On 30/10/13 11:10, Jean-Marc Vanel wrote:
Thanks for the answer , which is not surprising .
This suggests new questions:
- 1. would it be difficult to add this feature, and make timestamp
available through the API ?
Contributions are always welcome (but this looks like a distinctly
non-trivial change to Jena :-)
Have you considered using named graphs + unionDefaultGraph, using the
graph name to indicate time?
- 2. is the order of triples insertion also preserved in SPARQL results,
API SPO queries and other outputs ?
No. Nothing is gauranteed.
2013/10/30 Rob Vesse <[email protected]>
No it does
However since triples are stored in B+Tree based indexes there is an
ordering to the triples within the internal storage so when tdbdump
outputs the contents of the database the triples will be serialized out in
that order.
And, caveat, that does not form part of the contract.
Rob
On 30/10/2013 10:20, "Jean-Marc Vanel" <[email protected]> wrote:
Hi
Does Jena TDB record timestamp of triples?
( I noticed that tdbdump outputs triples in chronological order ).
If you add data for subjects that already exist in the DB, those triples
will appear in the dump at that subject, not at the end.
A backup is dumping by reading the SPO index and writing out the triples.
Each unique S gets an allocated number - it happens to increase with
time in the current implementation.
Andy
--
Jean-Marc Vanel
Déductions SARL - Consulting, services, training,
Rule-based programming, Semantic Web
http://deductions-software.com/
+33 (0)6 89 16 29 52
Twitter: @jmvanel ; chat: irc://irc.freenode.net#eulergui
