IRC log for #koha, 2006-07-18

All times shown according to UTC.

Time S Nick Message
15:45 russ morning all
15:45 chris morning
15:51 czokie chris - updatedatabase in 2.3 does not create all required preferences fields.....
15:52 czokie for new installs - would probably work for a 2.2 -> 2.3 upgrade ... but not for a fresh install
15:53 czokie also - am having problems creating borrowers or bib entries in 2.3 .... Not sure of the problem yet - will look over the next few days.
15:53 chris hmm interesting, because its the same script that runs in 2.2
15:53 czokie I had a look at the values in the script.
15:53 czokie There is a lot of parameters table data missing from 2.3 version
15:54 czokie To fix, I extracted 2.2 script, ran it, and then re-ran 2.3 script.
15:54 chris weird, i guess something went wrong in the merge from 2.2
15:54 czokie That may be what broke other things ....
15:55 chris if you could right it up in an email that would be great
15:55 chris i wont be working on it this week, but joshua and others might be
15:55 czokie I am thinking of doing a 2.2 db install, updatedatabase 2.2, updatedatabase 2.3 to see if it gets me going, but that can wait for a bit ...(Work work work)
15:56 chris right
15:56 chris i dont think that will work either :)
15:57 chris in fact i would have been hugely suprised if you got 2.3 to go... but knowing the things that broke .. will help us make a 2.3.1 that might actually be installable :-)
02:21 ToinS hi all
02:21 btoumi hi all!!!!
02:22 btoumi chris: are u around
02:31 ToinS salut bruno
02:32 btoumi salut toins ca va?
02:32 ToinS yep
02:32 ToinS ca va
02:32 ToinS et toi ?
02:33 btoumi ca va bien apre sun long week end
02:34 btoumi t'aurais pas vu passer chris ? :=)
02:58 paul hello world
03:11 btoumi HI PAUL
03:11 paul hello bruno
03:12 btoumi lol kool
03:12 paul (but still get no results, even i I should have 7)
03:12 paul (I mean querying with yaz-client give me 7 results, querying through Koha gives me 0
03:13 paul )
03:13 btoumi u search with koha interface?
03:13 paul yep.
03:13 btoumi not on head
03:13 paul nope, on dev_week
03:13 paul it's this one :
03:13 paul http://zoomopac.liblime.com/
03:13 dewey http://zoomopac.liblime.com/ is working now wher do i look
03:13 btoumi ok
03:14 paul (still no internet at my office, still working on my laptop, at ToinS home...)
03:14 btoumi and zebra works with unimarc?
03:14 paul it will work soon : just 1 file to modify ;-)
03:14 paul to 'map' bib1 attset to unimarc fields.
03:15 btoumi ok
03:15 paul the record.abs file says for marc21 :
03:15 paul melm 100author
03:15 paul I will change this to melm 200 author
03:15 btoumi i follow ur discussion with joshua about that
03:15 paul 120 lines to modify
03:15 btoumi ok
03:16 btoumi me zebra works but only on konsole not in koha interface
03:16 paul did you tried dev_week ?
03:17 btoumi now because i work on fines now
03:17 btoumi there is a lot of works
03:18 btoumi when i will finish i work on zebra
03:18 btoumi i try the url that u give me
03:18 paul don't forget my holidays will be august 7 => 23
03:18 btoumi like me
03:19 btoumi i begin the 28 july => 21 august
03:20 btoumi now i wait for chris i hope i see him this morning for my works on fine
03:21 paul did you "beep" him ?
03:21 btoumi no
03:22 paul I did it for you then ;-)
03:22 btoumi thank u
03:22 btoumi :=)
03:29 btoumi paul : do u work with ur pull over ;=)
03:29 paul yes, of course
03:29 paul quite cold in Marseille today. Some snow expected in the next hours
03:29 btoumi lol
03:30 btoumi i just see ice in my glass
03:31 btoumi i think that chris is not here
03:32 btoumi for the moment
04:11 btoumi paul: are u around ?
