← Previous day | Today | Next day → | Search | Index
All times shown according to UTC.
Time | Nick | Message |
---|---|---|
12:44 | kados | dewey: 007 is Physical Description Fixed Field (R) |
12:44 | dewey | OK, kados. |
12:45 | kados | 007 |
12:45 | 007? | |
12:45 | dewey | rumour has it 007 is Physical Description Fixed Field (R) |
12:45 | kados | nice |
13:59 | tumer | kados are you around? |
13:59 | kados | tumer: sure am |
13:59 | tumer: probably will be for the next 12 hours or so till I completely crash :-) | |
13:59 | tumer | i have recommited them resorts now |
13:59 | kados | sweet |
14:00 | sorry i didn't commit my stuff | |
14:00 | which resorts too | |
14:00 | I'm thinking about a creating a new attribute set | |
14:00 | for MARC21 | |
14:00 | used for indexing fixed fields and other special Koha fields | |
14:01 | for now I'm just adding on to bib1 | |
14:01 | tumer | i noticed you develoeed a few more entry_points I'll wait to see them |
14:02 | kados | yep, when I'm done I'll commit |
14:02 | (soon as everything's working) :-) | |
14:02 | tumer | somebody will have to write pqf.properties for that |
14:03 | by the way CQL does work even with resort | |
14:03 | but I modified resorting to be a sort after search | |
14:07 | kados | bbiab |
14:32 | tumer: looking at your code now | |
14:32 | ok, cool, I'll merge it with mine and re-commit | |
14:33 | tumer | i know you will oppose the new resort code but it was the quickest way around |
14:33 | kados | owen: search is gonna break for a few minutes :-) |
14:33 | owen | Okay |
14:36 | tumer | owen:sorry I removed the <Display for NPL> parts on this opac-results.tmpl i did not realise you are also using it |
14:36 | owen | No problem. It needed to be taken out |
14:43 | paul | hello world. |
14:43 | tumer | hi paul |
14:44 | kados | hey guys |
14:44 | guess we have a meeting today eh? | |
14:44 | not sure I'll have time to preside | |
14:44 | chris | hi all |
14:44 | ToinS | hi all |
14:45 | chris | i have tech meeting for work in 45 mins, so i will have to leave in about 25 |
14:45 | kados | and zebra config files |
14:46 | so do we have anything so important to talk about that we should stop working? | |
14:46 | ToinS | hehe |
14:46 | paul | yes, we have to speak of 2 things : |
14:47 | - the name of the next stable release + a future 1st ZOOM release | |
14:47 | - what ToinS will work on in 2 weeks. | |
14:47 | chris | right |
14:48 | kados | sure |
14:48 | 1. name of next stable release | |
14:48 | my opinion is to have: | |
14:48 | rel_2_2 => 2.2.6 | |
14:48 | dev_week => 2.3/4 | |
14:48 | head => 3.0 | |
14:49 | chris | i quite like that idea too |
14:49 | tumer | so do i |
14:49 | chris | assuming at some point we go |
14:49 | head -> rel_3_0 => 3.0 releases | |
14:49 | paul | I like it too, although I think we will have a lot of work to synch 3 cvs branches |
14:49 | kados | yep |
14:50 | chris | yep, once 3.0 is released .. should be back to 2 branches again rel_3_0 and head |
14:50 | kados | yep |
14:50 | I'm fine with that ... | |
14:50 | paul | what kind of timeline should we expect ? |
14:51 | kados | now that I've got zebra working, I'm less worried about zebra-based koha |
14:51 | well ... | |
14:51 | paul | I vote for 2.2.6 next week, 2.4 in july/august and 3.0 at the end of the year. |
14:51 | tumer | so do we merge dev_week with rel2_2 |
14:51 | kados | i have clients that need 2.4 within a month or so ... |
14:51 | paul: I agree with your timeline | |
14:51 | chris | me too |
14:51 | paul | it's very important for me to be able to announce 3.0 for the end of the year. |
14:52 | kados | ok |
14:52 | paul | because it will have many new features that are highly waited ! |
14:52 | kados | sure, I understand |
14:52 | chris | getting a 2.4 out, with zoom allows us the time to do 3.0 right |
14:52 | kados | but keep in mind that it's also very important for me that 3.0 be done right |
14:52 | paul | like holidays, news, letters, late returns management... |
14:52 | kados | no shoddy code allowed :-) |
14:52 | paul | shoddy ? |
14:52 | kados | shoddy = badly written |
14:53 | I will be auditing features at some point | |
14:53 | for rel_3_0 | |
14:53 | so if you have a feature you want included, make sure it's cleanly written ;-) | |
14:53 | paul | mmm... badly written from which pov ? |
14:54 | because it can be : | |
14:54 | - functionnaly deficient | |
14:54 | - poor perl | |
14:54 | - poor SQL | |
14:54 | (or even the 3 :-( ) | |
14:54 | kados | functionality / features aren't my primary concern |
14:54 | because we can assume that libraries sponsored that | |
14:54 | paul | kados: ++ |
14:55 | kados | mainly it's poor perl and poor SQL that makes mainteannce a nightmare |
14:55 | chris | the thing that we need to be careful of |
14:55 | is when we change the way something works | |
14:55 | kados | ryan and I have been tearing our hair out the last few weeks |
14:55 | :-) | |
