| 00:00.15 | *** join/#storm wallflower (n=wallflow@ip-205-246-113-216.pool.grokthis.net) |
| 02:26.04 | *** join/#storm jamesh_ (n=james@canonical/launchpad/jamesh) |
| 05:25.23 | *** join/#storm dobee (n=dobeee@85-124-200-100.static.xdsl-line.inode.at) |
| 05:41.22 | *** join/#storm jukart (n=jukart@lsfw01.lovelysystems.com) |
| 05:56.23 | *** join/#storm dobee (n=dobeee@lsfw01.lovelysystems.com) |
| 07:22.15 | *** join/#storm goschtl (n=goschtl@p5B0BD539.dip.t-dialin.net) |
| 07:31.19 | *** join/#storm mcella (n=mcella@ip-178-210.sn2.eutelia.it) |
| 07:43.57 | *** join/#storm atowers (n=atowers@203.206.162.25) |
| 07:48.35 | atowers | question to the floor: does Storm support the full-text searching in postgres 8.3? |
| 07:50.33 | atowers | as in, with some Storm-like syntax :) |
| 07:52.11 | jamesh | atowers: I don't think there is anything in storm.expr to directly support it, but it would be trivial to do |
| 07:52.23 | jamesh | atowers: and you can always use an SQL() token in your query |
| 07:54.19 | atowers | ok thanks, I'll look into adding some support :) |
| 07:55.40 | jamesh | you basically want something like "table.fti @@ ftq('search term')", right? |
| 07:58.03 | jamesh | (or at least that's what we did with 8.2) |
| 08:12.30 | atowers | yep, that's about all we'll be doing. just reading the 8.3 docs re building the 'table.fti' part and keeping it up to date :) |
| 08:13.13 | atowers | hmm, we might need to rank the results as well |
| 08:18.43 | atowers | which might be a little more tricky; an example from the docs: SELECT title, ts_rank_cd(textsearch, query) AS rank FROM apod, to_tsquery('neutrino|(dark & matter)') query WHERE query @@ textsearch ORDER BY rank DESC LIMIT 10 |
| 08:20.46 | jamesh | so, you might want to define some expression objects like tsquery(column, query) and tsrank(column, query) |
| 08:21.01 | jamesh | then define how to compile that to SQL for postgresql |
| 08:22.25 | jamesh | that example looks a bit weird |
| 08:23.49 | jamesh | I'd expect it to be able to handle something like: SELECT title, ts_rank_cd(textsearch, to_tsquery('...')) AS rank FROM apod WHERE to_tsquery('...') @@ textsearch ORDER BY rank DESC LIMIT 10 |
| 08:23.59 | jamesh | and handle it efficiently |
| 08:24.23 | atowers | assuming it can recognise that they're the same query, and not execute it twice :/ |
| 08:24.26 | jamesh | it should know that ts_query is const/pure |
| 08:24.40 | jamesh | see what explain tells you |
| 08:25.38 | jamesh | having a constant in the WHERE clause seems weird |
| 08:25.52 | atowers | will do :) time for me to head home, and I don't have a table set up to try this yet so it'll have to wait. I'll keep you posted, and thanks for your help :) |
| 08:28.14 | atowers | I imagine your example would build two different 'tsquery' objects in the implementation though, although they might cache query strings.. but even so it relies on them to check those two tsquery objects for equivalence and share the result set |
| 08:29.43 | jamesh | atowers: the query optimiser should see them as the same value (and that they are constant for all rows of the query) |
| 08:30.52 | atowers | that's true :) I'll try it out with explain and see what I get. really going this time! :) |
| 09:08.45 | *** join/#storm harnak (n=harnak@ip-118-90-37-181.xdsl.xnet.co.nz) |
| 11:59.59 | *** join/#storm niemeyer (n=niemeyer@200-103-245-169.ctame705.dsl.brasiltelecom.net.br) |
| 12:30.06 | *** join/#storm marianom (n=marianom@ubuntu/member/marianom) |
| 12:34.48 | *** join/#storm harnak (n=harnak@ip-118-90-37-181.xdsl.xnet.co.nz) |
| 13:25.52 | *** join/#storm oubiwann (n=oubiwann@209-181-152-150.omah.qwest.net) |
| 13:45.19 | *** join/#storm bigdog (n=scmikes@8.8.197.72) |
| 14:09.33 | *** join/#storm mcella_ (n=mcella@ip-179-71.sn2.eutelia.it) |
| 14:52.39 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 16:16.15 | *** join/#storm dobee (n=dobeee@85-124-200-100.static.xdsl-line.inode.at) |
| 16:28.51 | *** join/#storm marianom (n=marianom@ubuntu/member/marianom) |
| 16:42.32 | *** join/#storm bigdo2 (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 20:38.23 | *** join/#storm dobee (n=dobeee@85-124-200-100.static.xdsl-line.inode.at) |
| 20:49.15 | *** join/#storm dobee (n=dobeee@85-124-200-100.static.xdsl-line.inode.at) |
| 20:56.48 | *** join/#storm mcella (n=mcella@ip-179-71.sn2.eutelia.it) |
| 21:51.15 | *** join/#storm atowers (n=atowers@203.206.162.25) |
| 22:11.34 | *** join/#storm mcella (n=mcella@ip-179-71.sn2.eutelia.it) |