The *_ok methods in ILS.pm were targeting the wrong depth.
This also resolves a longstanding FIXME on to_bool() warning like:
Argument "\x{66}\x{61}..." isn't numeric in numeric ne (!=) at /ILS.pm line 94.
The example_institution_dump.sh essentially provides the proof test case for this patch.
Run it before/after on SIPconfig.xml where "MAIN" has checkout="true" and checkin="true".
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
GetPendingIssues did several bad things:
~ select * on a 4 table join,
~ including multiple namespace collisions,
~ including large fields marc and marcxml from biblioitems,
~ return ($count, \@array_being_counted).
Not everything is fixed here (see FIXMEs), but the situation is
improved considerably, with bug 2900 resolved. The "timestamp"
namespace collision in query should be resolved by separate patch.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
This affects transactions on items that may be covered by
a title or item-level hold, but have not yet been retrieved
from the stacks (i.e. "confirmed") by the librarian. This does
not affect waiting holds (found="W"), so they will still block
transaction unless they belong to the operating patron.
Also I cleanup any hold the patron has on a biblio/item when
they are allowed to checkout the item.
Here we are filling the hold because the checkout is allowed, regardless
of of the queue position. This is different from AddIssue's CancelReserve, that
only fills the hold if it is next in line, but in the future AddIssue should
adopt a similar logic.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Allow valid comparison of hold_queue to current user barcode and
permit checkout if the current user is at the front of the queue.
Effectively, this allows a user to checkout a book he has held.
Here are the example checkout (11) and checkout response (12)
statements in a SIP telnet session, showing failure (120) previously
and success (121) after patch application.
Before:
11YN20060329 203000 AOCPL|AA1|AB502326000820|ACterm1
120NUN20081009 140222AOCPL|AA1|AB502326000820|AJKidnapped :|AH|AF2008-10-09 : Koha Admin (1)|BLY|
After:
11YN20060329 203000 AOCPL|AA1|AB502326000820|ACterm1
121NNY20081009 150204AOCPL|AA1|AB502326000820|AJKidnapped :|AH2008-10-10|AFItem was reserved for you.|
This patch also resolves security/privacy issues related to the display
of "needsconfirmation" values that identify, for example, the user
currently issued an item, or with a hold on the item. These messages
should not be passed to an end-user interface.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
SIP actually relied on the AddIssue return that was not reliable.
AddRenew also updated to return C4::Dates object for datedue.
Please note, any running SIPServer will have to be restarted
*immediately* after applying this patch, because although Koha
C4 behaves as normal, the SIP server runs as a Net::Server with
components cached. Changes will not be applied until SIPServer
restarts, and so checkout actions may fail until then.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Following suggestion by Vincent Danjean for Debian
packaging, 0755 -> 0644 for non-executable
files.
Also removed shebang from a few modules in C4.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
SIP code uses UNIVERSAL modules that do not tolerate the circular deps in
C4, specifically, in Auth.pm. Changing the order so that UNIVERSAL follows
the other imports is important to avoiding failures like:
perl -w -e 'use UNIVERSAL qw(can); use C4::Auth;'
"MarkIssueReturned" is not exported by the C4::Circulation module
This patch should resolve the problem earlier seen with Shelves too.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
ILS::Patron is where most of the intelligence for SIP's representation
is lodged. Currently there is difficulty with C4::Members functions.
GetMemberDetails is required for the needed flags, but it returns
empty structure on bad barcodes, where it should be undef.
Signed-off-by: Joshua Ferraro <jmf@liblime.com>
The basic problem is that the SIP logic doesn't know where the
input is coming from. It might be a RAW socket, and it might
be telnet. If it is telnet, although the specs declare a
character set (from MS, unfortunately), they do not specify a telnet
implementation. So you might get telnet handshaking or
renegotiations in the middle of an otherwise peaceful session and
these should not be taken as SIP commands. Patches include a move
towards using $CRLF from Socket to avoid problems w/ foreign platform
mapping \n and \r to \015 or \012.
Signed-off-by: Joshua Ferraro <jmf@liblime.com>
* Move CheckItemPreSave to C4::Items (from C4::Biblio)
* Modified C4::Biblio::AddBiblioAndItems to use appropriate
internal routines from C4::Items
* Moved GetItemnumberFromBarcode to C4::Items
* Removed duplicate C4::Biblio::_koha_new_items
* Removed disused C4::Biblio::MARCitemchange
Currently AddBiblioAndItems is a special routine that
uses private subs from both C4::Biblio and C4::Items.
This needs to be refactored.
Signed-off-by: Chris Cormack <crc@liblime.com>
Signed-off-by: Joshua Ferraro <jmf@liblime.com>