14:55 | chris | is that we do that with a system preference |
14:55 | kados | yep |
14:55 | paul | yep too |
14:55 | chris | for eg, the changes in circulation |
14:56 | i think bob emailed about and arnaud answered, ill follow that up today | |
14:56 | kados | good example |
14:56 | chris | saying i think we should have a system preference for that |
14:56 | kados | definitely |
14:56 | me too :-) | |
14:56 | chris | :) |
14:57 | kados | ok ... so we all set on versioning questions for now? |
14:57 | chris | the change assumes that if you return a book at a library other than its homebranch, you want it back at its homebranch and sets up a transfer for it |
14:57 | paul | so, we are at point 2 it seems : ToinS 1st goal |
14:57 | chris | which is cool, if thats the behaviour u expect |
14:57 | kados | chris: NPL wants this if i understand correctly ... |
14:57 | paul | it seems it will be code cleaning and code cleaning and maybe a little bit code cleaning. |
14:57 | kados | hehe |
14:57 | ToinS | hehe |
14:57 | kados | a good plan I think |
14:57 | paul | I hate coca cola ! |
14:57 | (ads from them on TV just now) | |
14:58 | kados | :-) |
14:58 | chris | code cleaning is great |
14:58 | kados | yep, also a good way to get familiar with the project |
14:58 | paul | yep. |
14:58 | kados | ToinS: feel free to ask us questions |
14:58 | ToinS: on the list or on IRC | |
14:58 | ToinS | ok kados |
14:58 | kados | any time of day :-) |
14:58 | chris | or in cvs commits too |
14:58 | paul | so, 15 days investigating Koha organisation & 1,5 month cleaning code |
14:58 | ToinS | ok ok! |
14:58 | paul | in 2 months he should do a good job |
14:59 | kados | excellent |
14:59 | chris | :-) |
14:59 | kados | anything else to discuss? |
14:59 | paul | kados: & chris repeat this 10 times to ToinS, because he is 2meters far from me, and rarely ask me questions... |
14:59 | chris | and its good timing, because he can watch the world cup too :) |
14:59 | kados | hehe |
14:59 | paul | he must learn that "1 knowing is better than 10 searching" |
15:00 | (french proverb) | |
15:00 | kados | yep ... |
15:00 | no sense wasting valuable time searching when you could just ask :-) | |
15:00 | chris | just on a positive note |
15:00 | ToinS | i've noted |
15:00 | kados | :-) |
15:01 | ToinS | :-) |
15:01 | chris | in the last 24 hours, ive seen commits from paul, tumer, pierrick, bruno, mason, bob, me, joshua, paul, and i think arnaud too |
15:01 | tumer | kados:did you ask chris about contex.pm and xml |
15:01 | chris | its great to see lots of ppl working |
15:01 | paul | chris: right, & I was very happy to see so many commits ! |
15:02 | chris | hehe |
15:02 | kados | tumer: no ... I got it working |
15:02 | tumer: no more hardcoded stuff | |
15:02 | tumer: it just started working one day :-) | |
15:02 | tumer | :-d |
15:02 | chris | he asked me |
15:02 | and then it worked | |
15:02 | im magic :-) | |
15:02 | kados | yea :-) |
15:02 | hehe | |
15:03 | chris | ok, im off to eat breakfast with the katipo tech team |
15:03 | kados | ciao |
15:03 | tumer | bye |
15:03 | kados | many thanks to the katipo team |
15:03 | chris | be back in an hour or 2 |
15:03 | kados | sweet |
15:03 | chris | theyve caught the committing bug |
15:03 | kados | nice |
15:03 | chris | so expect commits to come much more often |
15:03 | paul | chris |
15:04 | a last question | |
15:04 | chris | we still have all the templates to go thru and commit the changes to prog, plus serials |
15:04 | paul | could you explain on koha-devel how branches categories work ? |
15:04 | chris | over the next couple of days |
15:04 | ahh yes | |
15:04 | will do paul | |
15:04 | paul | because i'm not able to explain to SAN |
15:04 | chris | ill email when i get back from meeting |
15:05 | paul | so the meeting is done ? |
15:05 | wow, I told my wife that we will end late ! | |
15:07 | kados | paul: now's your chance to work late :-) |
15:07 | paul | great ! |
15:07 | ToinS | hehe |
15:07 | paul | bye bye everybody |
15:07 | kados | bye paul |
15:07 | tumer | bon soir paul |
15:07 | ToinS | bye paul |
15:08 | tumer: s/bon soir/bonsoir/ ;-) | |
15:08 | tumer | thanks Toins |
15:09 | paul | however, bon soir is "good evening". |
15:09 | but it became bonsoir (in 1 word) some centuries ago. | |
15:11 | tumer | bon nuit then |
15:11 | is that bonnuit as well ? | |
15:11 | ToinS | no |
15:11 | bonne nuit | |
15:12 | tumer | well i give up |
15:13 | kados | tumer: can you set your editor to use 4-character tabs? |
15:13 | tumer | it does not have that facility |
15:14 | kados | bummer |
15:14 | tumer | i'll try and find another editor |
15:14 | kados | since we're working on the same code |
15:14 | it might be nice to agree on a format | |
15:14 | I propose: | |
15:14 | if { | |
15:15 | my $something; | |
15:15 | } | |
15:15 | so every block is indented one tab | |
15:15 | blocks within blocks also | |
