDS on OS X, you can use MacPorts:
+
+        sudo port install freetds
+
+        ****************************************************************
+        Configuration file freetds.conf does not exist and has been created using
+            /opt/local/etc/freetds/freetds.conf.sample
+        Configuration file locales.conf does not exist and has been created using
+            /opt/local/etc/freetds/locales.conf.sample
+        Configuration file pool.conf does not exist and has been created using
+            /opt/local/etc/freetds/pool.conf.sample
+        ****************************************************************
+
+
+  * Then edit your ODBC configuration and add the FreeTDS driver.
+
+    * Using a text editor: On OS X, open `/Library/ODBC/odbcinst.ini` and add
+      the following entries:
+
+            [FreeTDS]
+            Driver = /opt/local/lib/libtdsodbc.so
+            Setup  = /opt/local/lib/libtdsodbc.so
+
+    * You can also use a GUI for this (provided in Mac OS X 10.3 - 10.5;
+      [ODBCManager][odbcmanager] available for OS X 10.6)..
+      * Start ODBC Manager
+      * Go to *Drivers*, *Add...*
+      * Enter _FreeTDS_ as Driver Name.
+      * Enter `/usr/local/freetds/lib/libtdsodbc.so` as Driver File
+      * Enter `/usr/local/freetds/lib/libtdsodbc.so` as Setup File
+      * Select *System*
+      * Click *OK*.
+
+
+[rubyodbc]:http://www.ch-werner.de/rubyodbc/README
+[unixodbc]:http://www.unixodbc.org/
+[odbcmanager]:http://www.odbcmanager.net/index.php

data/LICENSE

@@ -0,0 +1,21 @@
+Copyright (c) 2009 Clifford Heath
+Copyright (c) 2009, 2010 Alex Coles
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

data/README.markdown

@@ -0,0 +1,110 @@
+# do_sqlserver
+
+* <http://dataobjects.info>
+
+## Description
+
+A Microsoft SQL Server adapter for DataObjects,
+
+## Features/Problems
+
+This driver implements the DataObjects API for the Microsoft SQL Server
+relational database.
+
+Problems with MRI implementation (unreleased):
+
+* Relies on DBI's support for either ADO or ODBC with FreeTDS
+* Has no tests and no data type conversion yet
+
+## Synopsis
+
+Examples of usage:
+
+    # default port (using SQL Server Express Edition)
+    DataObjects::Connection.new('sqlserver://user:pass@host/database;instance=SQLEXPRESS')
+    # port specified (using SQL Server Express Edition)
+    DataObjects::Connection.new('sqlserver://user:pass@host:1433/database;instance=SQLEXPRESS')
+
+    @connection = DataObjects::Connection.new("sqlserver://john:p3$$@localhost:1433/userinfo")
+    @reader = @connection.create_command('SELECT * FROM users').execute_reader
+    @reader.next!
+
+In the future, the `Connection` constructor will be able to be passed either a
+DataObjects-style URL or JDBC style URL, when using do\_sqlserver on JRuby.
+However, this feature is not currently working reliably and is a known issue.
+
+* See also the accompanying `CONNECTING.markdown`.
+
+## Requirements
+
+This driver is provided for the following platforms:
+ * JRuby 1.3.1 + (1.4+ recommended).
+
+Code for the following platform is in the repository, but is still under EARLY
+DEVELOPMENT and is neither RELEASED or SUPPORTED:
+ * Ruby MRI (1.8.6/7), 1.9: tested on Linux, Mac OS X and Windows platforms.
+
+Additionally you should have the following prerequisites:
+ * `data_objects` gem
+ * `do_jdbc` gem (shared library), if running on JRuby.
+ * `dbi` gem, if running on MRI.
+ * On non-Windows platforms, unixODBC and FreeTDS libraries.
+
+## Install
+
+To install the gem:
+
+    gem install do_sqlserver
+
+To compile and install from source:
+
+ * For MRI:
+  * Installation of do_sqlserver is significantly more involved than for other
+    drivers. Please see the accompanying `INSTALL.markdown`.
+
+ * For JRuby extensions:
+   * Install the Java Development Kit (provided if you are
+     on a recent version of Mac OS X) from <http://java.sun.com>.
+   * Install a recent version of JRuby. Ensure `jruby` is in your `PATH` and/or
+     you have configured the `JRUBY_HOME` environment variable to point to your
+     JRuby installation.
+   * Install `data_objects` and `do_jdbc` with `jruby -S rake install`.
+
+ * Then, install this driver with `(jruby -S) rake install`.
+
+Then:
+
+    sudo gem install do_sqlserver
+
+For more information, see the SQL Server driver wiki page:
+<http://wiki.github.com/datamapper/do/sql-server>.
+
+## Developers
+
+Follow the above installation instructions. Additionally, you'll need:
+  * `bacon` gem for running specs.
+  * `YARD` gem for generating documentation.
+
+See the DataObjects wiki for more comprehensive information on installing and
+contributing to the JRuby-variant of this driver:
+<http://wiki.github.com/datamapper/do/jruby>.
+
+To run specs:
+
+    rake spec
+
+To run specs without compiling extensions first:
+
+    rake spec_no_compile
+
+To run individual specs:
+
+    rake spec TEST=spec/connection_spec.rb
+
+(Note that the `rake` task uses a `TEST` parameter, not `SPEC`. This is because
+the `Rake::TestTask` is used for executing the Bacon specs).
+
+## License
+
+This code is licensed under an **MIT (X11) License**. Please see the
+accompanying `LICENSE` file.