04:11 paul yep
04:11 btoumi just one question
04:12 btoumi do u now the different possible value for itemlost ?
04:12 paul they can be defined in an authorised_value list. So there is no limit
04:13 btoumi because in fines2.pl
04:14 btoumi u have test on this value and test is like this if itemlost=2 && itemlost=1
04:14 btoumi itemlost ne 2 and itemlost ne 1
04:15 paul my bet is that =0 => item not lost. >=0 => item lost
04:17 btoumi i think like that but i don 't now why u have 2 value ,u must test only >0 not 1 or 2
04:17 btoumi t y for u answer
04:25 chris hi all
04:25 chris sorry bruno i was playing volleyball
04:25 chris paul is right
04:25 chris 1 and 2 are lost
04:25 paul hello chris.
04:25 chris 1 = lost, marked lost by the librarian
04:25 chris 2 = lost, marked lost by koha
04:25 btoumi ok thanks
04:25 paul do you have 5mn to speak of zebra things, or do you need a shower asap ?
04:25 btoumi hi chris  do u win?
04:26 chris at HLT and a few other libraries, after a certain time that a book is overdue, perhaps 4 weeks or 6 weeks
04:26 chris then koha marks it lost=2
04:26 chris bruno: no .. it was 2 - 2
04:26 chris a draw
04:27 btoumi lol
04:27 paul chris : do you have 5mn to speak of zebra things, or do you need a shower asap ?
04:28 paul I'm not sure to understand well the record.abs file syntax.
04:28 paul and the indexdata doc is unclear to me
04:28 chris ive had a shower
04:28 btoumi i take a ticket to speak about fines chris: just after paul ;=)
04:29 chris i can try, im not sure i fully understand it either ... yeah i find their documentation quite hard to read
04:29 paul what means :
04:29 paul melm 200$fauthor,author:p,author:s
04:29 paul author, author:p, author:s ?
04:29 chris ahh i knew this
04:30 paul I also see the !:w
04:30 paul and !:p
04:30 chris they are different types of indexes
04:30 chris !:p is the same as author:p
04:30 chris so you could go
04:31 chris author,!:p,!:s
04:31 chris and that would be the same as above
04:31 paul and what's the difference between nothing, :p :s and !:w ?
04:31 chris im trying to remember :)
04:31 paul s/!:w/:w/
04:32 chris ahh tumer will know
04:35 btoumi chris: can i show with u for fines?
04:36 chris ah ha :w = word
04:37 chris :w = word
04:38 paul2 and :p, :s, nothing ? (if you wrote something, I missed it)
04:38 chris The attributes can be qualified with  field types   to specify which character set should govern the indexing procedure for that field. The same data element may be indexed into several different fields, using different character set definitions. See the the section called “Field Structure and Character Sets â€?. The default field type is w for  word  .
04:38 chris somewhere it must tell us what p and s are .. ill try and find it
04:39 chris f a  Structure   attribute of  Phrase   is used in conjunction with a  Completeness   attribute of  Complete (Sub)field  , the term is matched against the contents of the phrase (long word) register, if one exists for the given  Use   attribute. A phrase register is created for those fields in the .abs file that contains a p-specifier.
04:40 chris p = phrase
04:40 paul2 exactly the kind of paragraph I don't understand ;-)
04:40 chris n = numeric
04:41 tumer hi chris you wanted something?
04:41 chris w = word, p = phrase, n=numeric, u=url
04:41 paul2 hello tumer. it's me that asked a question about record.abs*
04:41 chris i dont know what :s is
04:41 paul2 and nothing means :w isn't it?
04:42 chris yes i think nothing is :w by default
04:42 paul2 so title, title:p, title:s means : index this field as title, as word AND as phrase and as <s> ?
