Hi Bernhard,
Bernhard Woditschka wrote:
> Hi, here's the status of the Castor Pull Mapping gsoc project:
>
> Currently my gsoc mentor Joachim Grüneis and I are discussing the main use
> cases and the design of the pull mapping.
>
> Main use case:
>
> Processing of very large documents that are flat in their structuree.g. a
> table export from a database.
> In this case the whole document shouldn't be mapped at once because it would
> require a far to large memory footprint.
>
> Proposed design:
>
> 1) Create a PullMapper trough the XMLContext
> 2) initialize it with the xml source and a xpath to the xml fragments of
> interest.
What are your plans on providing xpath support ? Do you think we can
(should) provide full xpath support ? Would this technically be feasible
at all ?
> 3) pull the mapped fragments one by one from the PullMapper
What does this exactly mean ? Especially in the context of the existing
functionality of Castor XML ? Wouldn't it make more sense to use a pull
parser internally to move to a location as identified by the xpath
expression, and then stream the resulting XML into Castor XML (through a
DocumentHandler) ?
>
> no worries, Bernhard
>
Werner
---------------------------------------------------------------------
To unsubscribe from this list, please visit:
http://xircles.codehaus.org/manage_email
