This patch adds a script to allow checking the status of Koha's Perl dependencies from the
command line interface at any time. Run '$ ./koha_perl_deps.pl --help' for available options.
'$ ./koha_perl_deps.pl -c' is recommended if you like color.
This patch adds two modules: C4::Installer::PerlModule and C4::Installer::PerlDependencies. The latter provides a single point
of reference for cataloging Koha Perl dependencies. The former provides an OO style interface to the dependency catalog.
The format of C4::Installer::PerlDependencies is very simply an anonymous hash of hashes. Each second level hash takes this form:
Galen Charlton [Fri, 14 May 2010 20:10:52 +0000 (16:10 -0400)]
bug 4509: remove references to PINESISBN system preference
This system preference never worked; also, as far as I can tell,
the PINES oISBN service was experimental, not meant for production
use, and is not currently advertised.
Colin Campbell [Fri, 14 May 2010 16:24:13 +0000 (17:24 +0100)]
Bug 4507 Don't return null vendor to claims processing
The empty vendor created by the select if any subscriptions
have no associated vendor generates a 'wrong-looking' display
and selecting it gives no options
Colin Campbell [Sat, 15 May 2010 12:40:18 +0000 (13:40 +0100)]
Bug 2505 enable warnings in serials/reorder_members.pl
The only serials routine without warnings enabled was a
script with the least code to generate them
removed the numerous Module use declarations that were not
required
aligned the parameters for a more aesthetically pleasing effect
Jane Wagner [Tue, 4 May 2010 12:15:55 +0000 (08:15 -0400)]
Bug 4394 REVISED Allow opaccolorstylesheet syspref to use an external URL
Previous code would only allow for an auxiliary stylesheet on the same server. This checks for
http in the opaccolorstylesheet syspref. If it's found, sets a different variable pointing to the
remote server location. If not found, retains the previous css directory path.
Also modifies the description of opaccolorstylesheet in the installer files to reflect current
usage as an auxiliary stylesheet.
Signed-off-by: Chris Cormack <chrisc@catalyst.net.nz>
Bug 4141 Reconcile 3.0.x and HEAD database updates for 3.2.0
This is a partial fix addressing this error:
You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '' at line 1 at /home/koha/koha.prod/installer/data/mysql/updatedatabase.pl line 2916., referer: http://koha-staff.foobar.org/cgi-bin/koha/installer/install.pl
If there are no records in aqbudgets the result is an empty array which causes the sql to attempt to set AUTOINCREMENT to null.
This patch is being submitted on behalf of the Mapua Community Library.
This patch updates the URL for image links in ContentCafe
Sponsored by Plano Independent School District, Plano, TX, USA
[RM note: the previous book cover URL was for a varient of Baker
and Taylor's service that has to do with 'bookstore integration',
and if I remember correctly, the main difference is that the
book jacket image had "buy now" text. The version supplied
in this patch should work for normal use of ContentCafe.]
Colin Campbell [Wed, 12 May 2010 11:25:41 +0000 (12:25 +0100)]
Housekeeping in errors scripts
Enable warnings - remove Bug 2505 FIXME messages
Correct FSF address in License statement
To aid code analysis tools:
Call new unambiguously with arrow syntax
use noninterpolating quotes on noninterpolating strings
Obviously, Koha global pagination function could be improved. This patch
do something which belongs to this function. A CPAN module like
Data::Page could help or be an inspiration... 3.4?
There is also room for improvement in they way resultset are
constructed. The whole set is retrieved and just a subset is displayed.
It impacts performances for libraries with a great number of patrons.
The right solution would be:
- to find the resultset size with SELECT COUNT(*)
- to retrieved the exact subset with LIMIT and OFFSET SQL arguments
Colin Campbell [Sun, 9 May 2010 14:58:52 +0000 (15:58 +0100)]
Do not declare variables within (bogus) conditionals
if $index is unnecessary as we have made this true 5 lines above
variables should not be declared in conditionals if used outside of them
set $struct_attr to a sensible default to avoid generating warnings
in this assigment and elsewhere
Jane Wagner [Wed, 11 Nov 2009 15:47:07 +0000 (10:47 -0500)]
Bug 3093 add syspref to turn off multiholds button
Having the place multiple holds button at the top of the OPAC and staff search results has caused problems for some sites with tight holds policies; this syspref allows those sites to turn off the button.
Jane Wagner [Wed, 7 Oct 2009 13:52:02 +0000 (09:52 -0400)]
Bug 3671 Workaround for font display problem
If there is no 007 field, parts of the OPAC results data display in a
much smaller font than normal. Adding a blank space in the span section
fixes it.
(cherry picked from commit 2ff55ef2a2443933881186323f5dec3fe40842dd)
Colin Campbell [Thu, 6 May 2010 17:13:36 +0000 (13:13 -0400)]
Bug 4450 Use more consistent error returns in C4/Creators/*
A lot of routines were defaulting to return -1 in error conditions
but calling code was expecting a ref or object
use return with explicit undef (or emptyness in array context)
for these cases. Extended this to cases where return was not tested
( -1 might in some cases be legit data).
Signed-off-by: Chris Nighswonger <cnighswonger@foundations.edu> Signed-off-by: Galen Charlton <gmcharlt@gmail.com>
(cherry picked from commit 5cf2b78b6ffa401bc82b3f913c89828c38b8e4ed)
[RM note: ... thereby undoing the revert]
Galen Charlton [Thu, 6 May 2010 23:46:51 +0000 (19:46 -0400)]
move misc source file test and fix scripts to xt/
The license and non-UTF8 file checks are closer to
author tests and don't really belong at the top level of the
Koha source tree. Follow-up might be to converted them
to test cases in xt/author that can be run by prove(1).
Colin Campbell [Thu, 6 May 2010 17:13:36 +0000 (13:13 -0400)]
Bug 4450 Use more consistent error returns in C4/Creators/*
A lot of routines were defaulting to return -1 in error conditions
but calling code was expecting a ref or object
use return with explicit undef (or emptyness in array context)
for these cases. Extended this to cases where return was not tested
( -1 might in some cases be legit data).
Signed-off-by: Chris Nighswonger <cnighswonger@foundations.edu> Signed-off-by: Galen Charlton <gmcharlt@gmail.com>
Owen Leonard [Wed, 5 May 2010 14:50:38 +0000 (10:50 -0400)]
Fix for Bug 4416, renew all and return all buttons too close together
This patch removes the "return all" button altogether. Returning all items
is a step which one wouldn't take lightly or even regularly. Since the
interface has a means of selecting all checkboxes in the "Check in" column,
removing the "Return all" button wouldn't eliminate the means of doing so, it
would just require more deliberation.
Owen Leonard [Wed, 5 May 2010 14:15:49 +0000 (10:15 -0400)]
Fix for Bug 2375, Serials holdings data does not display in opac-detail
- Removing markup to display serial data in the holdings table
This functionality (if it worked) would seem to duplicate
the data displayed under the 'Subscriptions' tab
- Revised again to merge item status display changes
Kyle M Hall [Tue, 4 May 2010 16:33:46 +0000 (16:33 +0000)]
Fixed Patron Search Results
By default, the patron search was pulling results in *Descending* order,
then reordering the results on the page using javascript in *Ascending* order.
This was causing screwy results where the entire list of results of multiple
pages was not in complete alphabetical order.
Fixed by changing the SQL to pull in Ascending order.
Owen Leonard [Thu, 25 Feb 2010 14:43:15 +0000 (09:43 -0500)]
Use CSS3 box-shadow property for Cart tooltip
The OPAC Cart tooltip ("Your cart is empty," "1 item(s) added to your
Cart") currently uses a jQuery plugin to generate a drop shadow.
This works cross-browser, but has a high failure rate with regard
to hiding the shadow consistently, resulting in an orphan shadow
after the tooltip has been hidden. I propose to instead use the
new CSS3 property "box-shadow" (-moz-box-shadow and -webkit-box-
shadow for FireFox and Safari/Chrome). Currently Internet Explorer
does not support this property. I think this is an
acceptable tradeoff for more stable functionality.
Colin Campbell [Fri, 26 Mar 2010 16:41:28 +0000 (16:41 +0000)]
Bug 2889 Remove toggle in boraccount display
Also remove the reverse column if no contents to be displayed
( otherwise looks like a markup error )
Rationalize the main loop now toggle not required
Michael Hafen [Tue, 30 Mar 2010 20:58:50 +0000 (14:58 -0600)]
on checkout ask for confirmation if borrower from different branch
With IndependantBranches on check that the borrower is from the same branch as
the librarian. If not then ask for confirmation. This check is skipped for
SuperLibrarians.
Galen Charlton [Sat, 1 May 2010 16:43:07 +0000 (12:43 -0400)]
bug 4036, 4405: overdue and fine day block improvements
* Adopted wording suggested by Kyle Hall for the
USERBLOCKEDREMAINING and USERBLOCKEDOVERDUE circulation blocks
* Updated IsMemberBlocked so that if a patron has accrued
fine days, that will be tested for first before testing
to see if the patron has current overdue items; this solves
a problem introduced in the patch series for bug 4505 where
accrued fine days would be ignored if (a) the patron has
current overdue items but (b) the library has chosen to
set the OverduesBlockCirc syspref to noblock.
* Now correctly assigns the USERBLOCKEDREMAINING and USERBLOCKEDOVERDUE
blocks; prior to this patch, they had been swapped.
FIXME: IsMemberBlock ought to be split into two functions
Galen Charlton [Sat, 1 May 2010 18:32:44 +0000 (14:32 -0400)]
bug 4396: remove reerences to bi_notforloan
These were causing the OPAC details place hold button to disappear
in circumstances where a bib-level item type was set to a non-loanable
item type even when item-level_itypes was on.
This change does mean that if item-level_itypes is not turned on,
the hold button will now appear for bibs whose bib-level item type
has its not-for-loan flag set; however, if this is a problem, a
better long-term fix would be to use CanBookBeReserved() instead.