I tried manually just add the line @schema_name = nil to schema.rb but it did not change the behaviour (maybe in between there were other changes, I see the directory structure is changed slightly inside of lib/)

I will wait for the gem. And anyway it’s not critical since the parallel connections option works great.

thanks for your great work

PS. from next issue on will post to the discussion group :)

I pushed the patch to github – if this is critical for you then please clone it from github.com/rsim/ruby-plsql and run “rake install_gem”. Or you can wait until I publish new version of gem.

Btw if you need to operate with two different connections then you can use them in parallel like:

plsql(:one).connection = OCI8.new(’chris’,’*****’,’emeadb11′)
plsql(:two).connection = OCI8.new(’kcierpisz’,’*****’,’mktgcd’)
plsql(:one).emeadb_phone_pkg.prepare_phones_in(’EMEA_PHONES’)
plsql(:two).phone_pkg.create_table(’EMEA_PHONES_TMP’)

And also if you have any further issues with ruby-plsql then please use discussion group http://groups.google.com/group/oracle-enhanced – it is primarily used for oracle_enhanced adapter but ruby-plsql questions might be interesting for oracle_enhanced users as well.

I am very happy user of your gem.

I encountered today an obstacle when handling 2 connections in one script.
I have 2 databases emeadb11 and mktgcd

chris@chris-ub:~$ irb
irb(main):001:0> require ‘rubygems’
=> true
irb(main):002:0> require ‘ruby_plsql’
=> true
irb(main):003:0> con1 = OCI8.new(‘chris’,'*****’,'emeadb11′)
=> #
irb(main):004:0> con2 = OCI8.new(‘kcierpisz’,'*****’,'mktgcd’)
=> #
irb(main):005:0> plsql.connection = con1
=> #
irb(main):006:0> plsql
=> #<PLSQL::Schema:0x8cd9f94 @schema_name=nil, @procedures={}, @connection=#<PLSQL::OCIConnection:0x8cd95a8 @activerecord_class=nil, @raw_connection=#, @raw_driver=:oci>, @first=true, @schemas={}, @packages={}>
irb(main):007:0> plsql.emeadb_phone_pkg.prepare_phones_in(‘EMEA_PHONES’)
=> nil
irb(main):008:0> plsql
=> #<PLSQL::Schema:0x8cd9f94 @schema_name="CHRIS", @procedures={}, @connection=#<PLSQL::OCIConnection:0x8cd95a8 @activerecord_class=nil, @raw_connection=#, @raw_driver=:oci>, @first=true, @schemas={}, @packages={:emeadb_phone_pkg=>##[]}, @schema=#, @arguments={0=>{:emea_phones_table=>{:data_precision=>nil, :data_type=>”VARCHAR2″, :data_scale=>nil, :in_out=>”IN”, :data_length=>nil, :position=>1}}}, @argument_list={0=>[:emea_phones_table]}, @overloaded=false, @package=”EMEADB_PHONE_PKG”, @return={0=>nil}, @overloads=[0], @procedure=”PREPARE_PHONES_IN”>}, @override_schema_name=nil, @schema=#, @package=”EMEADB_PHONE_PKG”>}>
irb(main):009:0> plsql.connection = con2
=> #
irb(main):010:0> plsql
=> #<PLSQL::Schema:0x8cd9f94 @schema_name="CHRIS", @procedures={}, @connection=#<PLSQL::OCIConnection:0x97554c8 @activerecord_class=nil, @raw_connection=#, @raw_driver=:oci>, @first=true, @schemas={}, @packages={}>
irb(main):011:0> plsql.phone_pkg.create_table(‘EMEA_PHONES_TMP’)
ArgumentError: No PL/SQL procedure found
from /home/chris/.gem/ruby/1.8/gems/ruby-plsql-0.3.1/lib/plsql/schema.rb:127:in `method_missing’
from (irb):11
from :0
irb(main):012:0>

as you can see at irb(main):006 schema_name is nil (although it should already here have CHRIS I guess)
at irb(main):008 schema_name has value ‘CHRIS’ (probably because at irb(main):007 I ran a procedure)
at irb(main):009 I am assigning con2 (KCIERPISZ schema)
plsql though sees still CHRIS under schema_name
and that’s why irb(main):011 fails.

When I run only one connection per script everything is fine:
irb(main):001:0> require ‘rubygems’
=> true
irb(main):002:0> require ‘ruby_plsql’
=> true
irb(main):003:0>
irb(main):004:0* con2 = OCI8.new(“kcierpisz”,”*****”,”mktgcd”)
=> #
irb(main):005:0> plsql.connection = con2
=> #
irb(main):006:0> plsql.phone_pkg.create_table(‘EMEA_PHONES_TMP’)
=> nil
irb(main):007:0> plsql
=> #<PLSQL::Schema:0x8f33984 @schema_name="KCIERPISZ", @procedures={}, @connection=#<PLSQL::OCIConnection:0x8f338e4 @activerecord_class=nil, @raw_connection=#, @raw_driver=:oci>, @first=true, @schemas={}, @packages={:phone_pkg=>##[]}, @schema=#, @arguments={0=>{:emea_phones_tmp_table=>{:data_precision=>nil, :data_type=>”VARCHAR2″, :data_scale=>nil, :in_out=>”IN”, :data_length=>nil, :position=>1}}}, @argument_list={0=>[:emea_phones_tmp_table]}, @overloaded=false, @package=”PHONE_PKG”, @return={0=>nil}, @overloads=[0], @procedure=”CREATE_TABLE”>}, @override_schema_name=nil, @schema=#, @package=”PHONE_PKG”>}>
irb(main):008:0>

Is there a way I could overwrite that schema_name (if that causes the issue)?

thanks a lot for your help,
chris

I guess an enhanced version could support batch fetching similar to find_in_batches.
I have managed to call using a ref-cursor using the following jdbc code – not a generic solution and only for JRuby but found no references on google so i will paste it here….

def test
conn = ActiveRecord::Base.connection.raw_connection
cstmt = conn.prepareCall(“begin test(?); end;”)
cstmt.registerOutParameter(1, -10) # -10 == oracle.jdbc.driver.OracleTypes.CURSOR
cstmt.execute
rs = cstmt.getObject(1)
rsm = rs.getMetaData
columnCount = rsm.getColumnCount
rows = []
while (rs.next())
row = []
for j in (1..columnCount)
row << rs.getObject(j)
end
rows << row
end
return rows
ensure
rs.close if rs
cstmt.close if cstmt
end

How would you like to use returned cursor in Ruby? Would you like to fetch all rows from it or fetch one by one?

Note: the stored proc is like…

procedure pkg.test ( p_cursor out sys_refcursor)
is
begin
open p_cursor for
select
col1 ,
col2
from pkg.table1;
end test;