I still feel more comfortable doing the temp table version first, but
you should ask him for the patch so that we can base version 2 on it.
On Mar 13, 2009, at 9:54 AM, Igor Babaev wrote:
> Eric Jensen wrote:
>> On Mar 13, 2009, at 4:46 AM, Sergei Golubchik wrote:
>>>> 4. make st_select_lex_unit::exec use k-way merge logic relying on
>>>> one-result-at-a-time fetching to retrive up to LIMIT n results
>>>> treating our temp table with its ordered index as a heap
>>> Wouldn't it be better to use a heap as a heap, not btree as a heap ?
>>> Heap is faster than btree, and MySQL has a heap implementation in
>> would be better indeed! however, i could not find another example
>> of a select result class that reads results from a heap. from what
>> i remember, much of the logic for reading from an intermediate
>> result set is specific to reading from a temporary table. for
>> example, i don't know where to find all the proper comparators to
>> execute an arbitrary order by in the heap. the heap seems like
>> version two to me, unless i'm mistaken about the difficulty of
>> storing intermediate row results in it, performing comparisons on
>> them, and feeding them back to the user there a select result class
>> that behaves like the ones that read from temporary tables.
> Somewhere in 2005 ago Evgeny Potemkin (eugene@stripped) implemented
> ORDER BY <list> LIMIT N employing priority queue (heap) from mysys/
> I'm pretty sure he still has that patch in his archives.
> Would you like to contact Evgeny or you'd prefer me to ask him for
> that patch?
> The patch most probably will help you.
> MySQL Internals Mailing List
> For list archives: http://lists.mysql.com/internals
> To unsubscribe: http://lists.mysql.com/internals?