15:15 | makes it easier to see where the logic is | |
15:15 | tumer | once i ca find some editor with tabs i fully agree |
15:16 | even for me i use spaces to try t compensate that but its difficult | |
15:18 | kados: it seems i was cut off after: it makes it easier... | |
15:18 | kados | owen: do you know of a good editor for windows? |
15:19 | owen | I've struggled with that question repeatedly. |
15:19 | Lots of folks swear by jedit | |
15:20 | HTML-Kit is another free one (though not open source, I think) | |
15:22 | kados | tumer: just commited a merged opac-zoomsearch.pl |
15:22 | owen: do you want to merge tumer's changes into your template? | |
15:22 | owen | Sure |
15:22 | kados | give me a minute or two more |
15:22 | tumer | k |
15:22 | kados | well ... maybe its' the template, can't be sure |
15:23 | ahh ... | |
15:23 | HTML::Template::param() : attempt to set parameter 'searchdesc' with a scalar - parameter is not a TMPL_VAR! at opac-zoomsearch.pl line 64 | |
15:23 | musta missed that one ... | |
15:25 | huh ... I don't get it | |
15:25 | tumer: any clues? | |
15:25 | does that just mean it's not in the templates? | |
15:25 | tumer | searchdesc was an array ow is not |
15:25 | kados | ahh |
15:26 | in fact, it's not even in the templates | |
15:26 | tumer | it line 17 of zoosearch |
15:27 | it has to be my $searchdesc | |
15:27 | kados | I've got: |
15:27 | my $searchdesc; | |
15:28 | tumer | line 42 and line 66 |
15:28 | kados | then got: |
15:28 | $searchdesc=$cql_query.$pqf_query; | |
15:28 | $template->param(searchdesc => $searchdesc ); | |
15:28 | and that's it | |
15:28 | tumer | all ok |
15:28 | whats in template? | |
15:29 | kados | ahh ... |
15:29 | it's a loop | |
15:29 | so I just delete it and replae with yours | |
15:29 | owen: take note :-) | |
15:30 | tumer | well there were other changes on that template as well |
15:30 | kados | actually ... owen I'm going to just merge them with cvs |
15:31 | owen: no conflicts | |
15:31 | we're in business | |
15:32 | http://zoomopac.liblime.com | |
15:32 | tumer: thanks ... | |
15:32 | tumer: so at this point ... | |
15:32 | tumer: I'll be adding a bunch of new features | |
15:33 | tumer: to the results screen | |
15:33 | tumer: and will commit tonight when I'm done | |
15:33 | owen | kados: looks like I just lost my permissions again |
15:34 | kados | owen: yea ... on purpose :-) |
15:34 | owen: sec | |
15:34 | owen: k, go ahead | |
15:34 | owen: I merged opac-results.tmpl | |
15:34 | owen: appears to be working | |
15:35 | resort isnt' tho | |
15:35 | probably my fault | |
15:36 | tumer: can you briefly tell me how your resort works? | |
15:37 | tumer: you use ZEbra to cache the query? | |
15:37 | tumer | zebra supposed to cahce it itsel |
15:37 | i just do a sort after search on resorts | |
15:38 | kados | ahh ... i see |
15:38 | tumer | but the code for resort both in templates and code is changed |
15:38 | kados | yep |
15:39 | I like your way | |
15:39 | but ... | |
15:39 | tumer | i use "1=4 i< 1=30 >" and aplly teh sort to results->sort |
15:39 | go ahead | |
15:39 | kados | it's still important to store the original query seprately on the results page |
15:39 | because you need it for a 'search within results' | |
15:39 | I'm gonna rewrite it :-) | |
15:40 | tumer: thanks for getting the results integrated | |
15:40 | tumer | origionl query is aleyas stored |
15:40 | kados | (separate from sort_as?) |
15:40 | tumer | yes |
15:40 | kados | ok |
15:41 | tumer | variable forminputs keeps the origional query |
15:42 | kados | got it |
15:42 | tumer | we have a reorder_query variable that keeps the new sort |
15:43 | so you can ask for any variable any time | |
15:44 | but you may have brilliant iadeas like this search module so go ahead and change it | |
15:46 | i just looked at your page no author names? | |
15:50 | kados | owen: what was the variable you needed for tabs? |
15:51 | owen | search_type is what it was in opac-catalogue-home.pl (see line 243, for instance) |
15:53 | kados | ok |
15:53 | I was calling that search_form | |
15:55 | query_form actually | |
15:56 | owen: ok ... it's available now | |
15:56 | owen: knock yourself out :-) | |
15:56 | owen: want a default value for it? | |
15:56 | owen: cql maybe? | |
15:56 | owen | yeah, that'd be good |
15:57 | kados | k |
15:57 | k, it should work now | |
15:57 | owen: anything else I haven't made available to you that you need? | |
15:58 | owen | Not yet |
15:58 | kados | cool |
15:58 | tumer | kados:so are you committing these shall i wait? |
16:00 | owen | kados: where in the script is that default getting set? |
16:00 | ToinS | bye all |
16:05 | kados | owen: top line |
16:05 | my $query_form = $query->param('query_form') | "cql"; # which query form was submitted | |
16:06 | owen | Okay, that's not really what I need. |
16:06 | kados | ok, what do you need? |
16:06 | owen | I need something like we had in the old script: |
