[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: subtree filter spec vs. implementation



Martin Bjorklund wrote:
Andy Bierman <ietf@andybierman.com> wrote:
This is bad news.  Can we issue an errata for a spec?
I hope so.
We need to agree on the text.
I searched some old drafts and could not find my
original sentence regarding removal of dead-end sub-trees.

FWIW, I think it is about the same amount of work
to implement subtree filtering either way.

This is very subjective of course, but I think the current spec is
easier to implement.  The reason is that when a containment node is
found in the filter, the code can directly write the node to the
output stream if it exists.  With the proposed change, the code would
have to buffer more.  With the current spec, max one level of
buffering is needed (for content match nodes).  No rocket science of
course.

Implementation details will vary a lot here,
so it's hard to compare apples to apples.

It could be the same amount of data saved,
but you may have to go through the filter tree
twice -- once to prune it, and then again
to output the matching nodes (the ones still there
after the pruning).



/martin

Andy

--
to unsubscribe send a message to netconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/netconf/>