data/Rakefile

@@ -0,0 +1,56 @@
+require 'pathname'
+require 'rubygems'
+require 'rake'
+require 'rake/clean'
+
+ROOT = Pathname(__FILE__).dirname.expand_path
+
+require ROOT + 'lib/do_sqlserver/version'
+
+JRUBY    = RUBY_PLATFORM =~ /java/
+IRONRUBY = defined?(RUBY_ENGINE) && RUBY_ENGINE == 'ironruby'
+WINDOWS  = Gem.win_platform? || (JRUBY && ENV_JAVA['os.name'] =~ /windows/i)
+SUDO     = WINDOWS ? '' : ('sudo' unless ENV['SUDOLESS'])
+
+CLEAN.include(%w[ {tmp,pkg}/ **/*.{o,so,bundle,jar,log,a,gem,dSYM,obj,pdb,exp,DS_Store,rbc,db} ext-java/target ])
+
+begin
+  gem 'jeweler', '~> 1.4'
+  require 'jeweler'
+
+  Jeweler::Tasks.new do |gem|
+    gem.name        = 'do_sqlserver'
+    gem.version     = DataObjects::SqlServer::VERSION
+    gem.summary     = 'DataObjects SQL Server Driver'
+    gem.description = 'Implements the DataObjects API for Microsoft SQL Server'
+    gem.platform    = 'java'
+    gem.files       = FileList['lib/**/*.rb', 'spec/**/*.rb', 'tasks/**/*.rake',
+                               'LICENSE', 'Rakefile', '*.{markdown,rdoc,txt,yml}',
+                               'lib/*.jar']
+    gem.extra_rdoc_files = FileList['README*', 'ChangeLog*', 'INSTALL.markdown',
+                                    'FAQS.markdown', 'LICENSE']
+    gem.test_files  = FileList['spec/**/*.rb']
+
+    gem.add_dependency 'data_objects',      DataObjects::SqlServer::VERSION
+    gem.add_dependency 'do_jdbc',           DataObjects::SqlServer::VERSION
+    gem.add_dependency 'do-jdbc_sqlserver', '~>1.2.4'
+
+    gem.add_development_dependency 'bacon',         '~>1.1'
+    gem.add_development_dependency 'rake-compiler', '~>0.7'
+
+    gem.has_rdoc    = false
+
+    gem.rubyforge_project = 'dorb'
+    gem.authors     = [ 'Alex Coles' ]
+    gem.email       = 'alex@alexcolesportfolio.com'
+  end
+
+  Rake::Task['build'].clear_actions if Rake::Task.task_defined?('build')
+  task :build => [ :java, :gem ]
+
+  Jeweler::GemcutterTasks.new
+
+  FileList['tasks/**/*.rake'].each { |task| import task }
+rescue LoadError
+  puts 'Jeweler (or a dependency) not available. Install it with: gem install jeweler'
+end

