For historical reasons the SIPServer and SIP modules
have used an extra module path in addition to the
standard Koha one. This has caused numerous irritants
in attempting to set up scripts and basic tests. It
does not help in attempting to modify or debug
this code
This patch changes the package value in the modules
under the C4/SIP directory and makes calls to
them use the full package name.
Where the export mechanism was being short circuited
routines have been explicitly exported and imported
declarations of 'use ILS' when that module was
not being used and which only generated warnings
have been removed.
As a lot of the changes affect lines where
an object is instantiated with new. The opportunity
has been taken to replace the ambiguous indirect
syntax with the preferred direct call
In intializing ILS the full path is added as this
will not require any changes to existing configs.
I suspect this feature is unused, and adds
obfuscation rather than flexibility but have kept
the feature as we need this change in order to
rationalize and extend the testing of the server.
The visible difference is that with the normal Koha
PERL5LIB setting. Compilation of Modules under C4/SIP
should be successful and not fail with unlocated modules,
allowing developers to see any perl warnings
All the SIP modules can now be run through the tests
in t/00-load.t now except for SIPServer itself
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@biblibre.com>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
Koha's SIP server accepts timeout parameters, but those parameters
are only used for the login portion of a telnet transport session.
Other than that, they are ignored, and as long as whatever opened
the connection keeps it open, it will stay open indefinitely.
Test Plan:
1) Set the timeout setting on your SIP server to 10 seconds
2) Modify misc/sip_cli_emulator.pl, add "sleep 100;" directly after line 91
3) Start your SIP server
4) Run the modified sip script with valid parameters
5) Watch the SIP server stderr
6) Note that even though the script waits far too long before continuing,
the SIP server never kills the connection, and the requests the cli
script makes come back with valid data.
7) Apply this patch
8) Restart your SIP server
9) Repeat step 4
10) Note that this time you see "SIP Timed Out!" in the SIP server
stderr and when the script finally makes it's request, it doesn't
come back with valid data.
Signed-off-by: Colin Campbell <colin.campbell@ptfs-europe.com>
Signed-off-by: Jonathan Druart <jonathan.druart@biblibre.com>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
The SIP config has allowed you to specify an interface ip as
part of the listeners/service/port attributei
e. g. as port="127.0.0.1:6001/tcp"
with IPv6 the equivalent would normally be
as port="[::1]:5001/tcp"
However in this case incoming connections will get rejected because
Configuration constructs a string without the brackets
This patch makes tests both formats on incoming connections so that
they are accepted as they were previously
In future the best course is not to include a port identifier in the
port definition then if the server has ipv6 it will bind to all
interfaces and accept both IPv4 and IPv6 traffic
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
This small patch adds a check on the SIP2 socket connection if it is
IPv6 and resolves socket address accordingly.
Any newer Debian distro would probably default to IPv6 so it would
eventually affect all SIP servers.
Tests against running SIP server on an IPv6 box:
http://wiki.koha-community.org/wiki/Koha_SIP2_server_setup#Testing_with_Telnet
before patch:
disconnects immediately. Log output:
Bad arg length for Socket::unpack_sockaddr_in, length is 28, should be 16
after patch:
operates normally
Signed-off-by: Colin Campbell <colin.campbell@ptfs-europe.com>
Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
We should add the ability to apply a regular expression to screen
messages for the SIP2 server. This would allow libraries to not only
customize the screen messages the patron sees, but can also allow screen
messages to be translated.
Test Plan:
1) Apply this patch
2) Inspect etc/SIPconfig.xml, note the new screen_msg_regex tags
that can be nested inside a given login tag.
3) Add one or more screen_msg_regex tags to your own SIP config
Recommendation: s/Greetings from Koha./Welcome to your library!/g
4) Restart your SIP2 server
5) Test with a SIP2 machine, or use /misc/sip_cli_emulator.pl
6) Note your new AF fields!
Signed-off-by: Jason Burds <jburds@dubuque.lib.ia.us>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
If the patron has amassed charges that block borrowing, but we
allow staff override the information that the patron cannot
issue should be included the patron information response
This patch sets the appropriate status fields in the patron object
It restores the fee_limit member to the patron object
and calls a local subroutine to set it.
This could be done more elegantly but that would require more
major refactoring of the rather messy initializer code
in ILS::Patron
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
If a patrons fines exceed noissuescharge and we allow
staff to allow issue at their discretion via an override
the SIP process allowed charges to go ahead.
This patch closes the loophole which allowed self issue
to subvert the usual library loan policy
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
If you do a kill -s HUP sipserver_pid the server process
should restart its children.
This was not happening although it was logging the commencement
of the process.
Removed traces of Exporter usage this was always incorrect
this program inherits from Net::Server 'use base' to make inheritance
clearer. Removed unnecessary begin block
made file executable so it can run
added a shebang line so OS knows to invoke perl to compile and run
the file.
I've removed the debug code dumping the config to STDOUT,
it should not have gone into the production version.
One of the first things this program does is close stdout etc and
reopens the sip sockets to them. Assuming you can print to them other
than as part of the sip dialogue is not always safe.
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
Bug 12041 made xt/author/podcorrectness.t consider files in the 'Koha' namespace.
Some of them where failing. This patch fixes some of those POD problems.
Best regards
To+
Test:
1) run prove xt/author/podcorrectness.t
it fails
2) apply patch
3) run again, now it's ok
Signed-off-by: Bernardo Gonzalez Kriegel <bgkriegel@gmail.com>
Before patch test fails. After it, it pass
No koha-qa errors
Signed-off-by: Katrin Fischer <Katrin.Fischer.83@web.de>
Signed-off-by: Tomas Cohen Arazi <tomascohen@gmail.com>
Currently Koha's SIP server ignores the return date part of an 09 ( aka
CHECKIN ) message. Koha should backdate a return, and remove
fines accordingly.
Signed-off-by: Benjamin Rokseth <benjamin.rokseth@kul.oslo.kommune.no>
Works as notified, second date field in SIP checkin (return date) is
used as return date. Return shows up in history with correct date.
Comments:
- patron is fined if return date is before issue date, but that
is largely irrelevant anyway.
Signed-off-by: Martin Renvoize <martin.renvoize@ptfs-europe.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
This patch fixes various issues that prevented the account
type from being set correctly when recording SIP2 payments:
- the fixed fields in the fee paid message were not getting
parsed correclty
- accountlines.accounttype is only five characters wide, so
SIP2 payments are now records as 'Pay00', 'Pay02', etc. rather
than 'Pay-00'.
- removed regression on bug 2546 and made the new payment types
translatable
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
Adds support for recording in Koha accountlines a different accounttype
depending on how the fee was paid at the SIP2 station (cash, credit,
etc.)
Adds a new param to recordpayment(), $sip_paytype, which is appended to
the 'Pay' type if present. The payment description is also appended
with " (via SIP2)" if this is present.
In order for other scripts to keep working as expected, "eq 'Pay'"
needed to be replaced with a regex comparison "=~ /^Pay/", so that 'Pay'
and 'Pay-##' would continue to group together.
To test:
1. Make a payment over a SIP2 connection
2. Check the patron record in the staff client; you should see the
modified description
3. Attempt to print a invoice or a reciept for the borrower; the
payment should show up where expected
Signed-off-by: Benjamin Rokseth <benjamin.rokseth@kul.oslo.kommune.no>
Signed-off-by: Brendan Gallagher <brendan@bywatersolutions.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
The accounts->login tag in SIPconfig.xml can now accept a new
parameter, "encoding". It will be mostly used to encode to utf8.
For this, simply add the parameter: encoding="utf8"
Signed-off-by: Chris Cormack <chrisc@catalyst.net.nz>
Works as advertised, does nothing if encoding is not set.
Blows up all the machines that can't handled utf8 if it is set :) But
that's not Koha's fault. :)
Patch rebased by Christophe Croullebois <christophe.croullebois@biblibre.com>
Signed-off-by: Petter Goksoyr Asen <boutrosboutrosboutros@gmail.com>
But now I did it the right way! And I can confirm that this patch solves
all issues with mangled characters in SIP messages. Confirmed that it
looks good with Norwegian characters in patron name and in book titles.
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
Koha's SIP2 server implementation does not currently support the SIP2
protocol field "media type" ( CK ).
This patch implements the SIP2 media type by allowing an arbitrary
mapping of itemtypes to SIP2 media types.
