- Koha-devel - lists.koha-community.org

ALTER IGNORE new policy?
by Philippe Blouin 31 Jul '18

31 Jul '18

Good day, everyone! Small inquiry here: Simply put, bz 17234 made an effort to remove the ALTER IGNORE from updatedatabase.pl so that it could work with Mysql 5.7. But playing with the master today, I got many errors related to it * 12395 - 18.06.00.005 * 19524 - 18.06.00.006 * 13560 - 18.06.00.008 Was there a change of policy regarding that, or is this something a bug should be opened or reopened about? Right now it kills our community development effort. Later on, it'll be our customers maintenance and I'll start to plan right now. Thanks! -- Philippe Blouin, Responsable du développement informatique Tél. : (888) 604-2627 philippe.blouin(a)inLibro.com <mailto:philippe.blouin@inLibro.com> inLibro | pour esprit libre | www.inLibro.com <http://www.inLibro.com>

3 4

Good commit messages
by David Cook 30 Jul '18

30 Jul '18

Hi all, So Jonathan linked me to this the other day: <https://wiki.koha-community.org/wiki/Commit_messages#Good_commit_messages_2 > https://wiki.koha-community.org/wiki/Commit_messages#Good_commit_messages_2 I don't understand one of the "Bad commit messages" though. I thought everyone always starts a good commit message with "Bug 12345 - ". Am I missing something there? Personally, I don't think you can always explain what the *bug* is in 50 characters, although I'm super on board for the commit message containing what the *fix* does. I'm writing a commit message for https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=21131 and struggling a bit as the bug is actually kind of convoluted. David Cook Systems Librarian Prosentient Systems 72/330 Wattle St Ultimo, NSW 2007 Australia Office: 02 9212 0899 Direct: 02 8005 0595

2 1

Formatting control number searches
by Barton Chittenden 28 Jul '18

28 Jul '18

I'm working on generating links for 7XX linking fields in XSLT, a-la Bug 15140 - Add MARC21 776 to OPAC and staff display ( https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=15140) This Takes the 773$w, and uses the XSLT template 'extractControlNumber' which strips the leading MARCOrgCode, e.g. turning (OCoLC)2776588 into 2776588 ... and then the search URL is /cgi-bin/koha/opac-search.pl?q=Control-number:2776588 The problem is that 2776588 isn't the actual OCLC number -- The formatting rules are here: https://help.oclc.org/Metadata_Services/OCLC-MARC_records/Content_designato… In short: 0-8 digits: 0 pad to 8 digits, prepend 'ocm': 2776588 -> ocm02776588 9 digits, prepend 'ocn' > 9 digits, prepend 'on' I tried formatting the search with a leading '*': /cgi-bin/koha/opac-search.pl?q=Control-number:*2776588 as well as with leading and trailing '*' /cgi-bin/koha/opac-search.pl?q=Control-number:*2776588* Neither of these worked. Formatting the control number according to the OCLC rules above did work: /cgi-bin/koha/opac-search.pl?q=Control-number:ocm02776588 So I have three questions: 1) Is there a way to format a Control-number search that doesn't force me to generate the control number according to the OCLC rules above? 2) If not, does anyone have pointers about how to * Test for (OCoLC) -> * Test length of control number * < 9 -> 0 pad to 8 digits, prepend 'ocm' * 9 -> prepend 'ocn' * > 9 -> prepen 'on' 3) Are there other control number formatting rules for other MARCOrgCodes?

2 4

Koha Debian packages for 18.05.02 and 17.11.08 are up
by Mirko Tietgen 26 Jul '18

26 Jul '18

Hi, it's hot, no further text, ;) Cheers, Mirko -- Mirko Tietgen mirko(a)abunchofthings.net http://koha.abunchofthings.net

1 0

Koha 17.11.08 release
by Fridolin SOMERS 26 Jul '18

26 Jul '18

The Koha community is proud to announce the release of Koha 17.11.08. This is a maintenance release. The full release notes are available at https://koha-community.org/koha-17-11-08-release/ Debian packages will be upgraded in a few days. Regards, -- Fridolin SOMERS <fridolin.somers(a)biblibre.com> BibLibre, France - software and system maintainer

3 2

Koha reports wiki
by Cab Vinton 23 Jul '18

23 Jul '18

Hi, All -- Apologies for the cross-posting. The clean up of the Reports wiki page appears to have got stalled out: https://wiki.koha-community.org/wiki/SQL_Reports_Library (Large no. of duplicate listings of reports for holds, patrons, & circulation (HPC), which also appear on corresponding pages -- https://wiki.koha-community.org/wiki/SQL_Reports_Holds, etc.) I've identified the reports on the main page that aren't cross-listed (19 in total). I'd like to relocate these & then delete all the listings under the HPC categories on the main page. I suspect these pages see a lot of use, however, so I don't want to barge in and inadvertently make things worse. Anything I should be aware of before jumping in? All best, Cab Vinton, Director Plaistow Public Library Plaistow, NH

4 10

Client-side git hooks
by David Cook 21 Jul '18

21 Jul '18

Hi all, How many of you are using client-side git hooks? That is, git hooks that are in your working directory. A while ago, I added a "pre-commit" git hook in my Koha git working directory, which runs a little Python script every time I make a commit. It's nothing fancy. Basically it gets a list of all the files I'm changing in Git and for a ".pl" or a ".pm" file, it runs "perl -c", which compiles the Perl code without running it (with a caveat about BEGIN{} blocks which do get run at compile time so could be a problem if you're running untrusted code especially as a privileged user.). The Python script gets the exit code of that operation, and if there's an error, it uses an exit code of 1 itself and prevents the commit from happening. I've attached some sample code to the bottom of this email under "__PYTHON_SCRIPT__". I find that it's a nice way of catching errors. Maybe you write some code and you don't think you need to test it, or you tested it but you made a last minute change and that last minute change has a typo. this catches that kind of things - at least if it's an error which prevents successful compilation. Anyway, I was just porting a patch between different versions of Koha, and everything looked good at a glance and the code merge was successful, but the commit failed because one variable name in one line of the many lines changed was slightly different. The error messages told me exactly where to go and then it was obvious what the problem was and what to do. Anyway, I just thought I'd share that. Maybe everyone is already using client-side git hooks, but I thought I'd share just in case someone else finds it useful. Especially as it saved my bacon just now. David Cook Systems Librarian Prosentient Systems 72/330 Wattle St Ultimo, NSW 2007 Australia Office: 02 9212 0899 Direct: 02 8005 0595 __PYTHON_SCRIPT__ #!/usr/bin/env python import subprocess, os errors = 0 output = subprocess.check_output(["git","diff","--cached","--name-only","--diff-filte r=ACMR"]) lines = output.split('\n') for line in lines: if line: filename, file_extension = os.path.splitext(line) print(filename) print(file_extension) if file_extension == ".pl" or file_extension == ".pm": rv = subprocess.call(["perl","-c",line]) if rv: errors += 1 if errors > 0: exit(1)

5 4

Asynchronous standup
by Nick Clemens 20 Jul '18

20 Jul '18

Shib for a few sites Working on rancor to get dgi stuff rebased Pushing to master Education hour on authentication Took off to help build the pit for our friends annual pig roast, pics later -- Nick Clemens Sonic Screwdriver (Development Support) ByWater Solutions IRC: kidclamp

1 0

Update your misc4dev repo
by Jonathan Druart 20 Jul '18

20 Jul '18

Hi devs, Since bug 20287 there is not more AddMember subroutine and misc4dev was using it. You will need to update it (the repo from gitlab/koha-community), especially if you are using a devbox. Cheers, Jonathan

2 4

Easy sign off
by David Cook 20 Jul '18

20 Jul '18

Hi all, If anyone is looking for an easy sign off, consider checking out https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=21097. It has a bit of set up in the test plan in terms of the acquisitions interface, but the actual patch is just adding a closing tag for an optgroup element. Modern browsers automatically add the missing tag, but I just noticed it was missing by accident and thought I'd fix it. David Cook Systems Librarian Prosentient Systems 72/330 Wattle St Ultimo, NSW 2007 Australia Office: 02 9212 0899 Direct: 02 8005 0595

1 0