04:42 chris tumer trying to explain
04:42 chris melm 200$f               author,author:p,author:s
04:42 chris to paul
04:42 btoumi chris:can u  beep  me when u are not busy to discuss about fines
04:43 chris will do bruno
04:43 btoumi ty chris
04:43 chris i think maybe :s allows you to sort by that field
04:44 chris yep, im pretty sure thats what it is
04:45 chris ok bruno, lets talk about fines
04:48 tumer can anybody see me there i have aprolem with connection i think
04:48 tumer s/aprolem/ a problem
04:49 paul yep
04:49 chris i can see you tumer
04:49 paul tumer: you're here, but have been disconnected
04:49 paul & reconnected 3 seconds later
04:49 tumer thanks paul
04:51 tumer paul you ok with zebra indexing?
04:51 paul not sure, but i'll try to investigate, chris gave me some hints.
04:51 paul I have another question tumer
04:51 tumer shoot
04:51 paul when I search on yaz-client : "find delmas" I get 7 answers, as expected
04:52 paul but when I search on zoomopac-search, I get 0 answers
04:52 paul any idea why ?
04:52 tumer hmm no
04:52 paul I also tried a search by author :
04:52 paul [Mon Jul 17 10:18:28 2006] [error] [client 127.0.0.1] Boolean Operators:    at /home/paul/koha.dev/dev_we​ek/opac/opac-zoomsearch.pl line 534., referer: http://o19.portable.paulpoulai[…]pac-zoomsearch.pl
04:52 paul [Mon Jul 17 10:18:28 2006] [error] [client 127.0.0.1] Sort by:   at /home/paul/koha.dev/dev_we​ek/opac/opac-zoomsearch.pl line 536., referer: http://o19.portable.paulpoulai[…]pac-zoomsearch.pl
04:52 paul [Mon Jul 17 10:18:28 2006] [error] [client 127.0.0.1] PQF: @attr 1=1003 @attr 4=1 @attr 5=1 "delmas" at /home/paul/koha.dev/dev_we​ek/opac/opac-zoomsearch.pl line 537., referer: http://o19.portable.paulpoulai[…]pac-zoomsearch.pl
04:52 paul [Mon Jul 17 10:18:28 2006] [error] [client 127.0.0.1] Full PQF:        @attr 1=1003 @attr 4=1 @attr 5=1 "delmas" at /home/paul/koha.dev/dev_we​ek/opac/opac-zoomsearch.pl line 538., referer: http://o19.portable.paulpoulai[…]pac-zoomsearch.pl
04:52 paul [Mon Jul 17 10:18:28 2006] [error] [client 127.0.0.1] NICE:   delmas at /home/paul/koha.dev/dev_we​ek/opac/opac-zoomsearch.pl line 539., referer: http://o19.portable.paulpoulai[…]pac-zoomsearch.pl
04:53 paul and got no answer either.
04:53 paul what is very strange too is that
04:53 paul find @attr 1=1003 @attr 4=1 @attr 5=1 "delmas"
04:53 paul in yaz-client works well too
04:53 tumer remind me 4=1 is phrase?
04:55 tumer well i do not use opac-zoomsearch.pl so i do not know abou whether its buggy
04:56 ToinS there is often a #' just before a declaration of sub. Anyone knows what does it mean ?
04:56 paul what do you use ?
