On Wed, Apr 04, 2007 at 08:57:46PM +0100, Olly Betts wrote:
> You said:
>
> | Have we tested iterating through the
get_matching_terms() output
> | directly, as opposed to using the MSetIterator to do
it internally?
>
> OK, my best guess is that by
"get_matching_terms()" you mean
> "get_matching_terms_begin() and
get_matching_terms_end()" (and by
> "MSetIterator" you mean
"TermIterator"). The answer to that amended
> question is that this reportedly works fine.
>
> If that's not the question you were asking, what was?
No, the opposite. We know that get_matching_terms_begin()
and
get_matching_terms_end() work directly. However
get_matching_terms()
has extra code over them to make it all magic. (Yeah,
TermIterator.)
You were suggesting that the problem might be in running the
output of
get_matching_terms() through join(); I thought I saw
something
suggesting that someone had tested the output of
get_matching_terms()
directly, and while _begin() and _end() worked, that
didn't.
> I can't see an answer either way on this in the earlier
mails - if you
> saw it, please point it out.
I meant:
<http://permalink.gmane.org/gmane.comp.search.x
apian.devel/1048> (I
think).
> The other piece of information we definitely do have is
that this works:
>
> join(foo())
>
> where foo() returns the same array which the call to
> get_matching_terms() is supposed to.
Yeah, which is why I think it's inside
get_matching_terms().
J
--
/-----------------------------------------------------------
---------------
James Aylett
xapian.org
james tartarus.org
uncertaintydivision.org
_______________________________________________
Xapian-devel mailing list
Xapian-devel lists.xapian.org
http://lists.xapian.org/mailman/listinfo/xapian-devel
|