![http://le-gall.net/pierrick/im[…]koha-2.2-utf8.png](http://le-gall.net/pierrick/images/koha-2.2-utf8.png){kind=link}
← Previous day | Today | Next day → | Search | Index
All times shown according to UTC.
| Time | Nick | Message |
|---|---|---|
| 11:02 | kados | hi guys |
| 11:02 | I made a few very important commits to rel_2_2 hdl | |
| 11:02 | related to utf-8 | |
| 11:02 | hdl | which ones ? |
| 11:03 | kados | hdl: I made a minor change to addbiblio.pl and to bulkmarcimport |
| 11:04 | hdl: that encodes all MARC-8 data as UTF-8 before importing / displaying on the screen | |
| 11:05 | hdl: but I'm not sure if it will work with UNIMARC records | |
| 11:08 | hdl | pierrick_ : still, I have a problem. Maybe it is me. |
| 11:09 | But henridamienlaurent.dyndns.org:8085/cgi-bin/koha/admin/biblio_framework.pl | |
| 11:10 | When Périodiques is not displayed correctly. | |
| 11:10 | But When I go into phpmyadmin, it is ok. | |
| 11:13 | pierrick_ | hdl: I can't reach your servere :-/ |
| 11:14 | hdl | Enterprise firewall ? |
| 11:14 | pierrick_ | hdl: are you sure your data are truely UTF-8 ? |
| 11:14 | maybe firewall, you're right | |
| 11:14 | hdl | Yes. |
| 11:15 | pierrick_ | how can you be sure ? |
| 11:15 | hdl | In an utf-8 console, |
| 11:15 | on a mysql connection | |
| 11:15 | select * from biblio_framework; | |
| 11:15 | returns weird character : | |
| 11:16 | pierrick_ | did you "set names 'UTF8';" before "select" ? |
| 11:16 | hdl | | PER | P?iodiques | |
| 11:16 | then set names utf-8 | |
| 11:16 | no weird char and everything goes well : | |
| 11:17 | | PER | Périodiques | | |
| 11:17 | pierrick_ | OK, I agree your data are truely UTF-8 |
| 11:18 | hdl | Then is there a connection status in Koha that goes wrong, that's what I want to know. |
| 11:18 | pierrick_ | Can you adapt the script I send today on koha-devel to make it read biblioframework and write the result to a file ? |
| 11:27 | hdl | http://pastebin.com/594457 : No problem. |
| 11:27 | Utf-8 and utf8. | |
| 11:28 | I hate hyphens :( | |
| 11:28 | no. | |
| 11:29 | It was good. | |
| 11:31 | pierrick_ | so the problem is in communication between Koha, Apache and browser |
| 11:33 | hdl | or between Koha and Mysql.... |
| 11:33 | pierrick_ | Koha doesn't do anything special I think |
| 11:34 | You asked Koha to 'set names 'UTF8'" ? | |
| 11:34 | hdl | maybe set names is not done on each call. |
| 11:34 | I tried to. | |
| 11:34 | It was commented out in Context.pm | |
| 11:35 | pierrick_ | to make sure the problem doesn't comes from Koha+MySQL, just after the select, print the results to a file |
| 11:40 | hdl | Wide character un print. |
| 11:41 | Seems perl considers STDOUT not to be utf-8 | |
| 11:45 | pierrick_ | th same query in Koha doesn't give the same result as a simple Perl script ! |
| 11:50 | hdl | no simple perl script wil return you the Authetication page, unless Auth is desactivated |
| 11:52 | pierrick_ | I also have "Périodiques" in the biblioframework table Paul gave me |
| 11:52 | hdl | And does it work for you . |
| 11:52 | ? | |
| 11:53 | pierrick_ | if I "set names 'UTF8'" in mysql client and "select * from biblio_framework", the display is correct in my console... |
| 11:53 | ... even if the table says the the field is latin1. | |
| 11:54 | what is the charset of your field and your table ? | |
| 11:54 | hdl | Utf8. I converted everything. |
| 11:54 | How can I print mysql status in Perl ? | |
| 11:55 | pierrick_ | I don't know yet, I search |
| 12:00 | osmoze | hi all |
| 12:00 | pierrick_ | I don't find, but as the behaviour of readfile_insertdb.pl is different with and without "set names 'UTF8'", I'm sure it's used |
| 12:00 | Hi osmoze | |
| 12:25 | in fact, I don't understand very well were the information of the column/table charset in MySQL is used. As I see on my Koha 2.2 installation, my data are in UTF-8 while my column says it's in latin1... | |
| 12:54 | kados | pierrick_: have you tried switching to utf-8 in mysql? |
| 12:56 | pierrick_ | kados: what do you mean exactly ? |