Test Plan:
1) Apply this patch
2) Run updatedatabase
3) Edit an itemtype, select a SIP media type, and save the changes
4) Make a SIP2 Item Information Request
5) Verify that the CK field of the Item Information Response contains
the correct media type code.
Signed-off-by: Benjamin Rokseth <benjamin.rokseth@kul.oslo.kommune.no>
Signed-off-by: Martin Renvoize <martin.renvoize@ptfs-europe.com>
If a patron has a record-level hold that is unavailable, any patron
information request will send back an empty CD field when this field
should have an item barcode in it [RM note: this actually isn't
universally true -- the SIP2 standard is silent as to what is supposed
to go in the CD field. Some SIP2 devices do indeed want an item
barcode, but others are known to just want a display of the title
and author of the request in question. Providing an option is the
topic of a new enhancement request, however.]
This is due to a minor error in ILS::Patron::_get_outstanding_holds
where GetItemnumbersForBiblio is assumed to return an array but in
reality returns an arrayref.
Test Plan:
1) Create a record level hold for a patron and record
2) Using SIP2, make a patron information request
3) Note the empty CD fields
4) Apply this patch, restart SIP server
5) Repeat step 2
6) Note the CD field now has a barcode
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Jonathan Druart <jonathan.druart@biblibre.com>
I did not test this patch but the following code shows me it is correct:
use C4::Items;
use Data::Dumper;
my $biblionumber = 5035;
my $itemnumber = (GetItemnumbersForBiblio($biblionumber))[0];
say Dumper $itemnumber;
$itemnumber = (GetItemnumbersForBiblio($biblionumber))->[0];
say Dumper $itemnumber;
displays:
$VAR1 = [
'23168',
'23169',
'23170',
'23171',
'23172'
];
$VAR1 = '23168';
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
The return from GetReservesFromBiblionumber contains an unnecessary
extra variable. In scalar context an array returns its element count.
Maintaining a separate count can lead to unforeseen bugs
and imposes ugly constructions on the subroutine's users.
Remove the useless count variable from the return
This patch also changes the parameters: now the routine takes a hashref.
Signed-off-by: Marcel de Rooy <m.de.rooy@rijksmuseum.nl>
Placed biblio holds, future holds and item holds. Works as expected.
Tested Holds.t and Reserves.t. Pass.
Tested /cgi-bin/koha/ilsdi.pl?service=GetRecords&id=999 with two holds on
one item. Fine.
C4/SIP/ILS/Item.pm: Looked for "whatever" and "arrayref" and could not find
them anymore. Looks good.
Handled a few unneeded calls in QA follow-up.
Left only one point to-do for serials/routing-preview.pl. See Bugzilla.
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
The system preference 'maxoutstanding' is defined as the maximum amount
of fees owed by the patron before Koha should block placing holds (
terrible naming on this one ).
However, although the Koha OPAC respects this preference, placing holds
via a SIP2 device will not.
Test Plan:
1) Set maxoutstanding to $10
2) Pick a patron owning more than $10 in fees
3) Attempt to place a hold for this patron from a SIP2 device
This attempt should succeed
4) Apply this patch
5) Restart your SIP2 server
6) Attempt to place a hold for this patron from a SIP2 device again
This attempt should now fail
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Brendan Gallagher <brendan@bywatersolutions.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
Test:
* SIP: Have an old user and create a new user
- use either tenet sip test or
C4/SIP/interactive_patron_check_password.pl to check old
userid/password
- do the same for the new user
Signed-off-by: Bernardo Gonzalez Kriegel <bgkriegel@gmail.com>
Work as described
Test
1) using perl C4/SIP/interactive_patron_check_password.pl
can check current (short) and new (long) passwords
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
When requested by the summary flags the sipserver
should return in the patron info response barcodes of
the relevant titles.
For available holds this is the barcode of the captured items.
For unavailable holds ( i.e. current unsatisfied holds ),
we need to send a barcode so that the unit can use this to
request the title info. The barcode could be any one
belonging to the title.
This patch also corrects the erroneous return of empty items
in the patron information response. If the unit supplies a
range 1 - 100 unless the title has a hundred or more copies the
unit expects all copies. The server was erroneously stuffing
the returned arrays with null copies so that all summary requests
returned 100 copies (mainly without barcodes)
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Katrin Fischer <Katrin.Fischer.83@web.de>
Testing notes:
Using the test script provided on the bug report, but changed it
to match sip user and patron existing in my database.
Before applying the patch the SIP responses show the behaviour
pointed out above regarding the 100 items. After applying the
patch and restarting the SIP server responses are much more clean
not returning empty IDs.
64 Patron information response
AS = hold items
hold items count is correct.
AS contains barcodes of waiting holds.
Before patch, all AS were empty.
AT = overdue items
overdue items count is correct.
AT contains barcodes of overdue items.
AU = charged items
charged items count is correct.
AU contains barcodes of charged items.
AV = fine items
Judging from behaviour seen and comment in
Patrons.pm currently not implemented.
BU = recall items
Recalls are not implemented in Koha yet.
CD = unavailable hold items
unavailable items count is correct.
CD contains barcode for item level holds and is empty
for title level holds where no item can be determined.
Before patch, all CD were empty.
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
This patch gives you the option of sending a patrons home branch code
in an AF field for patron status requests. It is controlled at the account
login level, so it can be enable on a per-sip-login basis.
Test Plan:
1) Apply patch
2) Edit SIPconfig.xml, add the parameter 'send_patron_home_library_in_af="1"'
to the login you will be using to test.
3) Start your SIP2 server.
4) Connect to it via telnet ( something like: '9300CNterm1|COterm1|CPCPL|' )
5) Send a patron status request ( like: '2300120121110 82925AOCPL|AA23529000035676|ACterm1|ADletmein' )
6) Examine reponse you should see something like this:
"24 00120121210 085332AEHenry Acevedo|AA23529000035676|BLY|CQN|AFGreetings from Koha. |AFMPL|AO|"
Note the second AF field with the value MPL.
Signed-off-by: George Williams <georgew@latahlibrary.org>
Signed-off-by: Christopher Brannon <cbrannon@cdalibrary.org>
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
The tests were too restrictive. It should be possible to pass
a checkin/checkout test with different values defined for
magnetic media.
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
Add a terminator option to SIPConfig.xml, choices for 'terminator' are
'CR' or 'CRLF'. The default continues to be 'CRLF' if 'terminator' is
undefined.
Test Plan:
1) Apply patch
2) Start SIP server
3) Run C4/SIP/t/04patron_status.t
4) Stop SIP server
5) Add terminator="CR" for account login 'term1'
6) Run 04patron_status.t again, you should see no change
Signed-off-by: Frédéric Demians <f.demians@tamil.fr>
Signed-off-by: Adrien Saurat <adrien.saurat@biblibre.com>
Signed-off-by: Chris Cormack <chrisc@catalyst.net.nz>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
This patch updates the wthdrawn field in items and deleteditems to be
withdrawn instead. No functional changes are made.
Signed-off-by: Bernardo Gonzalez Kriegel <bgkriegel@gmail.com>
Comment: Save for translation files (that will be fixed on next release),
only occurrence of wthdrawn is on updatedatabase.pl
No koha-qa errors.
This touch many files, and I did not test everything,
but all seems normal. I think that any problem could
be fixed later.
Perhaps both entries in updatedatabase.pl could be joined
into one, but thats for QA.
Signed-off-by: Jonathan Druart <jonathan.druart@biblibre.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
- fix identation in one line
- remove a commented-out warn
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
This patch switches from using a combination of
biblionumber/borrowernumber to using reserve_id where possible.
Test Plan:
1) Apply patch
2) Run t/db_dependent/Holds.t
Signed-off-by: Maxime Pelletier <maxime.pelletier@libeo.com>
Signed-off-by: Jonathan Druart <jonathan.druart@biblibre.com>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Galen Charlton <gmc@esilibrary.com>
Some patron infos were hard coded instead
of using the variables defined in SIPtest.pm
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@biblibre.com>
Signed-off-by: Jared Camins-Esakov <jcamins@cpbibliography.com>
The debarred status in DB switched from a boolean to a date.
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Elliott Davis <elliott@bywatersolions.com>
Correctly pulls debared flag instead of debarment date.
Signed-off-by: Jared Camins-Esakov <jcamins@cpbibliography.com>
Adding a use English to mosules which dont may
waste a few CPU cycles and bytes of storage
but doesnt add anything to the sum of human happiness
Remove unnecessary use English's
Removed already commented out but equally unnrcessary
use Exporter in same modules
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Jonathan Druart <jonathan.druart@biblibre.com>
Signed-off-by: Jared Camins-Esakov <jcamins@cpbibliography.com>
Expansion of Liz Rea's original patch
Bug report specifically mentioned lack of the
Zip/Postal code but all used address fields should
be included. Moved the address
generation to its own sub.
Hopefully using an if else will read more clearly than
the original ternary operator proved to be
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Jared Camins-Esakov <jcamins@cpbibliography.com>
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
Signed-off-by: Jared Camins-Esakov <jcamins@cpbibliography.com>
All the modules in the SIP/ILS tree are objects
The addition of calls to Exporter or hand manipulation of
@ISA added unnecessary bloat
Removed the "self = shift or return" idiom as it is nonsensical
if the method can only be called via an object.
standardized inheritance via use parent
added a $self = shift in a couple of places where it
was not strictly necessary as its absence seemed to have
misled readers in the past
Signed-off-by: Kyle M Hall <kyle@bywatersolutions.com>
Passed-QA-by: Mason James <mtj@kohaaloha.com>
Signed-off-by: Jared Camins-Esakov <jcamins@cpbibliography.com>
Rebase and merge based on Mark Gavillet's patch to reduce
the loan period given to items on renewal when they are
in high demand for holds
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Renewals were being rejected for incorrect reasons
Checking was being done against the wrong object
Add more informative messages on failure
Correctly set due_date for renewal response
Avoid crashing the SIPServer because it handles RenewAll
incorrectly
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
SIPServer.pm requires that C4/SIP is added to its lib
path This has been done by passing this directory
to it via -I. By using FindBin it can set the path
for itself correctly. This will also work if the C4/SIP
directory tree is moved to a non-standard location
Removed the now redundant -I. from sip_run.sh
Added a variable to sip_run.sh for the koha tree to
highlight a problem with the script if you have multiple
directories in the PERL5LIB environment variable
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
Be liberal in what we accept, but strict in what we send:
Never exit the server process, but send a SC_RESEND message (96)
to the client if we received anything we don't understand.
This is consistent with SIP server implementations of other ILSs.
Signed-off-by: Colin Campbell <colin.campbell@ptfs-europe.com>
Signed-off-by: Ian Walls <koha.sekjal@gmail.com>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
Most units require or tolerate CRLF as the terminator
Using only \r causes many including 3M units to signal errors
due to delay this resets CRLF to be default but adds a single
constant variable and explanation so that if you have a
"strictly conforming" unit (the 'Baby' is the only I have encountered
you know where to make the change
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
Add C4/SIP to perlcritic tests. Fix those issues that were
generating perlcritic errors
Signed-off-by: Stacey Walker <stacey@catalyst-eu.net>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
If a discharged item is captured for a hold at another branch
the selfcheck should generate a branchtransfer for it.
This was complicated as the checkin routine was setting its own
location from a field not set in many scm machines (e.g. 3M)
also in ILS::Transaction the test for location was incorrect
assuming that an empty string or undefined was a valid location
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
Shelf Check was receiving messages saying item was wanted
for a hold but the item was discharged to the shelf
not associated with the hold or transited to
the pickup location. The message was also being sent
on discharge of items when a suitable item had already
been captured.
Checkin now associates the item with the hold and sets
the appropriate data for a correct checkin response
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
The server returns its branch as the institution in the config file
However in transactions it will default to the homebranch of the
sip user which may not be consistent with that. Needs to set branch
in the session if there is a value in config otherwise statistics and
behaviour can be wrong (esp in discharge)
Signed-off-by: Chris Cormack <chrisc@catalyst.net.nz>
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
Explicitly clear these arrays and update them directly.
In practice item numbers were already in them when called
so that more items were being returned than were actually
renewed (or failed)
Signed-off-by: Chris Cormack <chris@bigballofwax.co.nz>
Functionally sound patch, and tested in production
Signed-off-by: Paul Poulain <paul.poulain@biblibre.com>