16:06 | my $search_type = $query->param('search_type'); | |
16:06 | if ((!$search_type) || ($search_type eq 'cql')) { | |
16:06 | $template->param(cql_search => 1); | |
16:06 | } elsif ($search_type eq 'precise') { | |
16:06 | $template->param(precise_search => 1); | |
16:06 | } elsif ($search_type eq 'keyword') { | |
16:06 | $template->param(keyword_search => 1); | |
16:06 | } elsif ($search_type eq 'loose') { | |
16:06 | $template->param(loose_search => 1); | |
16:06 | } | |
16:06 | so that I can say <!-- TMPL_IF NAME="cql" --> | |
16:07 | kados | gotcha |
16:07 | owen | ...because I can't say <!-- TMPL_IF queryform = "cql" --> |
16:07 | kados | right |
16:07 | ok, no prob | |
16:07 | just a sec | |
16:10 | you want them named 'cql_search' or just cql? | |
16:10 | owen | doesn't matter |
16:12 | kados | ok ... done |
16:13 | cql_search | |
16:13 | advanced_search | |
16:13 | power_search | |
16:13 | proximity_search | |
16:15 | owen: btw ... feel free to add a link on every search point to a help pop-up | |
16:15 | help.pl | |
16:15 | I'll enable it eventually | |
16:15 | (eventually being sometime in the next 12 hours :-) | |
16:16 | tumer | goodnight all |
16:16 | kados | night tumer |
16:26 | owen | kados: Under the advanced search, the 'format' options we could customize with our own itemtype groupings? |
16:26 | kados | correct |
16:26 | under power search too | |
16:27 | owen: I've got them as: | |
16:27 | ta = regular print | |
16:27 | tb = large print | |
16:27 | vf = vhs | |
16:27 | vd = video disk | |
16:27 | sd = sound disk - book on CD or music CD | |
16:27 | ss = sound cassette | |
16:27 | co = CD-ROM | |
16:27 | (ignore the two letter code) | |
16:27 | go ahead and throw those inputs in there with blank values | |
16:28 | I'll populate the values when I've got the index updated | |
16:28 | owen | hunh? |
16:28 | inputs with blank values? | |
16:28 | kados | just put in: |
16:28 | <input value="">Video Disk</input? | |
16:28 | > even | |
16:29 | does that make sense? | |
16:29 | then ... I'll come along tonight | |
16:29 | and turn that int: | |
16:29 | <input value="@attr 1=8003 vd">Video Disk</input> | |
16:29 | or whatever it is | |
16:30 | (it'll actually be much more complicated than that | |
16:30 | owen | Are we forced by the leader values to accept books on CD and music CDs lumped together? |
16:30 | kados | no |
16:30 | in fact, they distinguish | |
16:30 | put what you want in the inputs | |
16:30 | owen | /whew/ |
16:30 | kados | I'll figure out how to make it work :-) |
16:31 | owen | We never /wanted/ to call DVDs 'video discs,' it was just a kludge to get them to show up next to videos in the format list |
16:31 | kados | right |
16:31 | well you can go: | |
16:31 | Videodisk (DVD) | |
16:31 | like worldcat does | |
16:32 | we're not tied to a specific order anymore | |
16:32 | at least not for the formats | |
16:44 | owen | kados: I don't know how much time I'll have this evening, but I'll try to check in |
16:44 | kados | don't sweat it |
16:45 | thanks for working on it today | |
16:45 | we've made excellent progress | |
16:45 | I think I can get it at least in a beta phase by tomorrow morning | |
16:45 | I'm actually gonna take dinner in a minute or so | |
16:45 | owen: anything you need before I do? | |
16:46 | owen | No, I'm about to take off too |
16:46 | kados | cool |
16:46 | have a good evening | |
16:46 | thanks again | |
16:46 | owen | Likewise |
17:15 | chris | back |
19:00 | kados | chris: seen this lately: |
19:00 | dewey | I haven't seen 'this', kados |
19:00 | kados | http://zoomopac.liblime.com/cg[…]?query_form=power |
19:00 | still got a ways to go | |
19:00 | ie, not everything works yet :-) | |
19:00 | but we're getting there | |
19:01 | in fact, the db is currently re-indexing | |
19:01 | so it's completely broken now :-) | |
19:01 | 15 minutes | |
19:02 | chris | :-) |
19:03 | kados | now that I understand it, zebra totally rocks |
19:03 | :-) | |
19:04 | and I'm actually kinda liking MARC21 at the moment | |
19:04 | strange as that may seem :-) | |
19:04 | chris | oh i dont mind it |
19:04 | its fine for what its desinged for | |
19:04 | just dont make people have to see it | |
19:04 | :-) | |
19:04 | kados | right :-) |
19:06 | chris | for a data transport/storage layer its fine |
19:06 | you dont want to search it | |
19:06 | you want to search a representation of it .. which is what zebra gives us | |
19:06 | kados | yep |
19:07 | chris | yeah i think that zebra gives the opportunity to do a lot of really cool things |
19:07 | and if we get 2.4 out | |
19:08 | kados | yep |
19:08 | chris | that will give us time to get 3.0 right |
19:08 | kados | yea |
19:08 | chris | :-) |
19:08 | kados | i think I'll have the beta of the search working tonight |
19:08 | chris | zebra stores its indexes as files right? |
19:09 | kados | well ... yes, they are files |
19:09 | but you can't really tell what they're for I think | |
19:09 | chris | ah right |
19:09 | kados | everything is stored as binary |
19:09 | chris | cos i was thinking |
19:09 | you could have a cron job watching them, and cvs committing any time they change | |
19:10 | but if they are binary that wont work | |
19:10 | was a 3am idea, so is probably daft :) | |
19:11 | kados | heh |
19:12 | well what we have | |
19:12 | is just as good i think | |
19:12 | we can just save the xml records | |
19:12 | in a directory | |
19:12 | chris | yeah true |
19:12 | and rebuild | |
19:12 | kados | the index isn't a big deal |
19:12 | you can rebuild it in 15 minutes | |
19:13 | and with shadow registers | |
19:13 | you can run the db at the same time | |
19:13 | chris | yep |
19:13 | kados | a commit takes under a minute |
19:13 | chris | sounds good |
19:13 | kados | with 150k records |
19:13 | chris | sweet |
19:13 | kados | so worse case scenerio |
19:13 | you're down 60 seconds :-) | |
19:13 | chris | :) |
19:14 | kados | the speed is just crazy fast |
19:14 | chris | another crackpot idea i had |
19:14 | kados | yea? |
19:14 | chris | was you could have multiple zebra |
19:14 | on different machines | |
19:14 | an update does them all | |
19:14 | a search round robins | |
19:15 | kados | yea, that'd be good for super large collections |
19:15 | but honestly | |
19:15 | our bottleneck here at this point is web traffic and cgi | |
19:15 | chris | yep |
19:15 | which is cool | |
19:15 | kados | zebra is really fast |
19:15 | and it doesn't get slower the more records you add | |
19:15 | because of the structure of the index | |
19:15 | chris | sweet |
19:16 | kados | npl has a language search now :-) |
19:16 | chris | sweet |
19:16 | kados | oh ... btw ... |
19:16 | any ideas on this | |
19:16 | I've got a form like this: | |
19:16 | <form> | |
19:16 | <input name="query1"> | |
19:16 | <input name="op1"? | |
19:17 | > even | |
19:17 | <input name="query1"> | |
19:17 | <input name="op2"> | |
19:17 | query2 I mean :-) | |
19:17 | chris | right |
19:17 | kados | so ... |
19:17 | op2 should only be used if query1 and query2 are used | |
19:17 | chris | right |
19:18 | kados | I'm stumped |
19:18 | chris | how to do it in that specific case, or in a generalised way? |
19:18 | kados | in a generalized way |
19:18 | well let me back up | |
19:19 | here's how i catch those inputs | |
19:19 | my @attributes = ("query","op") | |
19:19 | then ... | |
19:19 | foreach (1..10) { | |
19:20 | $query = $query->param('query$i'); | |
19:20 | $op = $qpery->param(op$i); | |
19:20 | } | |
19:20 | chris | k |
19:20 | kados | well ... slightly more complicated than that actually |
19:20 | chris | :) |
19:20 | kados | in fact, I really need a better way to handle the RPN query |
19:21 | it's a prefix notation | |
19:21 | so 4+5 | |
19:21 | looks like | |
19:21 | + 4 5 | |
19:21 | chris | bnf |
19:21 | thats called | |
19:21 | kados | cool |
19:21 | it gets fairly complex | |
19:21 | because each operand can be a whole query | |
19:21 | so you can have really complex structures | |
19:21 | chris | well a subset of it |
19:21 | right | |
19:22 | kados | any ideas? |
19:22 | i was looking at what Mike Taylor did in Net::Z3950::SimpleServer | |
19:22 | which is to have 'augmented classes' | |
19:22 | you basically override a render method on a bunch of classes | |
19:23 | and walkk the tree | |
19:23 | chris | yeah |
19:23 | kados | but I don't quite grok how to build that |
19:23 | chris | ur having to build a compiler :) |
19:23 | kados | hehe |
19:23 | cool :-) | |
19:24 | I hear there's some way called FIFO | |
19:24 | first in first out | |
19:24 | chris | yep |
19:24 | kados | but don't quite understand how that works |
19:24 | (well, haven't properly googled it either :-)) | |
19:24 | but I'm going to have to replace my current scheme | |
19:24 | it has one flaw | |
19:25 | which is that when you put in a query in the second, third, etc | |
19:25 | it adds the operator | |
19:25 | which makes it invalid RPN | |
19:25 | (if there is no first term) | |
19:25 | chris | right |
19:25 | kados | sorry if that doesn't make sense :-) |
19:25 | chris | no it does |
19:26 | there are 2 approaches to this | |
19:26 | kados | k |
19:26 | chris | 1 force the form to send compliant data before we parse |
19:26 | ie if you get something that has 2 and 3 filled but not one, force the user to fill in 1 | |
19:26 | kados | difficult to do in html ... imposible I think |
19:26 | chris | no do it in the script |
19:27 | but before you try to parse it into rpn | |
19:27 | kados | yea, that wont' work |
19:27 | because sometimes I'm gonna want to just search for everything in spanish | |
19:27 | chris | the other approach is to build a parser that tries to correct users mistakes |
19:27 | kados | which is not the first query |
19:27 | hmmm | |
19:28 | like 'count the number of operands' | |
19:28 | count the number of operators | |