04:56 tumer in yaz client if it works then its correct
04:57 tumer also write "help find" in yaz client and it gives you what attributes 4=1 is
04:57 tumer if you are doing a search on phrase you should index as phrase :p
04:57 tumer (:p)
05:00 tumer paul: is use opac-search.pl in dev_week
05:01 tumer or catalogue-home.pl in intranet
05:02 paul no answer with opac(search either :(
05:02 paul 12:02:24-17/07 [session] Connect from tcp:127.0.0.1
05:02 paul 12:02:24-17/07 zebrasrv(14) [session] Starting session 14 from tcp:127.0.0.1 (pid=6504)
05:02 paul 12:02:24-17/07 zebrasrv(14) [log] User perm for perm.anonymous: r
05:02 paul 12:02:24-17/07 zebrasrv(14) [request] Auth none
05:02 paul 12:02:24-17/07 zebrasrv(14) [request] Init OK - ID:81/81 Name:ZOOM-C/YAZ Version:1.78/2.1.22
05:02 paul 12:02:24-17/07 zebrasrv(14) [log] ResultSet '1'
05:02 paul 12:02:24-17/07 zebrasrv(14) [log][app2] zebra_register_open rw = 0 useshadow=0 p=0x8197478,n=,rp=(none)
05:02 paul 12:02:24-17/07 zebrasrv(14) [log] profilePath=.:/usr/local/share/idzebra​/tab/:/home/paul/koha.dev/dev_week/zeb​raplugin/zebradb/biblios/tab/:./tab/ cwd=/home/paul/koha.dev/dev_we​ek/zebraplugin/zebradb/biblios
05:02 paul 12:02:24-17/07 zebrasrv(14) [log] 0 keys, 0 sysnos, rank
05:02 paul 12:02:24-17/07 zebrasrv(14) [log] term="delmas" nn=0 type=void count=0
05:02 paul 12:02:24-17/07 zebrasrv(14) [log] user/system: 0/0
05:02 paul 12:02:24-17/07 zebrasrv(14) [request] Search OK 0 1 1+0 RPN: @attrset Bib-1 delmas
05:02 paul 12:02:25-17/07 zebrasrv(14) [session] Connection closed by client
05:02 paul 12:02:25-17/07 zebrasrv(14) [log][app2] zebra_register_close p=0x8197478
05:02 paul mmm... I suspect a database name problem
05:03 chris good suspicion paul
05:04 paul alleluia !!!
05:04 paul works fine !!
05:04 chris :)
05:05 paul the zebra DB name was biblio and default in my koha.xml
05:05 paul (I had 2, and removed one, but made a mistake)
05:06 paul and UTF-8 works fine too...
05:11 tumer so you can search accented characters and no problem occurs?
05:12 paul but what I say with utf-8 is that my database have moved from 8859-1 to utf-8
05:12 paul nothing related to zebra.
05:12 tumer well search even it does not exist
05:13 tumer the problem ocuring is this:
05:14 tumer the search term after the search displayed on screen as broken utf8
05:16 chris ahh
05:16 chris so it searches ok, but displays what you typed in wrong?
05:17 tumer yes chris it reencodes the term and anything accented on your template gets broken
05:17 chris ahh right
05:18 tumer i  had to re-write my USMARC.pm to correct this
05:19 tumer but i want to know whether anyone else is getting it
05:21 chris ill try to get it up and running here and test it
05:28 paul_lunch good night chris
07:08 paul kados around ?
07:08 paul if yes, a good news : I've made my 1st search successfully on opaczoom-search.pl !!!
07:15 paul tumer[A]: i've a problem with accented chars too : when I search for "probabilité", I should get many answers, but I just get :
07:15 paul Results for Search: biblio.title contains probabilit?
07:15 paul No results found.
07:34 tumer paul the accented character problem is due to new M:F:X
07:34 tumer i sent on devel a new USMARC.pm which corrects this
07:35 tumer the second is your .chr file in tab does not define your accented characters
07:36 tumer you have to define your whole alphabet in sort-string-utf.chr file
07:37 tumer i committed one with some accented characters in but i think kados overwritten that
07:39 tumer ping me if you need help with .chr file
07:50 paul tumer: i'm back
07:50 tumer hi again paul
07:51 paul my sort-string-utf is 1.1.2.1, which is from kados it seems.
07:51 paul it's only 47 lines long
07:51 paul #map éÉe
07:51 paul the # means it's commented, right ?
07:51 tumer yes
07:52 paul so uncommenting it should  be the solution ?
07:52 tumer also the alphabet should contain these charavters
07:52 paul ???
07:52 tumer top line a-z including accented characters
07:53 paul lowercase {0-9}{a-y}üzæäøöå
07:53 paul uppercase {0-9}{A-Y}ÜZÆÄØÖÅ
07:53 paul those ones ?
