[ thinks a bit ... ] At least for GIST, it is possible that whether
data can be regurgitated will vary depending on the selected opclass.
Some opclasses use the STORAGE modifier and some don't. I am not sure
how hard we want to work to support flexibility there. Would it be
sufficient to
On Mon, Jul 13, 2009 at 11:32 AM, Heikki
Linnakangasheikki.linnakan...@enterprisedb.com wrote:
Tom Lane wrote:
One thought here is that an AM call isn't really free, and doing two of
them instead of one mightn't be such a good idea. I would suggest
either having a separate AM entry point to
At the moment, amgettuple only returns pointers to heap tuples. There is
no way to return data from the index tuples. That needs to be changed to
support index-only scans.
I propose that we split index_getnext into two steps: fetching the next
match from the index (index_next()), and fetching the
On Mon, Jul 13, 2009 at 8:19 AM, Heikki
Linnakangasheikki.linnakan...@enterprisedb.com wrote:
I propose that we split index_getnext into two steps: fetching the next
match from the index (index_next()), and fetching the corresponding heap
tuple (index_fetch()).
A pretty trivial concern, but
Heikki Linnakangas heikki.linnakan...@enterprisedb.com writes:
At the moment, amgettuple only returns pointers to heap tuples. There is
no way to return data from the index tuples. That needs to be changed to
support index-only scans.
What are you going to do for index types that don't store
Tom Lane wrote:
Heikki Linnakangas heikki.linnakan...@enterprisedb.com writes:
At the moment, amgettuple only returns pointers to heap tuples. There is
no way to return data from the index tuples. That needs to be changed to
support index-only scans.
What are you going to do for index types
Heikki Linnakangas heikki.linnakan...@enterprisedb.com writes:
Tom Lane wrote:
What are you going to do for index types that don't store the original
data (e.g. hash)?
They will obviously not be able to regurgitate index tuples. I have not
yet decided how that's going to be signaled.
Well,
Tom Lane wrote:
One thought here is that an AM call isn't really free, and doing two of
them instead of one mightn't be such a good idea. I would suggest
either having a separate AM entry point to get both bits of data
(amgettupledata?) or adding an optional parameter to amgettuple.
I'm