| 12:57 | switching server, db, table, column, connection ? | |
| 12:57 | kados | table |
| 12:58 | pierrick_ | (how do you notify IRC things like "* doing the test" ?) |
| 12:59 | kados | dunno |
| 12:59 | pierrick_ | you and others do it all the time |
| 13:01 | kados | you mean: |
| 13:01 | ? | |
| 13:01 | type: | |
| 13:01 | /me does the test | |
| 13:03 | pierrick_ | roger roger, alpha tango charly |
| 13:08 | kados: "alter table biblio_framework modify frameworktext char(255) character set utf8 NOT NULL default '';" doesn't change anything. Data doesn't seem to have been modified (IMO it was already UTF-8 anyway), the only important thing is to "set names 'UTF8';" | |
| 13:09 | kados | pierrick_: first I think you need to convert it to binary |
| 13:10 | pierrick_ | kados: I don't think I have any problem with my data, I suppose them UTF-8 |
| 13:11 | kados | right, but mysql doesn't know that |
| 13:11 | so it will try to encode it | |
| 13:11 | pierrick_ | with my small script, I can extract them to a file, all in UTF-8* |
| 13:11 | kados | as utf-8 ... again |
| 13:11 | and it will probably mutilate the characters outside the ascii range | |
| 13:11 | (I could be wrong of course) | |
| 13:12 | pierrick_ | kados: in fact, I don't understand where you want me to go :-/ |
| 13:13 | kados | I'm trying to find the documentation :-) |
| 13:13 | pierrick_ | http://dev.mysql.com/doc/refma[…]t-conversion.html |
| 13:13 | kados | yep, I think that's it |
| 13:14 | it recommends going to binary first | |
| 13:14 | then to utf-8 | |
| 13:14 | pierrick_ | yes, if your data are not already UTF8 |
| 13:15 | let's do a try :-) | |
| 13:20 | intermediate step : biblio_framework.frameworktext is binary(255), I do not need "set names 'UTF8';" to see accentued characters in my UTF-8 console | |
| 13:22 | operation finished, biblio_framework.frameworktext is "char(255) character set utf8", if I don't specify "set names 'UTF8';" before the select, the accentued characters is not correctly displayed in my UTF-8 console | |
| 13:23 | if I "set names 'UTF8';", accentued characters are displayed correctly in my UTF-8 console (nothing changed from the initial step) | |
| 13:28 | kados | cool |
| 13:28 | so what do they look like in the opac? | |
| 13:34 | pierrick_ | where can I display content of biblio_framework ? |
| 13:38 | kados | in 'admin'-> biblio framework |
| 13:40 | pierrick_ | kados: there is a bug in http://www.kohadocs.org/Updating_Koha.html on the symlink for opac templates |
| 13:40 | I corrected by myself, but it could be good to modify the documentation :-) | |
| 13:41 | kados | pierrick_: could you email stephen with the correction? |
| 13:41 | pierrick_ | kados: right |
| 13:42 | in my 2.2 OPAC, I don't see 'admin' :-/ | |
| 13:44 | I found it in the intranet | |
| 13:46 | kados | marc framework isn't available from the opac |
| 13:46 | but really, the biblio_framework is probably a poor test | |
| 13:46 | since it doesn't hold any actual MARC data | |
| 13:47 | just information about what MARC tags/subfields the database recognizes | |
| 13:47 | pierrick_ | Firefox displays as iso-8859-1 as it's what precised in HTML header, and my frameworktext is correctly displayed. |
| 13:48 | if I "set names 'latin1';" and set my console charset to latin1, accentued characters are correctly displayed. | |
| 13:48 | kados | hmmm ... |
| 13:49 | pierrick_ | making the conversion from "*anything*" to utf-8 is possible, but the opposite is not available very often |
| 13:49 | for example, iconv doesn't propose to convert from utf-8 | |
| 13:49 | so my data are maybe in latin1 | |
| 13:49 | or not | |
| 13:49 | or I don't know | |
| 13:50 | kados | hmmm |
| 13:55 | hdl | kados, pierrick_ : with set names=latin1, with my data converted to utf8, the display is "magically" correct !!!!!! |
| 13:55 | STRANGE. | |
| 13:57 | pierrick_ | I made another test : I've inserted a character really not available in latin1 on the description of my biblio fwk. |