07:53 tumer those one should contain é etc
07:54 tumer in their normal sorting order
07:54 tumer according to french
07:54 tumer mine has lowercase {0-9}aâbcçdefgğhıiîjklmnoöpqrsştuüvwxyz
07:56 tumer also if you need to find probabilité when you search probabilite ..
07:56 tumer you should add "equivalent eé"
07:58 tumer paul:do you follow?
07:58 paul yep, many thanks
07:58 paul what is a shame is not to be able to have the name of this file as parameter.
07:58 paul (same for record.abs)
07:59 paul because i'll have to commit "sort-string-utf_for_french.chr"
07:59 paul and during install, we will have to remane it
07:59 tumer you can change the name of this file but its still valif for all
07:59 tumer s/valif/valid
08:00 tumer not ebing able to rename record.abs is already causing problems with me
08:00 tumer i am investigating other indexing filters in zebra currently
08:02 tumer we should probably have a more global with most european accents in
08:03 tumer but whatever we do KOHA will still need language adjustments at installation
08:03 paul tumer :
08:03 paul map (^The\s)    @
08:04 tumer that is a stop word
08:04 paul means "the" at beginning of a string must be eliminated for ordering isn't it ?
08:04 tumer correct
08:04 tumer all stop word has to go in here
08:04 tumer another localisation to be done
08:08 tumer "sort-string-utf_for_french.chr"
08:10 tumer paul:to make zebra read from  "sort-string-utf_for_french.chr" go into defaul.idx
08:10 tumer you will see that its related in there
08:10 tumer "sort-string-utf_for_french.chr"
08:10 tumer "sort-string-utf_for_french.chr"
08:11 tumer changing  "sort-string-utf.chr" to  "sort-string-utf_for_french.chr" will also work
08:13 tumer so you may have french.idx which you rename to default.idx at installation or rename sort-...chr
08:25 paul tumer ?
08:25 dewey somebody said tumer was here for a few seconds ;-)
08:25 paul it seems not to work atm.
08:25 paul 2 syntax questions :
08:25 tumer say again paul
08:25 paul is
08:25 paul equivalent eé
08:25 paul correct ?
08:25 tumer yes
08:25 paul map éèêëe
08:26 paul map ËÊe
08:26 paul ?
08:26 tumer i use equivalent map only for stop words
08:27 paul so equivalent eé should be enough ?
08:28 tumer use equivalent for lowercase and uppercase separately
08:28 paul so :
08:28 paul map éèêëe
08:28 paul map ËÊE
08:28 paul ?
08:29 tumer equivalent eé equivalent EE(accented) (i dont have it on keyboard)
08:30 tumer you need reindexing after changing this file
08:30 paul yep, I knew for reindexing
08:30 tumer mapping also supposed to work
08:31 tumer i could not get it to work for character mapping but for stop words
08:31 paul you mean you can't search on accented char ?
08:32 tumer i mean i can search on accented characters by using equivalent and not Map
08:33 paul strange syntax :
08:33 paul equivalent å(aa)
08:33 paul but equivalent eé
08:33 paul ?
08:33 paul are u sure it's not equivalent ée ?
08:33 tumer they are the same eé or ée
08:35 paul wow ...
08:35 paul [paul@portable zebraplugin]$ zebraidx -g iso2709 -d biblios -c /home/paul/koha.dev/dev_week/ze​braplugin/etc/zebra-biblios.cfg update export/
08:35 paul 15:35:31-17/07 zebraidx(7552) [log] update grs.marcxml.record export/test.iso2709 4044
08:35 paul zebraidx: extract.c:1305: extract_flushRecordKeys: Assertion `ch > 0' failed.
08:35 paul Abandon
08:35 paul I didn't had this before !!!
08:36 tumer i never had it either
08:38 paul ok, zebraidx init solves it
09:08 tumer[A] power cutoff UPS running out so I am off
10:45 paul2 kados around ?
10:46 paul2 bye bye world. see you on thursday

| Channels | #koha index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary