struts-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nick Afshartous <nick.afshart...@proactcorp.com>
Subject Re: Bean taglib TODO Tasks
Date Wed, 07 Mar 2001 16:11:55 GMT

Could someone please comment on my questions below, thanks.
-- 

	Nick


From: Nick Afshartous <nick.afshartous@proactcorp.com>
To: struts-dev@jakarta.apache.org
Subject: Re: Bean taglib TODO Tasks
Date: Thu, 1 Mar 2001 11:36:21 -0500 (EST)

 > Dave Bettin wrote:
 > > Before I start formulating my design ideas, what are
 > > some of the design goals(you had in mine) regarding
 > > the XPath, DOM, and JDBC issues?

Craig McClanahan writes:
 > I have not thought out design plans in detail, but the following goals
 > would be high on my list:
 > 
 > * For the existing custom bean tags, such as <bean:define>
 >   or <bean:write>, the page designer should not know or care
 >   whether the underlying data is JavaBeans (the currently supported
 >   property lookup mechanism), a DOM object (where attributes and
 >   nested elements look like "properties"), or a JDBC ResultSet
 >   or RowSet (which appear to be a collection of beans, where each
 >   row is a bean and each column is a property).

I'm interested in how JDBC/GenericDataSource will be integrated into
the tablibs.  Here are some questions just to stimulate some
discussion.

  * How will ResultSet values be created from within a tag ?  
    Maybe something like:

       <bean:define id="result" datasource="myds" 
                    sql="some sql query">


  * Would other tags (i.e. iterate) then be extended to 
    work with ResultSets ?  

-- 

	Nick


 > * For existing tags that take "name", "property", and "scope"
 >   arguments, define an "xpath" attribute that uses an XPath
 >   expression to access properties.  This would be an alternative
 >   to the existing name/property/scope access, extended as
 >   defined in the previous point.
 > 
 > The basic idea is we want to make the presentation pages as independent
 > as possible of the actual architecture of the data that is provided by
 > the business logic.  In addition, transparent support for DOM and JDBC
 > objects will enable simple applications that require little or no
 > processing to be created.

Mime
View raw message