data/lib/dbd_odbc_patch.rb

@@ -0,0 +1,50 @@
+#
+# Monkey patch DBD::ODBC to pass usernames and passwords along
+# to the DB driver correctly
+#
+
+# Are you running a new version of DBI?
+begin
+  require 'dbd/ODBC'
+rescue Exception
+end
+
+# Or an old version?
+begin
+  require 'DBD/ODBC/ODBC'
+rescue Exception
+end
+
+class DBI::DBD::ODBC::Driver < DBI::BaseDriver
+
+    def connect(dbname, user, auth, attr)
+        driver_attrs = dbname.split(';')
+
+        if driver_attrs.size > 1
+            # DNS-less connection
+            drv = ::ODBC::Driver.new
+            drv.name = 'Driver1'
+            driver_attrs.each do |param|
+                pv = param.split('=')
+                next if pv.size < 2
+                drv.attrs[pv[0]] = pv[1]
+            end
+            #
+            #  These next two lines are new
+            #
+            drv.attrs['UID'] = user unless user.nil?
+            drv.attrs['PWD'] = auth unless auth.nil?
+
+            db = ::ODBC::Database.new
+            handle = db.drvconnect(drv)
+        else
+            # DNS given
+            handle = ::ODBC.connect(dbname, user, auth)
+        end
+
+        return DBI::DBD::ODBC::Database.new(handle, attr)
+    rescue DBI::DBD::ODBC::ODBCErr => err
+        raise DBI::DatabaseError.new(err.message)
+    end
+
+end

