Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 12:33:09 +0200] rev 5105
[test] use new temporary_appobjects context manager
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 12:32:39 +0200] rev 5104
[test] that doesn't smell good but fix the test...
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 12:22:20 +0200] rev 5103
[vreg] we do want the opposite...
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 12:26:07 +0200] rev 5102
[testlib] simple helper method to register temporary appobjects (hooks for instance) with a context manager
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 12:21:39 +0200] rev 5101
[undo] fix transactions table name in cleanup operation
Pierre-Yves David <pierre-yves.david@logilab.fr> [Wed, 31 Mar 2010 11:52:15 +0200] rev 5100
Add context management methode to dbapi.Connection
the connection is automaticly closed when exiting the context.
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 11:41:20 +0200] rev 5099
i18n update
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 11:41:17 +0200] rev 5098
[i18n] use named substitution to avoid gettext warnings
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 10:56:43 +0200] rev 5097
[dataimport] print transaction id when we get one
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 09:58:44 +0200] rev 5096
[cleanup] add some notes
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 09:58:20 +0200] rev 5095
[optimisation] _select_best is used very intensivly. Peanuts optimization may count :)
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 09:57:18 +0200] rev 5094
[optimization] traced_selection is used very intensively in debug mode
optimizing it is worth the pain. Basically : don't do stuff unnecessary
when we don't trace anything.
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 09:55:19 +0200] rev 5093
[optimization] improve massive write performance by optimizing hooks selection
profiling on some massive deletion showed up that 2/3 of the time was spent
in hooks selection. Those changes make it much more acceptable (through
selection is still not negligeable):
* use one registry for each event, so we've much less hooks to check
when emiting an event as well as no more need for the match_event
selector. This required ability to put one appobject
into several registries, using a __registries__ class attribute.
* check for deprecated .enabled at registry initialization time instead
of at selection time
A very simple HooksManager class has been reintroduce to choose the
right registry on call_hooks.
Those optimisations leads to a ~x3 factor of time necessary to delete
16000 entities.
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 31 Mar 2010 09:45:14 +0200] rev 5092
[upgrade] after upgrade, fork command to start instance avoid grave uicfg pb (explicit rules all removed...)
Sylvain Thénault <sylvain.thenault@logilab.fr> [Tue, 30 Mar 2010 19:59:56 +0200] rev 5091
merge