Forums

Hi Mark,

It's pretty straightforward to access the response object from within the .load callback. You'd do something like;

The details of the DataTable.DataSource.load() command are at http://yuilibrary.com/yui/docs/api/files/datatable_js_datatable-datasource.js.html#l126, a little hard to find! Also, the "onDataReturnInitializeTable" is immediately below the load command.

You are free to add / delete whatever rows you like in the success callback function, but you need to handle the myDataTable.set('data',your_data) yourself.

As for "combining columns" as you said in your posting ... that could be a little more involved, because the DataTable is constructed assuming the columns you defined in the "columns" config, so hopefully you aren't planning to change that.

Todd

Mark,

Clearly your modifications to the DS response results won't be available in the cache, although the original pristine response results should be accessible.

You're correct that you don't necessarily need to use the DS, you can simply do your Y.io call and any parsing and/or other mods and feed them right into DT with set('data'). A real advantage is with DataSource you can hook-up any of several DataSchemas (JSON, XML, etc...) and can also handle conversions with DataSchema parsers on individual fields.

On a related note, amongst the many changes in the recent 3.5 version of DataTable, you can now define a Y.ModelList that has it's own sync layer built-in, essentially freeing you from using DS or Y.io on your own. I recently posted a forum item regarding some work using a RESTful sync layer at http://yuilibrary.com/forum/viewtopic.php?t=10217. This ModelList sync approach is pretty cool IMHO, so that is potentially another alternative for you.

Regards,

Todd

I don't really use the caching frequently. Looks like you can push your changed response back on the DS cache using the DataSourceCache.add method. You'd have to experiment with that if that's a requirement.