| 00:00.20 | *** join/#storm _mup_ (~mup@li37-90.members.linode.com) |
| 01:01.25 | *** join/#storm bigjools (~quassel@60-241-216-21.static.tpgi.com.au) |
| 01:01.26 | *** join/#storm bigjools (~quassel@canonical/launchpad/bigjools) |
| 02:47.38 | *** join/#storm jamesh (~james@canonical/launchpad/jamesh) |
| 04:52.06 | *** join/#storm munderwo-work (~markl@203-206-133-216.perm.iinet.net.au) |
| 04:53.00 | munderwo-work | Hi all. I've got a mapping table with a ReferenceSet. I've also go some other columns in the mapping table. Can you get to those through the ReferenceSet? |
| 07:04.31 | jamesh | munderwo-work: You won't easily be able to get at those fields if you use the four argument version of ReferenceSet(), since it doesn't give you rows from the linking table. |
| 07:04.51 | *** join/#storm ant384 (~ghoul@unaffiliated/ant384) |
| 07:05.19 | munderwo-work | jamesh: ahh righto.. what i've done instead, is to link to the matching table and then onto the final table.. seems to do what I need to do.. thanks! |
| 07:05.32 | jamesh | If the linking table is important, consider using the two argument form so that the reference set returns the linking table directly, and have that table contain ReferenceSets |
| 07:06.13 | jamesh | alternatively, if you want to get a lot of linked objects plus values from the linking table, consider using store.find() directly |
| 07:06.58 | jamesh | e.g. result = store.find((B, Link.field), B.id == Link.b_id, Link.a_id = a.id) |
| 07:07.13 | jamesh | would give you a result that iterates over (b, field) tuples |
| 07:21.43 | *** join/#storm stub (~stub@canonical/launchpad/stub) |
| 07:48.57 | munderwo-work | jamesh: thanks... If i've chosen the first route. ie use the resultset with only two arguments, is there a way I can filter a find based on something in the resultset? |
| 07:50.51 | jamesh | munderwo-work: are you asking about filtering a result set, or checking for membership in the result, or something else? |
| 07:54.56 | munderwo-work | I want to do a find, and use a column in the mapping table in the where clause of the find. |
| 07:55.54 | jamesh | you can filter a result set by using its find() method, which acts like store.find() minus the first argument |
| 07:56.27 | jamesh | you could probably use that on a four argument ResultSet(), but I'm not sure whether we count that as officially supported or not |
| 07:57.46 | jamesh | i.e. if class A has ResultSet(id, Link.a_id, Link.b_id, B.id), then doing find(Link.field == 42) on that result set would probably work |
| 07:58.30 | jamesh | actually, there probably isn't any reason why we shouldn't say that kind of thing is supported |
| 07:59.44 | jamesh | gar. I mean ReferenceSet above |
| 08:09.23 | munderwo-work | right. In my case i've moved from a 4 argument ReferenceSet to a two to get to the mapping table. How do I then create a where clause that will allow me to filter on a column in the mapping table that isnt involved in the join |
| 08:10.35 | jamesh | well, you can always construct the query manually with store.find() |
| 08:11.34 | jamesh | If you want to filter items inside a reference set (rather than return other results), then ReferenceSet.find() should work okay: you should be able to do extra joins if needed. |
| 08:12.17 | munderwo-work | I basically want to do this http://dpaste.org/IGaev/ |
| 08:13.56 | jamesh | You'd probably need to fall back to store.find() for that. Something like store.find((LicensingUpdate, LicensingUpdateAgreement), LicensingUpdate.id == LicensingUpdateAgreement.update_id, LicensingUpdateAgreement.owner_store_id == 0) |
| 08:14.30 | munderwo-work | Dont worry I think I worked it out. I was having trouble workingout how to reference the owner_store_id.. thanks.. also is there anything wrong with using three expressions when using Or()? |
| 08:14.52 | jamesh | nope. |
| 08:14.53 | munderwo-work | and by having trouble i mean, my brain wasnt working properly |
| 08:15.04 | munderwo-work | jamesh: AWSE! |
| 08:16.10 | jamesh | One other possibility is to create a subexpression with ResultSet.get_select_expr() |
| 08:16.47 | munderwo-work | jamesh: I think i'll go the simple route :) my brain probably cant handler ResultSet.get_select_expr() ;) |
| 08:17.33 | munderwo-work | jamesh: and thats not saying anything bad about storm, but rather that im not far off going home from work :) |
| 08:18.58 | jamesh | get_select_expr() was a fairly recent addition: for a long time we got by without it, but in some cases where there are multiple ways a result set could have been generated, it is quite useful |
| 09:09.58 | *** join/#storm esteve (~quassel@91.126.88.6) |
| 10:25.34 | *** join/#storm niemeyer (~niemeyer@nat/canonical/x-tnrlyzhsncgkejhy) |
| 10:25.34 | *** mode/#storm [+o niemeyer] by ChanServ |
| 11:52.37 | *** join/#storm stub (~stub@canonical/launchpad/stub) |
| 12:10.33 | *** join/#storm niemeyer (~niemeyer@nat/canonical/x-wduhiazwpzhsxxof) |
| 12:10.34 | *** mode/#storm [+o niemeyer] by ChanServ |
| 12:34.21 | *** join/#storm fcorrea (~fcorrea@177.28.216.130) |
| 15:02.27 | *** join/#storm fcorrea_ (~fcorrea@187.83.128.114) |
| 15:54.23 | *** join/#storm ferd (~ferd@106.106.85.200.dsl.dynamic.telviso.net.ar) |
| 15:54.44 | *** part/#storm ferd (~ferd@106.106.85.200.dsl.dynamic.telviso.net.ar) |
| 17:53.39 | *** join/#storm fcorrea (~fcorrea@177.120.49.181) |
| 23:05.22 | *** join/#storm bigjools (~quassel@60-241-216-21.static.tpgi.com.au) |
| 23:05.23 | *** join/#storm bigjools (~quassel@canonical/launchpad/bigjools) |