Nice. I definitely have a use for this.
On Feb 1, 11:06 am, Luca Molino <
molino.l...@gmail.com> wrote:
> Hi all,
>
> I'm happy to announce new OrientDB Fetch Plan strategies and improvement
> (trunk revision 4742):
>
> *Commons features*:
>
> 1. Centralized fetch algorithm
> 2. fetch depth is record centric: now every field that is not a record
> or a collection of record will be fetched.
> 3. fetch level 0 (zero) is no more "don't fetch" but means fetch current
> record
> 4. fetch level >0 means fetch x level of depth (Example: *:0--> fetch
> only root record, *:2 fetch all records two depht levels down)
> 5. fields fetch now works in dot notation
> 6. fetch algorithm now uses an improved OFetchListener and a new
> OFetchContext
> 7. Class fetch (Classname.classfield:x) it's not available anymore (will
> be reimplemented again soon)
>
> OFetchListener interface takes care of what to do in case of field
> fetching: standard, record to fetch, record not to fetch, map or collection
> entries.
>
> OFetchContext interface takes care of anything the user wants do do before
> and after fetching an entire record or a single field divided by its type
> (Standard, Document, Collection, Map,Array).
>
> ObjectDatabase users need to know that now if fetch reached its max level
> the field inside the user object will be set as *null*.
>
> *Examples*: