The aqorders.subscriptionid info is not sent to the template when the
order is edited. Which means we lose this link.
Test plan:
Create an order from a subscription
Edit the order
=> Without this patch, the aqorders.subscriptionid value is set to NULL
and items are created when receiving serial.
=> With this patch applied the link is preserved and expected behaviors
are preserved during all the acquisition workflow
You should also try and create several orders from the same subscription
Sponsored-by: BULAC - http://www.bulac.fr/
Signed-off-by: Séverine QUEUNE <severine.queune@bulac.fr>
Signed-off-by: Julian Maurice <julian.maurice@biblibre.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 4df24fdde1)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Moving the status to the invalid 'processing' might well have unwanted
side-effects. (The status column will be set to empty string and we have
a problem if it is not processed.)
This patch allows pass-through of DBIX's make_column_dirty in
Koha::Object (simple tests included) and uses it to force an update.
If the update does not return true, it still exits.
Test plan:
[1] Read the changes.
[2] Run t/db_dependent/Koha/Object.t
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Conflicts:
Koha/Object.pm
(cherry picked from commit 40e865ba16)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Last week, we had a database server whose disk filled, causing database writes to fail.
This meant that messages in message_queue marked 'pending' were not marked as 'sent' when they were added to the postfix mail queue; messages were sent every 15 minutes (as specified in the cron job) until the disk space issues were cleared.
I would suggest adding a token write to the start of process_message_queue.pl as a 'canary in the coal mine'. If the database write fails, process_message_queue should stop, because it's not safe to proceed sending emails that may not be marked 'sent'.
Test Plan:
1) Apply this patch
2) Make the message_queue table unwriteable somehow
3) Run process_message_queue.pl
4) Script should die with error
Signed-off-by: Mark Tompsett <mtompset@hotmail.com>
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit a1e044094b)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
If items.restricted == 1, CanBookBeIssued will not returned what we are
testing.
The easiest and global fix is to define a default value at TestBuilder
package level
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit e93241a717)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
The tags for home library and holding library in the 'refine your
search' list should be h5.
This patch changes the html tag for the availability, holding library and home library
facets.
To test:
- Search for an item in the OPAC where the relevant facet appears
- Use 'Inspect element' to check the tag
- Apply patch
- Confirm the facet now has the h5 tag
- Confirm spacing is improved
Signed-off-by: Mark Tompsett <mtompset@hotmail.com>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 8d5e9750af)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This is caused by bib_list not being deleted.
The path was not explicitly set.
This patch adds path=/ explicitly to ensure it
does get deleted.
TEST PLAN
---------
1) search for 'the' in OPAC
2) add several items to cart.
3) look at your cart.
4) empty and close the cart, select all->remove
-- either way window closes and it looks empty.
5) refresh the page
-- cart comes back.
-- in chromium inspecting the page, specifically
the "Application" area, bib_list is shown as
a cookie still.
6) apply patch
7) restart_all and empty cache forcing JS reload
8) repeat steps 1-5
-- this time cart stays empty.
-- and the bib_list cookie is gone
Signed-off-by: Roch D'Amour <roch.damour@inlibro.com> using firefox 59.0.2
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 16c6ff1e3c)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
To test:
1 - Open the advanced editor
2 - Make some changes to a record
3 - Click on 'Patrons' or leave the page in some way
4 - Note there is no warning and changes are not saved
5 - Apply patch
6 - Open the advanced editor
7 - Make some changes
8 - Leave the page
9 - Note you get a warning of unsaved changes and can stay on the page
Signed-off-by: Charles Farmer <charles.farmer@inLibro.com>
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 5a6e7a6ffb)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Jenkins fails (run 287) on a test in t/db_dependent/Letters/TemplateToolkit.t:
With the historical syntax:
# Your request for an article from tQYRS (c3Av58O0P5xkkIGu) has been canceled for the following reason:
With the TT syntax:
# Your request for an article from tQYRS_ (c3Av58O0P5xkkIGu) has been canceled for the following reason:
The last character of the biblio's title has been removed because it's a punctuation character.
It comes from: C4::Letters::_parseletter
893 $val =~ s/\p{P}$// if $val && $table=~/biblio/;
The same replacement is done for patron's attributes too.
Test plan:
- Confirm that the new tests pass. That should be enough to confirm this change make sense.
Test plan (manual):
- Create a biblio with a title ending with a punctuation (like "with_punctuation_"), or any other fields of biblio/biblioitem
- Generate a notice which will display this field (CHECKIN for instance)
Use the historical syntax and the TT syntax, both should display the title without the punctuation character at the end
CHECKIN historical:
The following items have been checked in:
----
<<biblio.title>>
----
CHECKIN TT syntax:
The following items have been checked in:
----
[% biblio.title %]
----
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit f5be2d525f)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
To test:
1 - Add a course (Course Reserves)
2 - Add an item to the course
3 - Ensure to change the holding branch for the course reserve
4 - Modify the course item
5 - Note the dropdwn for holding branch is unset
6 - Apply patch
7 - Modify the course item
8 - Dropdown should be correctly populated
Signed-off-by: Maksim Sen <maksim.sen@inlibro.com>
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 03b31bfc62)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Test Plan:
1) Enable EasyAnalytics
2) Disable QueryAutoTruncate
3) Create an analytic record, add some host items to it
4) Browser to the analytics tab for the record
5) Click the link in the 'used in' column of the table
6) No search results
7) Apply this patch
8) Reload the page, now you get the analytic record!
Signed-off-by: Séverine QUEUNE <severine.queune@bulac.fr>
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 14b70dc910)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This has been done on bug 9509 when barcodes are scaned, but not when using a text file.
Test plan (stolen from Barton on the bug report):
1) Create a text file containing duplicate barcodes
Here's a quick way to do that, substitute INSTANCE for the instance name of your test server:
sudo koha-mysql INSTANCE <<< "select barcode from items where barcode is not null limit 3" | sed '1d;p' > /tmp/dup_barcodes.txt
2) Go to Home › Tools › Batch item deletion
3) Click 'Choose File', select the file you created in step 1.
4) Click 'Continue'
5) Note that duplicate items appear in the list of items to batch delete
6) Check 'Delete records if no items remain'
7) This will trigger a software error
Note that entering the barcodes via the 'scan items one by one' text box removes duplicates in the list, so this only causes problems when uploading a file.
Signed-off-by: Mark Tompsett <mtompset@hotmail.com>
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 1f4baf31bb)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
When you want to merge invoices you have a page with a non editable
field 'Invoice number:' it shows the content of the first invoice to be
merged.
But if you validate by clicking the 'merge' button you arrive on the
next page which says that 'Invoice has been modified' and if you quit
this page without saving you have now a merged invoice without Invoice
number.
This tiny patch just prevents this issue to occur.
Test plan :
1° go to the acqui/invoices.pl page and search invoices to merge.
You must have at least 2 invoices on the same vendor
2° check boxes to select invoices to merge and click on 'merge selected
invoices' button
3° next page you see the non editable field 'Invoice number:'
4° click on the 'merge' button
5° next page you see 'Invoice has been modified'
6° leave this page i.e click on the left link 'Invoices'
7° search invoices you'll see the merged invoice without invoice number.
Apply the patch, replay the steps 1 to 3
4° on this page you can enter your invoice number and click on the
'merge' button.
If you leave this field empty and click merge, a message informs you
that it is required and you can not merge.
Signed-off-by: Séverine QUEUNE <severine.queune@bulac.fr>
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Conflicts:
koha-tmpl/intranet-tmpl/prog/en/modules/acqui/invoices.tt
(cherry picked from commit 3838dd8375)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit a444807b57)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
We lost the ability to place multiple holds when we are searching for
patrons. The multi_holds parameter is lost and not handled correctly in
the template.
Test plan:
- Make sure you can place multiple holds for a patron you will search for
- Same for simple hold
TODO the multiple holds view should not be displayed if only 1 record
has been selected from the search result.
Signed-off-by: claude brayer <claude.brayer@cea.fr>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit dc008b90cd)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
category_type must be != "X"
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Conflicts:
t/db_dependent/Patron/Borrower_PrevCheckout.t
(cherry picked from commit 082e4e2571)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
NOTE: Confirmed that the typo of 17641 was part of bug 17941.
Signed-off-by: Mark Tompsett <mtompset@hotmail.com>
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 6852c592f7)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This patch will have to be tested deeply to make sure it will not
introduce regression!
The idea is to display an empty option in the patron attributes select
and ignore it. That way we can deal with false values 0 and "" which
were skipped before.
Test plan:
Add several patron attributes
Use the batch patron modification tool to add/update/remove them
Play with empty "" and 0 values, as well as other values
Modify several attributes in a row
Signed-off-by: Charles Farmer <charles.farmer@inLibro.com>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Conflicts:
koha-tmpl/intranet-tmpl/prog/en/modules/tools/modborrowers.tt
(cherry picked from commit 16dfb0309b)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
We may need to apply this change for all other occurrences in our
codebase.
Test plan:
- Make sure sample patron data is loaded and you have a patron without a
date expired defined
- Make an empty search for all patrons or another search that gives you
a good amount of results
- Try to sort on the date expired column
=> Without this patch you get
Odd number of elements in anonymous hash
at /home/vagrant/kohaclone/C4/Utils/DataTables/Members.pm line 189.
output_pref is called with 'dateonly' (ref SCALAR), not a DateTime
object
Signed-off-by: Mark Tompsett <mtompset@hotmail.com>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 09eaaaebcd)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
To test:
1 - Run the circulation wizard with itemtypes as columns, shelving
locations as rows - display on screen
2 - Note all itemtype values are blank
3 - Switch rows/columns - same issue
4 - Apply patch
5 - Re-run reports
6 - Itemtypes should display correctly
Signed-off-by: Claire Gravely <claire.gravely@bsz-bw.de>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit b3bab56c31)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Bug 18571 added default ES configuration for packaging config.
This patch adds it to dev install in koha-conf.xml.
Database name is used in index_name to allow multiple installs.
Test plan :
- Run dev install
- Install ElasticSearch server and Koha deps
- Enable ElasticSearch in Koha
- Check indexing and searching works directly
Signed-off-by: Brendan Gallagher <brendan@bywatersolutions.com>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 219e427f29)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
In subscription-add.pl, the two params aren't passed to
NewSubscription() but they are to ModSubscription()
== Test plan ==
1. Enable syspref "makePreviousSerialAvailable"
2. Create a minimal subscription with a value in the "Item type" and
"item type for older issues" fields.
3. Edit the subscription
4. You should see that the two fields are empty. This is the bug.
5. Apply this patch
6. Do step 2 and 3
7. You should see that the two fields have the right value
8. Express the joy of a successful and easy sign off.
(this is important, otherwise the sign off spell won't work!)
Signed-off-by: delaye <stephane.delaye@biblibre.com>
Signed-off-by: Séverine QUEUNE <severine.queune@bulac.fr>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit b02528d8bb)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
When bibliographic records are merged, the suggestions.biblionumber
field should be replaced with the new record.
Ideally we should have kept an history of the merges, to know what was
the original record. Now we cannot fix the broken links.
Test plan:
1) Create a suggestion in OPAC
2) Accept the suggestion in staff
3) Add order from this suggestion
4) Go to cataloguing search and search for the record and another
5) Merge the suggestion record with the catalog record - catalog record wins (should be the more common case when a patron suggests something that already exists)
6) Verify that after merging, the new title displays in the acquisition record
7) Verify that the suggestion info no longer displays in basket
From Katrin's test plan:
8) Verify that the new title doesn't display in suggestions. The old title will still show as suggested (not updated)
=> This is the title from suggestions.title, I think it makes sense to show this one.
Maybe we should open a new bug report to improve the display of the suggestion, and link to the bibliographic record
Signed-off-by: Lucie Gay <lucie.gay@ens-paris-saclay.fr>
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 3f118604af)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Fix the same error in another place
Signed-off-by: Julian Maurice <julian.maurice@biblibre.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 174dc4cb2e)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
If you use bulkmarcimport.pl to import records with items it looks
like the successfull insert of the record is reported multiple time,
but the second and subsequent "ok" is really related to importing
the item(s).
This patch changes the log message on successfully inserting an item
to match the log message given when inserting an item fails.
To test, the easy way:
- Look at lines 530 and 536 of bulkmarcimport.pl, and note that the
"op" in those two lines are different
- Apply the patch
- Look at lines 530 and 536 again, and note that the "op" is now
identical, and that this makes sense, since they are both related
to the same operation, specifically inserting an item
To test, the hard way
- Have some records with items
- Import the records with bulkmarcimport.pl, and make sure to specify
the -l option, to create a log of the actions taken
- Look at the log and verify it looks something like this:
id;operation;status
1;insert;ok
1;insert;ok
2;insert;ok
2;insert;ok
- Apply this patch and import some more records with items. The log
should now be similar to this:
id;operation;status
1;insert;ok
1;insertitem;ok
2;insert;ok
2;insertitem;ok
Signed-off-by: Maksim Sen <maksim.sen@inlibro.com>
Signed-off-by: Julian Maurice <julian.maurice@biblibre.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit aa7954907b)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This patch adds a check in admin-home.pl to see whether plugins are
enabled. If plugins are disabled, the link to plugins management is
hidden even if the user has plugin management permission.
To test, view the administration home page with plugins enabled and
disabled via the enable_plugins flag in the Koha configuration file.
Also test using users with different plugins permissions:
- Plugins enabled
- CAN_user_plugins = 1
-> Plugins link appears
- CAN_user_plugins = 0
-> Plugins link hidden
- Plugins disabled
- CAN_user_plugins = 1,
- CAN_user_plugins = 0
-> Plugins link hidden
Signed-off-by: Séverine QUEUNE <severine.queune@bulac.fr>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 8450d2020b)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Differences in markup around some toolbars has the potential to create
CSS styling problems. This patch changes the markup to be consistent
with other instances of <div id="toolbar"></div>
To test, apply the patch and view these pages:
- Course reserves (main page)
- Course reserves -> Course detail page
- Reports -> Dictionary
On each of these pages the toolbar should look correct.
Signed-off-by: Claire Gravely <claire.gravely@bsz-bw.de>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit c9ee7c8d3b)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Test Plan:
0.1 - find at least one biblionumber that has at least one item checked out
0.2 - find at least one biblionumber that has zero items checked out.
1 - Go into the module "Tools > Batch record deletion"
2 - in "List of biblionumbers or authority ids (one per line)" put the values you found in step 0.1 and 0.2
3 - click continue
4 - biblionumbers with at least one item checked out should not be clickable.
5 - biblionumbers with zero items checked out should be clickable.
6 - click "Select all" and "Clear all", both buttons should change the "checked" status of the enabled AND disabled biblionumber checkboxes.
7 - apply patch
8 - click "Select all" and "Clear all", both buttons should change the "checked" status of ONLY the enabled biblionumber checkboxes.
Signed-off-by: JM Broust <jean-manuel.broust@univ-lyon2.fr>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Conflicts:
koha-tmpl/intranet-tmpl/prog/en/modules/tools/batch_delete_records.tt
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 336e7faaf4)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Followed test plan. Patch functions as described and both commands execute.
Signed-off-by: Dilan Johnpullé <dilan@calyx.net.au>
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 931db23ee3)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This attempts to handle issues arising when running
koha-create on a system that never had MySQL installed.
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit a85d33314c)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This is a QA follow-up of bug 19436, submitted separately for ease of
backporting it. (The bug was not raised by 19436.)
Obviously, the number of authority servers not biblio servers should
be the key factor for showing or hiding the button for new authority
records via Z39.50/SRU.
Trivial fix.
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 7852d967d1)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Actually catalog detail page shows for damaged and lost items the authorized value description if exists.
We must do the same for withdrawn information.
Test plan :
1) Set for items a subfield linked to items.withdrawn and using the authorized value category 'WITHDRAWN'
2) Edit the authorized value category to add value '2' and description 'Eaten by dog'
3) Edit an item to set this withdrawn value
4) Go to staff interface and look at items in record detail page /cgi-bin/koh/catalogue/detail.pl :
=> Without patch you see 'Withdrawn' and with patch you see 'Eaten by dog'
5) Go to OPAc and look at items in record detail page /cgi-bin/koha/opac-detail.pl :
=> Without patch you see 'Withdrawn' and with patch you see 'Eaten by dog'
Signed-off-by: delaye <stephane.delaye@biblibre.com>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 830ce35795)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Fixes this error:
Undefined subroutine &main::MarkIssueReturned called at
misc/cronjobs/longoverdue.pl line 316.
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Julian Maurice <julian.maurice@biblibre.com>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 4aaf30dfd8)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
The longoverdue.pl option --mark-returned doesn't work unless the
--charge option is used as well.
Test Plan:
1) Run long overdue with --mark-returned and not --charge,
note your items are marked lost but not returned
2) Apply this patch
3) Repeat step 1, the items should now get returned!
Tested with (for example):
misc/cronjobs/longoverdue.pl --lost 10=1 --mark-returned --verbose
--confirm
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Julian Maurice <julian.maurice@biblibre.com>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 93c5c8e4c8)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This test does obviously not achieve the desired result:
[ "chdir" != "no" ]
Trivial fix.
Adding the same quotes around starting_dir (just as for Bug 19546).
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 7c55ea5917)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
In order to patch production sites we need to adjust the shipped
cronjobs so they are called with the --chdir option switch.
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 37750b2f20)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Until Perl 5.26, the current directory is added to @INC when running a
Perl script [1]. Having the current directory in @INC means it can be
tried to be traversed when performing a lib lookup. Since version 5.18,
Perl dies when it finds an unreadable directory (permissions) in @INC
that needs to be traversed. This behaviour won't change because Perl
devs consider it an enhancement to security. [2]
Because of this, we need to make sure our scripts are ran **from** a
directory in which they have read permissions.
Ths patch adds a --chdir option switch to the **koha-foreach** wrapper
script, that makes the inner shells/scripts to be ran within the Koha
instance's user home directory.
The change is trivial and should be QAed easily. I tested this on a prod
server:
- Create a /tmp/test.pl file containing:
use Modern::Perl;
use Cwd;
my $dir = getcwd;
warn $dir;
1;
A) then create a cronjob entry to run it using koha-foreach:
(in /etc/cron.d/test):
1/* * * * * root koha-foreach perl /tmp/test.pl
- Once I noticed the cronjob ran, I used mutt to read the emails in the
root user.
=> FAIL:
...
Subject: Cron <root@koha> koha-foreach --enabled perl /tmp/test.pl
"/root"
"/root"
"/root"
"/root"
"/root"
...
B) I then used the patched koha-foreach with different results:
=> SUCCESS:
...
Subject: Cron <root@koha> /root/koha-foreach --chdir --enabled perl /tmp/test.pl
"/var/lib/koha/acaderc"
"/var/lib/koha/agro"
"/var/lib/koha/anc"
"/var/lib/koha/arico"
"/var/lib/koha/artes"
...
So this patch's approach works. But...
C) master's koha-foreach seems to work just the same... I think it is
because of my previous attempt to fix this by using sudo in koha-shell.
So I think environmental conditions affect the behaviour (which shell is
configured for cron, sudo configuration, etc).
====
In conclusion, I think we should go ahead with this patch as it will solve
peoples issues, and it is a right solution (option #5 on the list) to
this Perl behaviour change. It doesn't cover other commands, but
followup patches could do.
I avoided /tmp as it is writable by any user... so it is an easy path
for both exploiting by replacing some lib, and also because the
existence of an unreadable dir that the interpreter could try to
traverse (unreadable /tmp/Authen or /tmp/Koha will trigger the same
error, and I assume people know what they are putting on the instance's
dir, at least it will be easier to track).
A followup patch takes care of making the cronjobs use --chdir when
calling koha-foreach
[1] https://lists.debian.org/debian-devel-announce/2016/08/msg00013.html
[2] https://rt.perl.org/Public/Bug/Display.html?id=123795
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 3db7e1a63a)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
This patch simplifies the koha-foreach script, making it use koha-shell
to gain instance privileges and have all environment variables set.
To test:
- Apply the patch
- Have one or more instances created (kohadev already exists in kohadevbox, add
another onewith $ sudo koha-create --create-db test)
- Run:
$ sudo kohaclone/debian/scripts/koha-foreach echo "This is instance __instancename__"
=> SUCCESS: The script runs gracefuly, and outputs something like:
This is instance kohadev
This is instance test
Note: this means the command was executed for each instance and the current behaviour
of replacing the placeholder __instancename__ for the actual instance name still
works as expected.
- You can try the different filter switches, but the logic has not been touched. only the
command call.
- Sign off :-D
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 75994dc94a)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Bug 19568 use the TT filter url whereas it must have escaped the URI
characters with the uri filter.
Test plan:
Subscribe to an OPAC search
=> Without this patch you will get an empty page
=> With this patch applied you will see result list.
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Brendan Gallagher <brendan@bywatersolutions.com>
Signed-off-by: Julian Maurice <julian.maurice@biblibre.com>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
(cherry picked from commit 5de7dc040d)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
Today Feb 28th 2018:
If we subtract 18y to this DateTime object 29/02/2000 and so the patron is not major yet
use Koha::DateUtils qw( dt_from_string );
say dt_from_string('28/02/2018')->add(years => -18);
say dt_from_string('28/02/2018')->add(years => -18, end_of_month => 'limit');
will display:
2000-02-29T00:00:00
2000-02-28T00:00:00
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 72de4ddaee)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>
The service description is very old and outdated. This patch fixes this.
Sponsored-by: Orex Digital
Signed-off-by: Hugo Agud hagud@orex.es
Signed-off-by: Josef Moravec <josef.moravec@gmail.com>
Signed-off-by: Jonathan Druart <jonathan.druart@bugs.koha-community.org>
(cherry picked from commit 362c0e0cd1)
Signed-off-by: Fridolin Somers <fridolin.somers@biblibre.com>