What makes you think it replaces the entire dataset? Have you seen this
happen? HTTP PUT on the dataset URI wil do that - s-put does not
because it replaces the graph, default or named, not the whole dataset.
See the "SPARQL Graph Store Protocol".
If your file is not in its own graph, you'll need to use a SPARQL Update
s-post is "add RDF to a graph". in HTTP, POST should always be
"append-like" when on some container-like resource.
On 16/09/16 10:02, Sandor Kopacsi wrote:
Dear List Members,
In a previous thread (extending an existing dataset) Andy suggested to
use s-put to replace a deficient data file (delete the old one and
upload a new one) in an existing dataset.
I am not sure, but I am afraid that the suggested s-put scrip is not
appropriate to upload new data file in my case, since I have already
several other data files in the dataset in the TDB that I want to keep
intact. I only would like to replace a certain data file with a new one
(extended with some additional triplets) under the same graph name. As
far as I know s-put will clean the entire dataset first before it
uploads the new datafile. But I might use s-post, that - according to
may knowledge - can add new datafile without cleaning the existing data
files from the TDB. Am I right?
Another the question is, how can I delete a dedicated graph from the
dataset. The utility s-delete might help?
Thank you very much and best regards,