this patch change the attachment encoding, else if the file is too large, koha throws an error.
This solution was found on koha list in a Joshua's e-mail
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
This patch add the form for sending shelf by e-mail, and add a button in shelves that allow the user to send a shelf.
It's an approximate copy of the send basket feature.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
The default syspref setting for intranetuserjs should be empty. At one
time, an unwanted value accidently got put in there. This removes the
syspref only if it equals the unwanted value.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Modified the borrower attribute and borrower password fields to support a maxlength of 64 characters.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
This patch change the linktext as the label of the link(in 856$z), and set the url from the 856$u.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
An ending </div> was not placed correctly and encompassed both
privateshelves and publicshelves
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
If a value being substituted in was NULL, then warnings like this
would be emailed to the crontab owner (or MAILTO):
Use of uninitialized value in substitution (s///) at /home/user/kohaclone/C4/Letters.pm line 508.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
* more portable way to determine uname for
crash messages
* C4::Context->userenv now returns undef instead
of 0 if no userenv has been set yet; returning
undef when necessary is almost always better
than returning a scalar when a hashref is
normally expected.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Refactoring KOHAVERSION to be less crufty, enabling use warnings, change formatting slightly in places.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Previous patch broke date comparison that sets
overdue flag - now that ISO-formatted date
strings are being compared, must use "it"
instead of "<".
"2009-01-02" lt "2009-01-08" is true
"2009-01-02" < "2009-01-08" is *false*
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 ($count, \@array_being_counted).
Not everything is fixed here (see FIXMEs), but the situation is
improved considerably, with bug 2900 resolved. The "timestamp"
namespace collision in query should be resolved by separate patch.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
despite the same values being supplied each time. Then
the conditional assignments would repeat the same calls again!
That means execution was liable to query the DB at least once
and as many as four times per item. With a large number of items
this is an unnecessary burden. By moving the calls outside
the loop, we can guarantee that we never have to call the DB for
that info more than twice (once for lost, once for damaged).
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
This adds a little javascript to the Patron Selection form input after searching for a patron to check out to.
On click it redirects to checking out to the clicked patron.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
t/Dates.t is essentially restored from its previous state with
the revision that it now does not use C4::Context or check syspref
for the default date format. Instead it sets the C4::Dates default
directly, taking cue from command line argument or ENV. ISO format
revised to accept "T" separator and "Z" (zulu) terminator. POD
expanded and corrected.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Moving towards being able to enable warnings. Use one dbh, and stop
redeclaring it in each conditional chunk. ($sth still need to be cleaned.)
Toggle in script removed (tmpl should use loop context var __odd__).
$sth->finish calls removed where unnecessary.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
I've duplicated the "Add to Order" block on the search results page to display after a user has searched for an existing record to add to an order. This block is displayed whether or not results were found on the assumption that a non-empty result set may not contain the desired result. Changes to neworderbiblio.pl allow supplier name to appear in breadcrumb nav.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Change made to GetMarcItem to quell a warning
changed sense of a test; defined($foo) is *not*
the same as $foo ne ''.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
When cloning the set of inputs, the new js
increments the IDs of each form field (to keep them unique) and automatically
selects the option that was selected in the cloned group.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
CalcNextLabelPos pulled out of each subblock.
Print of header and initial structure deferred until later to allow fatals_to_browser to
display any fatal errors encountered in data preparation. This does not fix Unicode "wide character" crashes.
Created subroutine for debug dump lines.
Comparisons corrected (numerical not string compare).
Note: this script is still insecure with NO AUTH CHECK.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
I added 'use warnings' to C4::Biblio and made a handful of changes to
reduce the number of warnings emitted.
One notable spot is the change in the regex in
C4::Biblio::GetNoZebraIndexes. I have replaced the parens with a character
class. The parens change the way 'split' works, making it return elements
for each delimiter. We did not want those elements returned, and they
only resulted in "'' => undef" being added to the final hash. They also
resulted in two "undefined" warnings for each pass through the loop. I've
included a simple test for this function.
There may be a few warnings still emitted in spots, but either I haven't
seen them yet, or I have chosen to not fix them yet because they require
too much change.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
This patch amends the function to return barcode, in particular when
filter is not defined. It also adds an optional 2nd argument to
allow the filter to be specified by caller, enabling testing.
Non-DB-dependent test script included. Note: T-prefix style
barcode filter is not documented, and drops the first nonzero
digit after the T. This seems mistaken, but is not corrected here
to avoid any surprises.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
Consolidated error catching after evals.
Removed unnecessary $sth->finish calls and some unused variables.
Pulled query for itemtype mappings outside DrawSpineText and added
a class level caching variable to eliminate repeated queries for
*each piece of text* on *each label*! This was a major performance
downside.
Note: this does not fix Unicode problems, but it does add some notes
on unsuccessful attempted workaround using utf8::encode.
C4::Labels should likely be broken up to separate out the pieces that
do not touch the database (wrappers of PDF::Reuse) and those that are
CRUD API for table data.
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
there is some old 2.2 code for more complete labels-searching, thats been commented out
ive logged a 'bug' for this - 2777
http://bugs.koha.org/cgi-bin/bugzilla/show_bug.cgi?id=2777
this fix...
- enables the index-searching functionality
- adds date-ranges
- uses the numSearchResults syspref, not a hardcoded '20'
- rewrote old 2.2 page-num code, (didnt work for 3.x)
- some indent and whitepages tidys.
- unused 2.2 search-code removed.
Mason
Signed-off-by: Galen Charlton <galen.charlton@liblime.com>