data/lib/do_sqlserver.rb

@@ -0,0 +1,296 @@
+require 'data_objects'
+if RUBY_PLATFORM =~ /java/
+  require 'do_jdbc'
+  require 'java'
+  require 'do_jdbc/sqlserver'   # the JDBC driver, packaged as a gem
+else # MRI and Ruby 1.9
+  require 'dbi' unless defined?(DBI)
+  require 'dbd_odbc_patch'      # a monkey patch for DNS-less connections
+  #require 'core_ext/dbi'       # a hack to work around ODBC millisecond handling in Timestamps
+end
+
+require 'bigdecimal'
+require 'date'
+require 'base64'
+require 'do_sqlserver/do_sqlserver' if RUBY_PLATFORM =~ /java/
+require 'do_sqlserver/version'
+# JDBC driver has transactions implementation in Java
+require 'do_sqlserver/transaction' if RUBY_PLATFORM !~ /java/
+
+if RUBY_PLATFORM !~ /java/
+  module DataObjects
+    module SqlServer
+      Mode = :odbc
+#      The ADO mode requires a very old DBI to work with the unmaintained DBI::ADO adapter. Don't enable this unless you fix that.
+#      Mode = begin
+#          require "ADO"
+#          :ado
+#        rescue LoadError => e
+#          :odbc
+#        end
+
+      class Connection < DataObjects::Connection
+        def initialize uri
+          # REVISIT: Allow uri.query to modify this connection's mode?
+          #host = uri.host.blank? ? "localhost" : uri.host
+          host = uri.host.blank? ? nil : uri.host
+          user = uri.user || "sa"
+          password = uri.password || ""
+          path = uri.path.sub(%r{^/*}, '')
+          port = uri.port || "1433"
+          if Mode == :ado
+            connection_string = "DBI:ADO:Provider=SQLOLEDB;Data Source=#{host};Initial Catalog=#{path};User ID=#{user};Password=#{password};"
+          else
+            # FIXME: Cannot get a DNS-less configuration without freetds.conf to
+            # connect successfully, i.e.:
+            # connection_string = "DBI:ODBC:DRIVER=FreeTDS;SERVER=#{host};DATABASE=#{path};TDS_Version=5.0;Port=#{port}"
+            #
+            # Currently need to setup a dataserver entry in freetds.conf (if
+            # using MacPorts, full path is /opt/local/etc/freetds/freetds.conf):
+            #
+            # [sqlserver]
+            #   host = hostname
+            #   port = 1433
+            #   instance = SQLEXPRESS
+            #   tds version = 8.0
+            #
+            connection_string = "DBI:ODBC:DRIVER=FreeTDS;SERVERNAME=sqlserver;DATABASE=#{path};"
+          end
+
+          begin
+            @connection = DBI.connect(connection_string, user, password)
+          rescue DBI::DatabaseError => e
+            # Place to debug connection failures
+            raise
+          end
+
+          @encoding = uri.query && uri.query["encoding"] || "utf8"
+
+          set_date_format = create_command("SET DATEFORMAT YMD").execute_non_query
+          options_reader = create_command("DBCC USEROPTIONS").execute_reader
+          while options_reader.next!
+            key, value = *options_reader.values
+            value = options_reader.values
+            case key
+            when "textsize"                     # "64512"
+            when "language"                     # "us_english", "select * from master..syslanguages" for info
+            when "dateformat"                   # "ymd"
+            when "datefirst"                    # "7" = Sunday, first day of the week, change with "SET DATEFIRST"
+            when "quoted_identifier"            # "SET"
+            when "ansi_null_dflt_on"            # "SET"
+            when "ansi_defaults"                # "SET"
+            when "ansi_warnings"                # "SET"
+            when "ansi_padding"                 # "SET"
+            when "ansi_nulls"                   # "SET"
+            when "concat_null_yields_null"      # "SET"
+            else
+            end
+          end
+        end
+
+        def using_socket?
+          # This might be an unnecessary feature dragged from the mysql driver
+          raise "Not yet implemented"
+        end
+
+        def character_set
+          @encoding
+        end
+
+        def dispose
+          @connection.disconnect
+          true
+        rescue
+          false
+        end
+
+        def raw
+          @connection
+        end
+      end
+
+      class Command < DataObjects::Command
+        # Theoretically, SCOPE_IDENTIY should be preferred, but there are cases where it returns a stale ID, and I don't know why.
+        #IDENTITY_ROWCOUNT_QUERY = 'SELECT SCOPE_IDENTITY(), @@ROWCOUNT'
+        IDENTITY_ROWCOUNT_QUERY = 'SELECT @@IDENTITY, @@ROWCOUNT'
+
+        attr_reader :types
+
+        def set_types *t
+          @types = t.flatten
+        end
+
+        def execute_non_query *args
+          DataObjects::SqlServer.check_params @text, args
+          begin
+            handle = @connection.raw.execute(@text, *args)
+          rescue DBI::DatabaseError => e
+            handle = @connection.raw.handle
+            handle.finish if handle && handle.respond_to?(:finish) && !handle.finished?
+            DataObjects::SqlServer.raise_db_error(e, @text, args)
+          end
+          handle.finish if handle && handle.respond_to?(:finish) && !handle.finished?
+
+          # Get the inserted ID and the count of affected rows:
