Bill Erickson [Tue, 27 Dec 2011 20:47:00 +0000 (15:47 -0500)]
Alter hold pickup lib from shelf avoids transit
When the pickup library for a hold is changed while the hold is on the
shelf, we no longer put the copy into transit to the new location.
Instead, we clear the shelf_expire_time (to prevent premature
expiration) and leave the hold and copy otherwise untouched.
This makes it possible for staff to identify items that are sitting on
the holds shelf that need to be put into transit to a new pickup
library. The transit will be created when the item is checked in at the
originating library.
Signed-off-by: Bill Erickson <berick@esilibrary.com>
Bill Erickson [Wed, 28 Dec 2011 16:17:37 +0000 (11:17 -0500)]
New hold status for "Wrong Shelf"
Adds a new hold status code (8) to indicate when a hold is sitting on
the wrong holds shelf.
In the OPAC, a hold status of 8 is displayed as in-transit. Neither
waiting-for-copy or in-transit are entirely accurate. In-transit makes
slightly more sense, since the copy is captured and (slowly) making its
way to the correct shelf.
Signed-off-by: Bill Erickson <berick@esilibrary.com>
Bill Erickson [Tue, 27 Dec 2011 20:55:33 +0000 (15:55 -0500)]
Hold current_shelf_lib for shelf browse API
Use the new current_shelf_lib column on hold requests to determine which
copies are on a given shelf. This also removes the shelf_time filter
since shelf_time is guaranteed to be set if current_shelf_lib is set.
Signed-off-by: Bill Erickson <berick@esilibrary.com>
Bill Erickson [Tue, 27 Dec 2011 18:18:08 +0000 (13:18 -0500)]
Hold current_shelf_lib API for setting/clearing.
Sets and clears the hold.current_shelf_lib value as appropriate during
hold capture, transit checkin, hold resetting, etc.
Also, we now create hold transits (instead of regular transits) when a
captured copy is (unexpectedly) found at the wrong branch and needs to
transit to the pickup lib. This can happen if the captured copy simply
transits to the wrong place (existing behavior) or the pickup lib for
the hold was changed while the copy was on the shelf (new behavior).
Signed-off-by: Bill Erickson <berick@esilibrary.com>
Bill Erickson [Tue, 27 Dec 2011 17:19:05 +0000 (12:19 -0500)]
Hold current_shelf_lib DB and IDL
Schema, upgrade, and IDL pieces for a new
action.hold_request.current_shelf_lib column. The goal of the column is
to track the location where a hold is currently sitting on a hold shelf.
Typically, this will be the pickup library, however, if the pickup
library has changed while the hold was on the shelf, the current shelf
lib will be different.
The upgrade script sets current_shelf_lib to match the pickup_lib for
all holds that are currently active and on the shelf.
Signed-off-by: Bill Erickson <berick@esilibrary.com>
Dan Scott [Tue, 22 Nov 2011 16:55:45 +0000 (11:55 -0500)]
Avoid updating deleted call numbers for 2.0 -> 2.1 upgrade
25% of the call numbers on our production system are deleted, so
filtering out the deleted call numbers from the update in the 2.0 -> 2.1
upgrade saves a significant amount of time.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Mike Rylander <mrylander@gmail.com>
Uses the OILSProxyPermissions and OILSProxyLoginType settings, but prompts
are controlled by the Apache AuthName directive.
This module allows things like Directory Indexing and non-perl or otherwise
incompatible with the normal proxy module Response (or similar) handlers to
be password-protected.
It also supports the OpenILS::WWW::Proxy user/passwd CGI params to allow
an easier transition between the two.
For example:
<Location /authen_login/>
Options +Indexes
PerlOptions +GlobalRequest
PerlSetVar OILSProxyPermissions "VIEW_REPORT_OUTPUT"
PerlSendHeader On
AuthType Basic
AuthName "Notices Login"
PerlAuthenHandler OpenILS::WWW::Proxy::Authen
require valid-user
allow from all
</Location>
Signed-off-by: Thomas Berezansky <tsbere@mvlc.org> Signed-off-by: Mike Rylander <mrylander@gmail.com>
Bill Erickson [Tue, 3 Jan 2012 21:41:42 +0000 (16:41 -0500)]
Coded value map index normalizer
New index normalizer which maps a coded value from the record to the
display value configured in config.coded_value_map. The normalizer is
applied with one parameter, the ctype of the coded_value_map to use for
the mapping.
E.g config.metabib_field_index_norm_map:
id | field | norm | params | pos
---+-------+------+---------------+-----
57 | 31 | 17 | ["item_lang"] | -1
* pos = -1 causes this normalizer to be applied to facet_entry's
This is primarily useful for facets. For example, you could create a
metabib_field for language and mark it as a facet_field. Applying this
index would allow the mapped language names (e.g. "French" instead of
"fre") to display as facets.
Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Mike Rylander <mrylander@gmail.com>
Dan Scott [Tue, 27 Dec 2011 17:31:03 +0000 (12:31 -0500)]
TPAC: Uniform search lib / depth settings
The current TPAC code uses logic embedded within the templates to
determine which org unit and depth to apply to searches.
This commit attempts to standardize the logic within the new
OpenILS::WWWW::EGCatLoader::Util::_get_search_lib() function and sets
the result as a new context variable, "search_ou". The org unit is
determined as follows (first matching condition wins):
* specific CGI "loc" value
* logged-in user's preferred search library
* logged-in user's home org unit
* specific CGI "physical_loc" value
The depth is set as follows (first matching condition wins):
* specific CGI "depth" value
* depth of the chosen org_unit
As of this commit, a library that wishes to set the a default search
org unit for any given user should set the "physical_loc" CGI param;
after the user performs their first search, the "loc" CGI param will
propagate through most interfaces. When a user clicks "Your account log
in", the user's preferred search library or home org unit will be set as
the new search lib.
At any time after that, the user can select a different org unit from
the org unit selector to explicitly set the "loc" CGI param.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Jason Stephenson <jstephenson@mvlc.org>
Working towards parity with the JSPAC, give the user the ability to set
their preferred search library in account preferences via the org
selector. The value defaults to the user's home OU.
Also set a sane default of 10 search results per page in the absence of
any preference, as the default had been set to 5.
While we're in there, remove some of the inline styles and weird markup
in the user preference page, and refactor the O:W:EGCatLoader:Account
module to accept an easily-appended list of preferences to be set.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Jason Stephenson <jstephenson@mvlc.org>
Dan Scott [Wed, 28 Dec 2011 02:17:02 +0000 (21:17 -0500)]
TPAC: Set search_ou context in common loader
By setting the search_ou context in load_common() after checking for a
valid user session, we can catch the remaining pages in TPAC in which
the default search library was not being set properly.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Jason Stephenson <jstephenson@mvlc.org>
Dan Scott [Tue, 27 Dec 2011 17:31:03 +0000 (12:31 -0500)]
TPAC: Uniform search lib / depth settings
The current TPAC code uses logic embedded within the templates to
determine which org unit and depth to apply to searches.
This commit attempts to standardize the logic within the new
OpenILS::WWWW::EGCatLoader::Util::_get_search_lib() function and sets
the result as a new context variable, "search_ou". The org unit is
determined as follows (first matching condition wins):
* specific CGI "loc" value
* logged-in user's preferred search library
* logged-in user's home org unit
* specific CGI "physical_loc" value
The depth is set as follows (first matching condition wins):
* specific CGI "depth" value
* depth of the chosen org_unit
As of this commit, a library that wishes to set the a default search
org unit for any given user should set the "physical_loc" CGI param;
after the user performs their first search, the "loc" CGI param will
propagate through most interfaces. When a user clicks "Your account log
in", the user's preferred search library or home org unit will be set as
the new search lib.
At any time after that, the user can select a different org unit from
the org unit selector to explicitly set the "loc" CGI param.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Jason Stephenson <jstephenson@mvlc.org>
Reverse Copy Location picker labels via WS Setting
If the setting is set then place the name before the org unit. This allows
for typing of the start of the name without having to deal with the org
unit shortname "eating" the start of the label (and thus type-in), but
preserves the information the org unit provides.
As an example, instead of having to type:
CONS : St
To get close to the default "Stacks" in a long list you could just type
St
And likely get to, or close to, Stacks.
The option can be toggled from the Workstation Administration menu.
Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Conflicts:
Dan Scott [Fri, 23 Dec 2011 06:56:02 +0000 (01:56 -0500)]
TPAC: Display monograph parts
In search results and record details, display monograph parts if
associated with a copy. Uses unapi for search results and json_query for
record details. Only generates the "Part" column in the copy table if
that record actually has a monograph part; otherwise the column is not
generated.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Bill Erickson [Thu, 22 Dec 2011 15:13:03 +0000 (10:13 -0500)]
SVF in bib record summary: extract attrs on the server
This repairs a FIXME within the custom bib overlay code related to
parsing metabib.record_attr stores within the JS. Instead, now we use
an attr parsing utility function which compiles a regular hash so the
client no longer needs to parse it.
Signed-off-by: Bill Erickson <berick@esilibrary.com>
And then add something like this to server/skin/custom.js:
try {
netscape.security.PrivilegeManager.enablePrivilege("UniversalXPConnect");
var prefs = Components.classes['@mozilla.org/preferences-service;1'].getService(Components.interfaces['nsIPrefBranch']);
} catch(E) {
dump('Error in custom.js trying to set oils.bib_brief.*.dynamic_grid_replacement.data: ' + E + '\n');
}
Finally, re-ingest any records with tag 915d (you can simply edit the MARC via
the MARC Editor for a specific record, or if the ingest.reingest.force_on_same_marc
row in config.internal_flag is enabled and you're not doing this during library
hours, you can re-ingest all records with UPDATE biblio.record_entry SET marc = marc;)
Signed-off-by: Jason Etheridge <jason@esilibrary.com> Signed-off-by: Bill Erickson <berick@esilibrary.com>
Bill Erickson [Tue, 29 Nov 2011 21:54:52 +0000 (16:54 -0500)]
TPac; indicate search results previously checked out
When a user is logged in to the tpac and performs a search, indicate in
the results set when any of the result items were ever checked out by
the logged in user.
Items will only be tagged when the related org setting is enabled and
the user has opted-in to circ history tracking.
New org unit setting is "opac.search.tag_circulated_items" /
"Tag Circulated Items in Results"
In the search results, just below title/author/callnumber/copy-count, a
new line is displayed that says "I have checked out this item before"
with a purdy little checkmark.
Bill Erickson [Mon, 19 Dec 2011 17:57:08 +0000 (12:57 -0500)]
TPAC: Change title tooltip in results table to useful message
From dbs's original commit:
The tooltip that would display when hovering over the title of a result
in the results table was simply the URI-escaped version of the title
text - not particularly helpful. Change this to say "Display record
details" as an actual hint about what action will occur when the user
clicks on a title in the results list.
This modified commit changes the title to 'Display record details for
"<title>"'
This commit also introduces the HTML plugin as a new
loaded-by-default template plugin. Among other things, it's useful for
ensuring HTML attributes are safely encoded, moresoe than |html, which
does not escape quotes.
https://bugs.launchpad.net/evergreen/+bug/902979
Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Dan Scott <dan@coffeecode.net>
Dan Scott [Wed, 21 Dec 2011 17:58:46 +0000 (12:58 -0500)]
TPAC: Split result copy counts into a separate template
As Thomas Berezansky suggests, sites are likely to want to customize the
display of copy counts depending on whether they're consortial sharers,
how deep they want to display, whether they want to just show the
matching search scope... make it easy to customize by moving the
copy counts portion of the results table into a separate template.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Wed, 21 Dec 2011 06:06:06 +0000 (01:06 -0500)]
TPAC: Improve display of holds/copy counts on record details
Based on some suggestions from Thomas Berezansky in LP 906168:
* Change the "Copies" header to "Available copies"
* Simplify the copy language to "# copies at <library>" (we still need
the noun "copies" but "available" has been factored out)
* Create a "Holds" header and move the holds information under that
* Put the "Available copies" and "Holds" into inline-block elements
so that they can take less vertical space on a wide screen but
will wrap on a narrow screen.
* Fix the line spacing between LI elements
In addition, break up the lengthy summary TT2 file into a few smaller
parts. In particular, make a copy_counts.tt2 file as sites are very
likely to customize this to, in some cases, not show the consortial
counts, or in other cases to only show the consortial count and current
search library.
Also fix some horrible CSS issues in the header in which the "Place
hold" and "Add to list" buttons were above the title / image, as well as
remove some of the CSS manipulation to plain H2 elements that gave them
no top margin.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Mon, 19 Dec 2011 04:54:15 +0000 (23:54 -0500)]
TPAC: Display additional levels of copy counts
In line with the MassLNC TPAC suggestion, display the copy counts for
search results and record details not only at the consortial level, but
also at up to two additional levels of the org unit hierarchy (but no
deeper than the user has chosen to search).
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Bill Erickson [Thu, 6 Oct 2011 18:34:03 +0000 (14:34 -0400)]
TPac: holds placement on monographic parts
Ability to place holds on monographic parts. In the holds placement
form, if a record has parts, a parts selector will display in the form
allowing the user to optionally specificy a monographic part for the
hold.
If a record has no non-part copies, the user is required to select a
part, because, in such cases, the hold cannot be fulfilled without
selecting a part.
Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Jason Stephenson <jstephenson@mvlc.org>
Bill Erickson [Tue, 8 Nov 2011 15:56:14 +0000 (10:56 -0500)]
bookbag CSV gets bib attrs; A/T unapi method
* Added a bib record unapi retrieval utility method for action/trigger
templates.
* Updated the bookbag CSV template to include data for the "item_type"
record attribute both to have it and as an example of how the unapi
retrieval works.
Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Lebbeous Fogle-Weekley <lebbeous@esilibrary.com>
Dan Scott [Tue, 20 Dec 2011 21:19:08 +0000 (16:19 -0500)]
Add "Last circulation date" field to new rcl view
Some libraries will want the option of filtering out copies which have
truly never circulated, so offering up the bare "last circulation date"
field rather than coalescing it with "create date" will give report
writers the ability to use an IS NOT NULL clause.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Tue, 20 Dec 2011 16:33:33 +0000 (11:33 -0500)]
Escape the backslash to avoid warning in PostgreSQL 9.0-
In PostgreSQL 9.1, "standard_conforming_strings" is ON by default and
there's no warning - but previous versions of PostgreSQL will see the
backslash and throw a warning about how you haven't escaped the
backslash. Thanks to Thomas Berezansky for noticing.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Fri, 9 Dec 2011 17:38:55 +0000 (12:38 -0500)]
Add concerto.sql dataset with callnums, copies, and URIs
Useful for testing a fresh Evergreen database, the concerto.sql dataset
is an SQL file that adds 100 bib records and automatically generates
assorted call numbers and copies for 3 of the branches. 50 of the
records include unadorned 856 fields and 25 of them have located URIs
(25 scoped at SYS1, and 10 scoped at both SYS1 and BR3).
In the effort to make the one-shot load of sample bib / copy / call
number data broad and realistic, delete some of the entities, add
prefixes and suffixes to some of the call numbers, and include some
monograph parts and conjoined items.
Signed-off-by: Dan Scott <dan@coffeecode.net> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Tue, 20 Dec 2011 16:42:17 +0000 (11:42 -0500)]
Remove top-level templates_custom directory
This top-level directory and version of advanced.tt2 slipped into commit c9cdea869b5 and was noticed by Bill Erickson. It should not be here.
Send it into oblivion.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Bill Erickson <berick@esilibrary.com>
Dan Scott [Mon, 19 Dec 2011 17:55:10 +0000 (12:55 -0500)]
Add a .gitignore file suitable for master
"git status" is painful in a source directory in which you're built and
installed Evergreen. This commit adds a .gitignore file that masks most
of the files and directories that you don't want to even think about, in
the hopes that the default Evergreen development experience can be mo'
betta.
Variations on this could be backported to rel_2_1 and rel_2_0 but take
care as some of the files that used to be source files in older releases
are now build from *.in files via autoconf.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Mon, 19 Dec 2011 16:17:12 +0000 (11:17 -0500)]
Wrap upgrade script for CREATE_REPORT_TEMPLATE permission
Permissions to create report templates will automatically be added to
any group that currently has the ability to run reports as part of the
upgrade process; if you want to make use of the ability to prevent some
groups from creating ad-hoc reports, remove the CREATE_REPORT_TEMPLATE
permission from permission.grp_perm_map for that group.