| 13:57 | With Firefox, I ask to display in UTF-8, and my characters are not displayed correctly | |
| 13:59 | still in my 2.2 working copy, I've added "set names 'UTF8'" just after dbh creation and my characters are displayed correctly :-) | |
| 14:01 | hdl | Confusing confusing. |
| 14:02 | What is in utf8 what is not.... | |
| 14:02 | Is Périodiques really in utf8 (biblioframework) is it not ? | |
| 14:03 | I DID ALTER Table convert character set utf8 collation utf8_general_ci. | |
| 14:03 | But is it true utf8 ? | |
| 14:04 | On other parts of Koha, display is disturbed. | |
| 14:05 | pierrick_ | http://le-gall.net/pierrick/im[…]koha-2.2-utf8.png |
| 14:08 | hdl | pierrick_: what is your keyboard layout ? |
| 14:08 | Are you workin on a 2.2. version ? | |
| 14:09 | What are your collation and charset ? | |
| 14:09 | pierrick_ | how do I know my keyboard layout ? |
| 14:09 | 2.2, yes | |
| 14:09 | hdl | By me, in /etc/sysconfig/keyboard : |
| 14:09 | pierrick_ | charset utf8 for this table/column |
| 14:10 | default collation for this charset (but it has nothing to do with our problem) | |
| 14:10 | hdl | KBCHARSET=UTF-8 |
| 14:10 | KEYBOARD=fr | |
| 14:10 | KEYTABLE=fr_FR-latin1 | |
| 14:10 | DISABLE_WINDOWS_KEY=no | |
| 14:10 | ||
| 14:10 | pierrick_ | I don't have this file |
| 14:10 | hdl | so with export, you should get your variables. |
| 14:11 | No. I donot have them. | |
| 14:12 | pierrick_ | I don't have them neither of course |
| 14:12 | But what is the importance of my keyboard layout ? I copied/paste these characters from the web | |
| 14:12 | on an UTF8 page | |
| 14:13 | hdl | ok. |
| 14:14 | pierrick_ | I'm leaving now, have a nice WE you all :-) |
| 15:44 | Flash | Anyone out there willing to show me the ropes? |
| 15:46 | I have a real basic question. Don't be afraid. | |
| 16:04 | kados | hi Flash |
| 16:04 | what's your question? | |
| 16:05 | Flash | How can I see a list of my books and who has them? |
| 16:06 | kados | you might try going to the 'reports' section |
| 16:06 | Flash | Which report |
| 16:08 | kados | well there are several you might want to try |
| 16:08 | the default reports that come with a stock Koha might not include all the specific reports your library needs | |
| 16:09 | typically you could get your vendor to help you generate the specific report you need | |
| 16:09 | Flash | I tried Inventory/Stocktaking but it didn't work. |
| 16:09 | vendor? | |
| 16:09 | kados | yea, that's the problem with the default reports -- not all of them work |
| 16:10 | you don't have a Koha vendor? :-) | |
| 16:10 | Flash | No, I downloaded and istalled it myself. I'm an IT guy. |
| 16:10 | kados | http://koha.org/support/pay.html |
| 16:10 | ahh | |
| 16:10 | in that case, you could probably build yourself a report that you need | |
| 16:10 | the reports are quite simple scripts | |
| 16:10 | they just query the mysql database | |
| 16:11 | Flash | Perl |
| 16:11 | kados | yep |
| 16:11 | though you could use php or python if you're more familiar with them | |
| 16:11 | Flash | Is there a repository of reports that people have created? |
| 16:11 | kados | or you could just query mysql directly |
| 16:12 | well ... that's the default reports :-) | |
| 16:12 | Flash | Is there an ERD of the database? |
| 16:12 | kados | there are a lot of libraries out there using Koha that don't contribute their modifs back to the main repo |
| 16:12 | (well, all of them anyway) | |
| 16:12 | yea ... I think there's on on kohadocs.org | |
| 16:13 | Flash | Ok, I guess I'll either have to get the default reports working or write my own. |
| 16:14 | kados | sweet |
| 16:14 | Flash | Caio. |
| 16:14 | kados | bye |
| 17:29 | if I'm reading correctly, Tumer has got Zebra running in a production environment | |
| 17:30 | chris: did you get that from his email too? | |
| 00:19 | Heavenlydemo | hey everyone just have a question. wondering if koha would be of use for me |
| 00:20 | i want a database for my dvds, and a circulation so would koha be able to do that on my laptop? |
← Previous day | Today | Next day → | Search | Index