19:28 | make operators = operands - 1 | |
19:28 | chris | yep |
19:28 | kados | ok |
19:28 | easy enough | |
19:28 | only time it would ever mess up | |
19:28 | chris | one way |
19:28 | is | |
19:28 | kados | is with @or i think |
19:29 | chris | for (my $i=0;$i<11;$i++){ |
19:29 | actually | |
19:29 | for (my $i=1;$i<11;$i++){ | |
19:29 | in this case | |
19:29 | kados | right |
19:29 | chris | hmm lemme think ok, here goes |
19:30 | kados | the way I do it now |
19:30 | is to run through them twice | |
19:30 | once for the operator | |
19:30 | s | |
19:30 | once for the queries | |
19:30 | so instead of putting them into scalars | |
19:30 | I'll just push them into arrays | |
19:31 | that'll make it a piece of cake | |
19:31 | chris | yep |
19:31 | kados | so the question is, which one to drop :-) |
19:31 | chris | you could construct an thing |
19:31 | a thing | |
19:31 | kados | you've got @and, @or, @not |
19:31 | ahh ... | |
19:32 | it's aleays going to be the first one | |
19:32 | always | |
19:32 | chris | yep |
19:32 | kados | cool |
19:32 | so that's a since | |
19:32 | sinch even :-0 | |
19:32 | my new macbook has a slightly different keyboard | |
19:32 | still getting used to it | |
19:32 | btw ... triple booting :-) | |
19:32 | chris | :) |
19:32 | kados | got debian installed last night |
19:32 | etch | |
19:32 | and winxp too | |
19:33 | chris | cool |
19:33 | kados | if only there was some way to have them all running at the same time |
19:33 | chris | good for testing |
19:33 | kados | definitely |
19:33 | cool, thanks for brainstorming with me | |
19:33 | chris | np |
19:34 | kados | index is rebuilt |
19:34 | if you want to give some searches a shot | |
19:34 | just keep in mind that limitation | |
19:34 | in the power search, I put an 'a' in the first search box | |
19:34 | which catches it | |
19:34 | chris | ah ok |
19:36 | sweet | |
19:36 | kados | maori :-) |
19:37 | hehe | |
19:37 | is the title 'maori'? | |
19:37 | chris | legends of aotearoa |
19:37 | kados | ahh |
19:37 | well you can try out the exact title search | |
19:37 | it's in the advanced tab | |
19:37 | one-word titles even work | |
19:38 | chris | ill give that bash |
19:38 | kados | like 'cell' |
19:38 | :-) | |
19:38 | chris | if i can spell |
19:39 | sweet got it | |
19:40 | so i guess the next thing to try | |
19:40 | is making the stemming work? | |
19:40 | eg | |
19:40 | http://zoomopac.liblime.com/cg[…]d&query10=&sort_b | |
19:40 | hmm big url | |
19:40 | title search (not exact) | |
19:40 | kados | stemming doesn't work |
19:40 | chris | on legend of aotearoa (ie i left the s off) |
19:40 | kados | but ... |
19:41 | truncation does | |
19:41 | so if you set truncation to right-truncated | |
19:41 | it'll find it | |
19:41 | chris | ahh sweet |
19:41 | that will work | |
19:41 | kados | what i want to sponsor |
19:41 | is stemming and phonetics | |
19:41 | it's simple really | |
19:41 | chris | yeah thatd rock |
19:41 | kados | you implement them as indexes |
19:41 | chris | so currently |
19:41 | kados | simple I htink |
19:41 | chris | if i went |
19:41 | legend's | |
19:41 | it wouldnt find legends eh? | |
19:42 | kados | ' is equiv to a space |
19:42 | and if you did it as a fuzzy search | |
19:42 | it would allow that one error | |
19:42 | chris | sweet |
19:42 | cos that would rule | |
19:42 | kados | yep :-) |
19:42 | chris | so many ppl leave ' out |
19:42 | or add them | |
19:42 | kados | part of the learning process |
19:42 | is what the defaults should be | |
19:42 | chris | yeah |
19:42 | kados | you can also specify 'process # in term' |
19:42 | chris | i wonder if we could have different sets of defaults |
19:42 | kados | and then you could go: |
19:43 | chris | that you can choose at install? |
19:43 | kados | legend#s |
19:43 | well | |
19:43 | in fact ... | |
19:43 | we are going to need to store everything in the db | |
19:43 | right now | |
19:44 | you basically put the @attr 1=X 2=X, etc. | |
19:44 | right in the html | |
19:44 | but eventually | |
19:44 | chris | ahh right |
19:44 | kados | (in 3.0) |
19:44 | we are going to want to have a tmpl var | |
19:44 | chris | yeah |
19:44 | kados | once I'm done with this |
19:44 | I'll propose a table structure for it | |
19:44 | I've already been working on some ideas | |
19:44 | chris | sweet |
19:44 | kados | the idea is |
19:45 | once you get your indexes built | |
19:45 | you can basically don any query | |
19:45 | anyway ... back to coding :-) | |
19:45 | I've got a demo tomorrow :-) | |
19:45 | chris | good luck :) |
20:16 | kados | hehe |
20:17 | it works :-) | |
20:17 | now you can just do a search on all the spanish books | |
20:17 | or french | |
20:17 | etc. | |
20:22 | chris | sweet |
23:04 | kados | chris: got a sec? |
23:04 | I go: | |
23:04 | $zoom_query_obj = new ZOOM::Query::CQL2RPN($query,$zconn); | |
23:04 | use Data::Dumper; | |
23:04 | warn "CQL STUFF:"; | |
