| 00:00.44 | *** join/#storm wallflower (n=wallflow@ip-205-246-113-216.pool.grokthis.net) |
| 00:09.31 | *** join/#storm jml (n=jml@li2-200.members.linode.com) [NETSPLIT VICTIM] |
| 01:50.35 | *** join/#storm dotz (n=dotz@chello089078009126.chello.pl) |
| 02:17.10 | *** join/#storm bigdo1 (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 05:08.19 | *** join/#storm hazmat (n=kapil@plone/hazmat) |
| 05:18.11 | *** join/#storm jukart (n=jukart@lsfw01.lovelysystems.com) |
| 06:12.38 | *** join/#storm bigdo1 (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 07:03.34 | *** join/#storm dobee (n=dobee@lsfw01.lovelysystems.com) |
| 07:23.36 | *** join/#storm mcella (n=mcella@ip-176-30.sn2.eutelia.it) |
| 08:09.58 | *** join/#storm jukart (n=jukart@lsfw01.lovelysystems.com) |
| 08:40.45 | *** join/#storm mcella (n=mcella@ip-176-30.sn2.eutelia.it) |
| 10:44.50 | *** join/#storm andrea-bs (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 12:08.10 | *** join/#storm tlynn (n=tlynn@host81-149-244-57.in-addr.btopenworld.com) [NETSPLIT VICTIM] |
| 13:01.23 | *** join/#storm kov (n=kov@debian/developer/kov) |
| 13:18.46 | *** join/#storm jrydberg (n=johan@212.112.160.94) |
| 14:54.33 | *** join/#storm reacocard (n=reacocar@rccy-06-1010.dsl.iowatelecom.net) |
| 15:53.05 | *** join/#storm andrea-bs_ (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 16:10.30 | *** join/#storm reacocard (n=reacocar@rccy-06-1010.dsl.iowatelecom.net) |
| 16:31.04 | reacocard | quick question: If I create a store, load and object from it, close the store, then open another storn on the same db and add the object to it, does it add another copy of the object, or overwrite the existing object in the db? |
| 16:32.19 | reacocard | assume that I haven't changed the Id of the object |
| 16:33.41 | therve | store.add produces a SQL INSERT |
| 16:33.56 | therve | so if you didn't change the id, it's likely to raise an IntegrityError |
| 16:34.10 | therve | but you shouldn't add it to the store the second time |
| 16:34.18 | reacocard | so how do I make it overwrite then |
| 16:35.19 | therve | well, you either get the original object and update it, or get the original object, delete it, and add a new one |
| 16:35.27 | therve | the first solution is of course better |
| 16:35.43 | reacocard | hrm |
| 16:36.10 | therve | storm is not an object database |
| 16:36.19 | reacocard | yeah I know |
| 16:36.28 | reacocard | this just makes things little trickier |
| 16:36.37 | reacocard | ah well, better than dealing with SQL |
| 16:36.48 | therve | if you think of a relational db, this is fairly natural |
| 16:38.52 | reacocard | so, when someone requests to add an object again, I should check to see if there's one with that id already, and if so, update its values from the new object? |
| 16:40.09 | therve | reacocard: how do you know if you add an object or if you overwrite an object? |
| 16:40.40 | reacocard | not necessary in this case. all that matters is that the data that was requested to be added ends up in the db |
| 16:41.28 | therve | reacocard: then add it? I don't see the problem |
| 16:41.47 | reacocard | but I don't want a duplicate object, hence the need to overwrite |
| 16:42.06 | reacocard | I can't keep the store open all the time since that could be hours |
| 16:42.37 | therve | I don't see how "not necessary in this case" can be an answer to my previous question then |
| 16:42.46 | reacocard | so, when a change is made I have to re-open the store, commit the changes, and close it again |
| 16:43.00 | reacocard | perhaps I dont understand the question properly |
| 16:43.38 | therve | reacocard: you have an object. How do you know this object is new, or if it has to overwrite a previous one? |
| 16:43.54 | reacocard | by the Id |
| 16:44.06 | reacocard | if it has an Id, it overwrites the one with that Id |
| 16:44.19 | reacocard | if not, we add nromally and it gets an Id |
| 16:44.51 | therve | ok |
| 16:45.04 | reacocard | the only problem is I can't keep the store open continuously while the object exists, hence my original question |
| 16:45.36 | therve | I don't see why you can't keep the store open. |
| 16:45.44 | therve | and I don't see how it's related |
