Mailing-List: contact dev-help@cocoon.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cocoon.apache.org
In-Reply-To: <397A40AC-8EF8-11D8-844A-000A95AF004E@apache.org>
References: <E5547B42-8EF4-11D8-844A-000A95AF004E@apache.org>
 <2DEA31D6-8EF6-11D8-B513-000A959D0312@mac.com>
 <397A40AC-8EF8-11D8-844A-000A95AF004E@apache.org>
Mime-Version: 1.0 (Apple Message framework v613)
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Message-Id: <DBCEBBC0-8EF8-11D8-B513-000A959D0312@mac.com>
Content-Transfer-Encoding: quoted-printable
Cc: dev@cocoon.apache.org
From: jastrachan@mac.com
Subject: Re: GroovyMarkup syntax works!
Date: Thu, 15 Apr 2004 17:20:49 +0100
To: Bertrand Delacretaz <bdelacretaz@apache.org>

On 15 Apr 2004, at 17:16, Bertrand Delacretaz wrote:
> Le 15 avr. 04, =E0 18:01, jastrachan@mac.com a =E9crit :
>
>> ...BTW you could just create the Sql object with a DataSource which=20=

>> is-a JDBC connection pool, then the Sql object takes care of all the=20=

>> pooling for you. Is there support for DataSource in Cocoon or some=20
>> kinda adapter between DataSource and ConnectionProvider?...
>
> Dunno about javax.sql.DataSource support, I'll have to check.
>
> ConnectionProvider does not exist yet, I just made it up with the=20
> following idea:

Ah OK :)

> public class ConnectionProvider {
>   // Groovy scripts which need a Connection will use this to get
>   // it from the Cocoon pool
>   public Connection getConnection(String connectionName);
>
>   // then, once the Groovy script has been executed, ScriptGenerator
>   // calls this to return all connections provided by getConnection to=20=

> the pool
>   public void releaseConnections()
> }
>
> In this way, the Groovy script can access any Connections that are=20
> configured, and does not have to care about releasing them. It's easy=20=

> to implement in case we don't have DataSource support.

You might as well just use DataSource then as this is pretty close to=20
all DataSource is. (Or at the least get ConnectionProvider to implement=20=

DataSource)

There's an implementation in Jakarta Commons dbcp and various JDBC=20
drivers come with an implementation of DataSource for you already (e.g.=20=

Oracle & Axion).

If nothing else just provide a DataSource facade to whatever connection=20=

pool you use and the groovy.sql.Sql will work fine with your pool.

James
-------
http://radio.weblogs.com/0112098/