23:04 | warn Dumper($zoom_query_obj); | |
23:04 | and I get: | |
23:04 | $VAR1 = bless( { | |
23:04 | '_query' => bless( do{\(my $o = 150526776)}, 'ZOOM_query' ) | |
23:04 | }, 'ZOOM::Query::CQL2RPN' ); | |
23:04 | chris | wk |
23:04 | -w | |
23:05 | kados | does that look like cacheing or something? |
23:05 | (I'd like to be able to derive the RPN from the CQL | |
23:05 | chris | hmm i cant tell |
23:06 | kados | so I can pass on the RPN to the results page for reordering, etc.) |
23:06 | chris | right |
23:06 | kados | btw ... with advanced search the reordering works really nicely |
23:06 | chris | cool |
23:07 | kados | hmmm |
23:07 | chris | is $o always the same |
23:08 | if you do that twice | |
23:08 | is the $o= the same? | |
23:08 | kados | no |
23:08 | it changes every time | |
23:08 | but ti's not incremental | |
23:08 | chris | twice in the same script i mean |
23:08 | kados | ahh ... lemme check |
23:09 | chris | it looks like a timestamp to me |
23:09 | kados | in that case it's the same |
23:09 | chris | ok |
23:09 | what if you go sleep 5 between the calls | |
23:09 | kados | hehe |
23:09 | ok | |
23:09 | chris | does it increase by 5? |
23:10 | kados | still identical |
23:10 | chris | right |
23:10 | so it looks like it caches | |
23:10 | kados | but where ... |
23:10 | chris | but only while the script is running |
23:10 | kados | in yaz? |
23:10 | or in zebra? | |
23:10 | chris | which makes sense |
23:10 | not zebra | |
23:11 | kados | so it transforms the CQL to RPN |
23:11 | and stores it somewhere | |
23:11 | how do I get it out? | |
23:11 | chris | good question |
23:11 | can you make zebra give it to you? | |
23:12 | kados | dunno |
23:12 | I'm reading the ZOOM docs | |
23:12 | don't see anything on it | |
23:13 | chris | zebra does the transform right? using the pqf.properties file? |
23:15 | it outputs it into the log | |
23:15 | kados | yea |
23:15 | the log? | |
23:15 | chris | so i wonder if there is some way to query for it |
23:15 | kados | I'm not sure it's zebra doing the transform |
23:15 | chris | ie if you watch the output of the zebrasvr you can see the rpn |
23:16 | kados | right |
23:16 | chris | i think it must be |
23:16 | because net::z3950 doesnt know about the pqf.properties | |
23:16 | kados | well ... actually ... :-) |
23:16 | you have to tell it about pqf if you use the CQL2RPN | |
23:17 | chris | ahh |
23:17 | so in that case it must convert it | |
23:17 | or the c libraries do | |
23:17 | im guessing the c it wraps does | |
23:17 | kados | ZOOM::Query::CQL2RPN |
23:17 | Implements CQL by compiling it on the client-side into a Z39.50 Type-1 (RPN) query, and sending that. | |
23:18 | chris | cool so if its in the perl thats easy |
23:18 | but i suspect its in the c | |
23:18 | kados | yea, prolly |
23:18 | not a big deal | |
23:18 | it only affects the CQL | |
23:18 | I can concern myself with that later | |
23:18 | chris | right |
23:18 | kados | bigger fish to fry ... |
23:18 | :-) | |
23:19 | or in my case, bigger tofu to fry :-) | |
23:22 | chris | :) |
23:29 | kados | chris: is there a way to send an array of variables through in CGI? |
23:29 | can I have more than one value with the same name? | |
23:29 | and pick it up as an array? | |
23:29 | chris | yes |
23:29 | easy peasy | |
23:30 | my @array=$input->param('variablename'); | |
23:30 | kados | ahh, sweet |
23:30 | thx | |
23:30 | chris | wanna send someone free koha cds? |
23:30 | :-) | |
23:31 | thats what the foundation could do :) | |
23:31 | kados | yep |
23:31 | next week even :-) | |
23:31 | chris | id offer to send john one, but itd cost a lot from nz :-) |
23:31 | kados | heh |
23:31 | I need to catch up on list email | |
23:32 | hopefully after tomorrow's demo I'll have a few hours | |
23:32 | chris | :-) |
23:32 | someone else might offer | |
02:01 | Comete | bonjour à tous ! |
02:01 | hi all | |
02:01 | paul | hello Comete & osmoze |
02:01 | Comete | hi paul and osmoze |
02:01 | osmoze | hello too :) |
02:01 | Comete | paul: a little bug http://o5.bureau.paulpoulain.c[…]etail.pl?bib=2932 |
02:02 | paul: the display is bad here | |
02:08 | hi pierrick | |
02:08 | pierrick | hi Comete |
02:21 | chris | evening all |
02:21 | or morning :-) | |
02:23 | what time is your demo? | |
02:23 | paul | hello kados & chris |
02:23 | chris | will you have a chance to sleep before it? |
02:24 | heya paul | |
02:25 | kados | it's at 9:00am |
02:25 | it's 3:30 right now | |
02:25 | so I'll probably get a few hours of sleep | |
02:25 | I'm cleaning up the results set currently | |
02:25 | and reindexing | |
02:25 | :-) | |
02:25 | chris | :) |
02:27 | ToinS | hi all |
02:27 | chris | hi toins |
02:29 | ToinS | hi chris |
02:32 | pierrick | hi all :-) |
02:33 | chris | hi pierrick, thanks for fixing that sql |
02:33 | pierrick | fixing? I only reformatted things :-) Do you like the presentation? |
02:33 | chris | yes, much easier to read |