| 16:47.09 | therve | you have an object to update |
| 16:47.12 | reacocard | whatever. if I just check for existing Id it works, no? |
| 16:47.30 | therve | you do something like obj = store.get(MyClass, id); obj.stuff = "stuff" |
| 16:47.40 | therve | but you don't have to check for existing Id |
| 16:47.52 | therve | because when you have an Id you know it exists |
| 16:48.04 | therve | if you don't have an Id it's a new object |
| 16:48.16 | reacocard | the thing is, the obj is gotten, then the store is closed, then re-opened. does it still save my changes to this obj? |
| 16:48.56 | reacocard | assuming changes are made while store is closed |
| 16:49.13 | therve | well, I think you can guess this one? |
| 16:49.20 | reacocard | since I have other tasks going on that will be altering the DB in the meantime and thus need store access |
| 16:49.23 | reacocard | right, they arent |
| 16:49.26 | therve | a store is equivalent to a db connection |
| 16:49.29 | reacocard | right |
| 16:56.19 | reacocard | so what I should do is just enforce creating new stores every time something needs ot be changed, and leave one store open all the time for my UI's read purposes? |
| 16:57.16 | therve | I would just have one store opened all the time |
| 16:58.29 | reacocard | so, use just one store and comiit(), even though I have several threads? I guess that works |
| 16:58.48 | therve | oh several threads |
| 16:58.52 | reacocard | yeah |
| 16:58.58 | therve | you'd better have one store per thread |
| 16:59.05 | reacocard | that's what I thought |
| 16:59.30 | reacocard | so, new stores when things need to be written or for alt threads, one persistent one for the UI |
| 17:00.01 | reacocard | if that works, I can manage that |
| 17:00.20 | therve | why don't keep the stores opened in the threads? |
| 17:00.40 | reacocard | becasue they won't run very long, no more than a few minutes, and then the thread ends |
| 17:00.52 | therve | ok |
| 17:00.58 | reacocard | so you open it for the duration of the task, commit(), and close out |
| 17:01.01 | therve | and why not call something in the main thread? |
| 17:01.48 | reacocard | because the main thread is GTK |
| 17:03.27 | therve | I mean, make your threads call specific methods in the main thread where you store resides? |
| 17:03.40 | *** join/#storm dobee (n=dobee@85-124-200-100.static.xdsl-line.inode.at) |
| 17:03.52 | *** join/#storm lightyear (n=ben@nat/fluendo/x-f2ed0fa8e4b57417) |
| 17:03.56 | lightyear | hey guys |
| 17:03.59 | reacocard | I dont follow |
| 17:04.03 | lightyear | have a storm related question |
| 17:04.31 | lightyear | can I join a reference-set using the find on it? |
| 17:04.50 | lightyear | lets say I have an artist and that has tracks what is a reference-set |
| 17:05.17 | lightyear | and then I have albums, which I can find using a track.album (that is a reference on its own) or track.album_name |
| 17:05.28 | lightyear | is it possible to do something like |
| 17:05.55 | lightyear | artist.tracks.find(Album, Album.name == Track.album_name) then? |
| 17:06.12 | lightyear | to get a list of all the albums? |
| 17:06.57 | lightyear | i tried to do it but then sqlite tells me that it is missing the column for album (which works fine in any other case) |
| 17:16.42 | reacocard | incidentally, is it possible to make values() return only unique results? a 500-item list is a lot easier to deal with than a 30,000 one. |
| 17:18.11 | *** join/#storm jukart (n=jukart@85-124-221-45.static.xdsl-line.inode.at) |
| 17:27.16 | jkakar | reacocard: result.config(distinct=True); result.values(...) |
| 17:27.39 | reacocard | jkakar: awesome, thanks |
| 17:27.51 | jkakar | reacocard: np |
| 18:39.46 | *** join/#storm andrea-bs (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 18:58.08 | *** join/#storm jukart (n=jukart@85-124-221-45.static.xdsl-line.inode.at) |
| 19:42.01 | *** join/#storm jukart (n=jukart@85-124-221-45.static.xdsl-line.inode.at) |
| 23:37.13 | *** join/#storm jbot (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 23:37.13 | *** topic/#storm is The Storm Python ORM - http://storm.canonical.com/ - 0.12 released! || Review branches: https://code.launchpad.net/~storm/storm/trunk/+landing-candidates |