+          inserted_id, row_count = nil, nil
+          if (handle = @connection.raw.execute(IDENTITY_ROWCOUNT_QUERY))
+            row1 = Array(Array(handle)[0])
+            inserted_id, row_count = row1[0].to_i, row1[1].to_i
+            handle.finish
+          end
+          Result.new(self, row_count, inserted_id)
+        end
+
+        def execute_reader *args
+          DataObjects::SqlServer.check_params @text, args
+          massage_limit_and_offset args
+          begin
+            handle = @connection.raw.execute(@text, *args)
+          rescue DBI::DatabaseError => e
+            handle = @connection.raw.handle
+            DataObjects::SqlServer.raise_db_error(e, @text, args)
+            handle.finish if handle && handle.respond_to?(:finish) && !handle.finished?
+          rescue
+            handle = @connection.raw.handle
+            handle.finish if handle && handle.respond_to?(:finish) && !handle.finished?
+            raise
+          end
+          Reader.new(self, handle)
+        end
+
+      private
+        def massage_limit_and_offset args
+          @text.sub!(%r{SELECT (.*) ORDER BY (.*) LIMIT ([?0-9]*)( OFFSET ([?0-9]*))?}) {
+            what, order, limit, offset = $1, $2, $3, $5
+
+            # LIMIT and OFFSET will probably be set by args. We need exact values, so must
+            # do substitution here, and remove those args from the array. This is made easier
+            # because LIMIT and OFFSET are always the last args in the array.
+            offset = args.pop if offset == '?'
+            limit = args.pop if limit == '?'
+            offset = offset.to_i
+            limit = limit.to_i
+
+            # Reverse the sort direction of each field in the ORDER BY:
+            rev_order = order.split(/, */).map{ |f|
+              f =~ /(.*) DESC *$/ ? $1 : f+" DESC"
+            }*", "
+
+            "SELECT TOP #{limit} * FROM (SELECT TOP #{offset+limit} #{what} ORDER BY #{rev_order}) ORDER BY #{order}"
+          }
+        end
+      end
+
+      class Result < DataObjects::Result
+      end
+
+      # REVISIT: There is no data type conversion happening here. That will make DataObjects sad.
+      class Reader < DataObjects::Reader
+        def initialize command, handle
+          @command, @handle = command, handle
+          return unless @handle
+
+          @fields = handle.column_names
+
+          # REVISIT: Prefetch results like AR's adapter does. ADO is a bit strange about handle lifetimes, don't move this until you can test it.
+          @rows = []
+          types = @command.types
+          if types && types.size != @fields.size
+            @handle.finish if @handle && @handle.respond_to?(:finish) && !@handle.finished?
+            raise ArgumentError, "Field-count mismatch. Expected #{types.size} fields, but the query yielded #{@fields.size}"
+          end
+          @handle.each do |row|
+            field = -1
+            @rows << row.map do |value|
+              field += 1
+              next value unless types
+              if (t = types[field]) == Integer
+                Integer(value)
+              elsif t == Float
+                Float(value)
+              else
+                t.new(value)
+              end
+            end
+          end
+          @handle.finish if @handle && @handle.respond_to?(:finish) && !@handle.finished?
+          @current_row = -1
+        end
+
+        def close
+          if @handle
+            @handle.finish if  @handle.respond_to?(:finish) && !@handle.finished?
+            @handle = nil
+            true
+          else
+            false
+          end
+        end
+
+        def next!
+          (@current_row += 1) < @rows.size
+        end
+
+        def values
+          raise StandardError.new("First row has not been fetched") if @current_row < 0
+          raise StandardError.new("Last row has been processed") if @current_row >= @rows.size
+          @rows[@current_row]
+        end
+
+        def fields
+          @fields
+        end
+
+        def field_count
+          @fields.size
+        end
+
+        # REVISIT: This is being deprecated
+        def row_count
+          @rows.size
+        end
+      end
+
+    private
+      def self.check_params cmd, args
+        actual = args.size
+        expected = param_count(cmd)
+        raise ArgumentError.new("Binding mismatch: #{actual} for #{expected}") if actual != expected
+      end
+
+      def self.raise_db_error(e, cmd, args)
+        msg = e.to_str
+        case msg
+        when /Too much parameters/, /No data found/
+          #puts "'#{cmd}' (#{args.map{|a| a.inspect}*", "}): #{e.to_str}"
+          check_params(cmd, args)
+        else
+          e.errstr << " running '#{cmd}'"
+          #puts "'#{cmd}' (#{args.map{|a| a.inspect}*", "}): #{e.to_str}"
+          #debugger
+        end
+        raise
+      end
+
+      def self.param_count cmd
+        cmd.gsub(/'[^']*'/,'').scan(/\?/).size
+      end
+
+    end
+
+  end
+
+else
+
+  # Register SqlServer JDBC driver
+  java.sql.DriverManager.registerDriver Java::net.sourceforge.jtds.jdbc.Driver.new
+
+  DataObjects::SqlServer::Connection.class_eval do
+
+    def quote_boolean(value)
+      value ? 1 : 0
+    end
+
+  end
+
+end