Hi!
I would break out the StorageEngine class to its own file. I've still got some
major changes that will be coming through that before it is done. The next big
change will be TableIdentifiers and those will touch all of the do*() methods.
Also, we are nearing the end of where an engine will need store_lock(), and I
suspect that for some engines this will mean they will no longer need a share.
Cheers,
-Brian
On Nov 18, 2009, at 9:21 AM, Jay Pipes wrote:
> Paul McCullagh wrote:
>> Thanks, Jay, we missed that one.
>> Just push your changes, and they will find their way into the PBXT trunk.
>> I think we will have to put an #ifdef DRIZZLED around it because the
>> create_info->options version is required by MySQL.
>
> I see, yes. Or...it may be possible to have an ha_pbxt_mysql.cc (or .c) file
> and a pbxt_storage_engine.cc file for Drizzle only. The reason I suggest
> this is that there are some massive changes to the StorageEngine API for
> Drizzle. Using references instead of pointers is just the tip of the
> iceberg. But, since PBXT's code is quite clean and modularized, making a
> Drizzle wrapper and a MySQL wrapper really shouldn't be too difficult, and
> creating the Drizzle wrapper is something we are eager to help do! :)
>
> Cheers!
> Jay
>
>> But we can do that when we merge into the PBXT trunk.
>> On Nov 18, 2009, at 5:51 PM, Jay Pipes wrote:
>>> Vladimir Kolesnikov wrote:
>>>> Hi all!
>>>> I have just pushed an important fix for a bug that sometimes caused index
>>>> scans to hang. Rev.1052.
>>>
>>> Hi!
>>>
>>> Just pulled latest and am getting the following compile error:
>>>
>>> btw, I am using: gcc 4.3.2 and configure'ing --with-debug
>>>
>>> It is the --with-debug that will show the failures....
>>>
>>> plugin/pbxt/src/ha_pbxt.cc: In member function ‘virtual int
>>> PBXTStorageEngine::doCreateTable(Session*, const char*, Table&,
>>> HA_CREATE_INFO&, drizzled::message::Table&)’:
>>> plugin/pbxt/src/ha_pbxt.cc:5412: error: base operand of ‘->’ has
>>> non-pointer type ‘st_ha_create_information’
>>>
>>> The fix is simple. Line 5412 must change from:
>>>
>>> XT_PRINT2(self, "create (%s) %s\n", table_path, (create_info->options &
>>> HA_LEX_CREATE_TMP_TABLE) ? "temporary" : "");
>>>
>>> to:
>>>
>>> XT_PRINT2(self, "create (%s) %s\n", table_path, (create_info.options &
>>> HA_LEX_CREATE_TMP_TABLE) ? "temporary" : "");
>>>
>>> After that, compiles just fine. :)
>>>
>>> Cheers!
>>>
>>> Jay
>>>
>>> _______________________________________________
>>> Mailing list: https://launchpad.net/~drizzle-discuss
>>> Post to : [email protected]
>>> Unsubscribe : https://launchpad.net/~drizzle-discuss
>>> More help : https://help.launchpad.net/ListHelp
>> --
>> Paul McCullagh
>> PrimeBase Technologies
>> www.primebase.org
>> www.blobstreaming.org
>> pbxt.blogspot.com
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~drizzle-discuss
> Post to : [email protected]
> Unsubscribe : https://launchpad.net/~drizzle-discuss
> More help : https://help.launchpad.net/ListHelp
_______________________________________________
Mailing list: https://launchpad.net/~drizzle-discuss
Post to : [email protected]
Unsubscribe : https://launchpad.net/~drizzle-discuss
More help : https://help.launchpad.net/ListHelp