02:34 | paul | although I think listing each field on a single line will make too long queries |
02:34 | (hi pierrick) | |
02:34 | chris | i was converting things to capitalised SQL words as I was going yesterday |
02:34 | but then it got towards the end of the day | |
02:34 | and i just started committing madly :-) | |
02:35 | pierrick | paul, many lines is less important than readibility IMO |
02:35 | paul, think of the result of a "grep" for example | |
02:35 | paul | yes, but I think that 15 lines just for a SQL query is not the best for readability |
02:36 | pierrick | paul, how long should a line in a SQL query be? |
02:37 | kados | has everyone seen Koha 2.3 (or what will be 2.3 very soon): http://zoomopac.liblime.com/ |
02:37 | (the search anyway) | |
02:37 | paul | kados: yes, I saw. quite impressive... |
02:38 | I don't understand every term in "power search". So i'm not a power user ;-) | |
02:38 | kados | hehe |
02:38 | there is a new 'resort by' feature | |
02:38 | chris | :) |
02:38 | kados | it works for the non-simple searches only right now |
02:38 | simple search will work soon ... | |
02:39 | pierrick | kados, I like the design of zoomopac :-) |
02:40 | kados | and I think if you put multiple terms in a non-simple search box, you need to enclose them in quotes still |
02:40 | chris | i like the speed :-) |
02:40 | kados | though obviously that will be fixed |
02:40 | I like the speed and the nice search feature | |
02:40 | for istance, you can find all the french books in NPL's catalog | |
02:40 | something previously impossible | |
02:41 | hmmm | |
02:42 | chris | no books on koha |
02:42 | kados | hehe |
03:24 | Icez | . |
03:28 | Icerz | hellol |
03:28 | chris | hi |
03:28 | dewey | hola, chris |
03:29 | Icerz | anyone here from new zealand |
03:29 | chris | i am |
03:29 | Icez | . |
03:29 | Icerz | . |
03:31 | paul | ... /me suspect icexxx to be a bot... |
03:32 | Icez | really |
03:32 | nah i'm normal just like you | |
03:32 | paul | so what means winamp... ? |
03:33 | & sorry if you are not a bot, because your 1st message was quite strange ;-) | |
03:33 | Icez | it's a script |
03:34 | chris | so what brings you to #koha? |
03:37 | Icez | nm aye |
03:37 | just looking on the net for irc servers (NZ) | |
03:37 | chris | ahh well this is for the discussion of the koha project |
03:37 | www.koha.org | |
03:37 | so wont be interesting unless you are interested in that | |
03:39 | Icez | you delevope software? |
03:39 | chris | yes |
03:39 | for use in libraries | |
03:40 | Icez | cool |
03:49 | chris | paul: http://www.bigballofwax.co.nz/[…]6/06/08#cvs-stats |
03:50 | paul | mine are better : http://bureau.paulpoulain.com/cvslog/ |
03:50 | ;-) | |
03:50 | chris | that using statcvs ? |
03:51 | paul | nope, it uses cvschangelogbuilder |
03:51 | (link at end of page) | |
03:51 | in fact, they are quite differents things. | |
03:51 | chris | ahh |
03:52 | paul | mine show cvs history while your show cvs weekly activity |
03:52 | chris | yep, all right if i link to yours from my blog? |
03:52 | paul | of course. |
03:53 | who commits as sushi ? | |
03:53 | chris | mason, from katipo |
03:56 | icez: would you be able to turn that part of your script off please? | |
03:56 | pierrick, you are plg on savannah? | |
03:57 | pierrick | I am |
03:57 | Pierrick Le Gall | |
03:59 | chris | of course, silly me :) |
04:28 | tumer | morning all |
04:28 | chris | hi tumer |
04:28 | paul | hi tumer |
04:29 | tumer | hi all |
04:30 | in october i have 80 university library directors coming for a meeting. All very interested in koha | |
04:31 | paul | mmm... sounds very interesting. |
04:31 | chris | wow |
04:31 | tumer | i have to get this zoom perfercted and guess what kados is now 'our man' for zoom |
04:31 | chris | thats a lot of directors |
04:31 | paul | if you need someone to come to speak, don't hesitate to ask, i'm sure someone will volunteer to come to cyprus ;-) |
04:31 | chris | heh |
04:32 | tumer | you can come to cyprus anytime no need to speak |
04:33 | has anybody noticed a bug with addbiblio.tmpl | |
04:33 | paul | what kind of bug ? |
04:33 | tumer | you cannot repaeat a subfield more than once |
04:34 | or is it a windows problem? | |
04:35 | chris | ahh i think i remember kados talking about that |
04:35 | i think he may have fixed it somewhere, maybe it was in head? | |
04:36 | tumer | i think paul also made some changes recently but dont know which version now (having 3!) |
04:37 | chris | ill take a quick look |
04:37 | tumer | so yours ok than paul? |
04:38 | chris | http://cvs.savannah.nongnu.org[…]oot=koha&view=log |
04:39 | tumer | thanks chris |
04:56 | paul: yours seems to have the same problem. Once duplicated it can not be moved or repeated again | |
05:14 | Comete | paul: did you see the url i gave you this morning ? |
10:05 | paul | kados around ? |
11:31 | bye world. |
← Previous day | Today | Next day → | Search | Index