Dan Scott [Tue, 22 Nov 2011 16:55:45 +0000 (11:55 -0500)]
Avoid updating deleted call numbers for 2.0 -> 2.1 upgrade
25% of the call numbers on our production system are deleted, so
filtering out the deleted call numbers from the update in the 2.0 -> 2.1
upgrade saves a significant amount of time.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Bill Erickson <berick@esilibrary.com>
James Fournie [Sun, 11 Dec 2011 00:49:01 +0000 (16:49 -0800)]
Fix Dewey call number sorting
Ported over Koha commit aef8358c - fix for Koha Bug 4265.
Further documented in Evergreen LP # 902667
Here's the description from the commit message by Magnus Enger:
C4::ClassSortRoutine::Dewey turns "306 Les" into "306_Les"
for items.cn_sort and MARC-field 952$6, which results in
"306.46 Les" being sorted before "306 Les" in the OPAC.
With this patch, "306 Les" is turned into "306_000000000000000_Les".
Signed-off-by: James Fournie <jfournie@sitka.bclibraries.ca> Signed-off-by: Dan Scott <dscott@laurentian.ca>
I lied when I said it includes Data::Dumper, so you don't have to.
It does include Data::Dumper, but if you want to use Data::Dumper
in your script, you still have to use Data::Dumper.
Signed-off-by: Jason Stephenson <jstephenson@mvlc.org> Signed-off-by: Bill Erickson <berick@esilibrary.com>
Dan Scott [Sat, 10 Dec 2011 15:43:47 +0000 (10:43 -0500)]
TPAC: Assume that templates are encoded in UTF-8
Per http://template-toolkit.org/docs/faq/index.html
under "Why do I get rubbish for my utf-8 templates?", Template::Toolkit
does not assume that templates are encoded in UTF-8, which runs counter
to current assumptions circa 2011.
Practically, this means that before this change any UTF8 characters
used for purposes such as Unicode art or otherwise would be garbled in
the generated Web page. This commit tells Template::Toolkit to assume
that the templates are UTF8 without forcing us to monkey with byte-order
markers.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Bill Erickson <berick@esilibrary.com>
bott [Wed, 14 Dec 2011 17:31:30 +0000 (12:31 -0500)]
Allow for the page offset when listing the "Result x of y" values.
The record details page did not show the page offset correctly, so
result 13 would show up as "Result 3 of ###", and result 23 would also
show up as "Result 3 of ###".
Signed-off-by: bott <bott@grpl.org> Signed-off-by: Dan Scott <dscott@laurentian.ca>
Bibs aren't required to have sources. Thus, adjust all appropriate locations
to support a default state of "no source bibs can have copies", without
erroring out.
Signed-off-by: Thomas Berezansky <tsbere@mvlc.org> Signed-off-by: Dan Scott <dscott@laurentian.ca>
Bill Erickson [Wed, 14 Dec 2011 18:36:05 +0000 (13:36 -0500)]
Process still-shelved, canceled holds in clear-shelf process
Holds that are canceled but still physically on the hold shelf are
processed like other holds during the clear-shelf process. They will be
grouped in the same manner and if a custom post-clear status is
configured, copies linked to already-canceled holds will get the custom
status. The only difference is already-canceled holds will not be
re-canceled during the clear-shelf process.
This change has the added benefit of leveraging another API call to
retrieve the shelf-expired holds instead of using its own custom
json-query.
See https://bugs.launchpad.net/evergreen/+bug/904332
Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Tue, 6 Dec 2011 07:16:16 +0000 (02:16 -0500)]
TPAC: Add prefixes and suffixes to call number displays
Call numbers can be associated with prefixes and suffixes meant for
display purposes; display these with the call numbers in search results
and record details.
Thanks to Lebbeous and Bill for the assist in sorting out the remaining
portion of the hairy json_query for the record display.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Bill Erickson [Fri, 9 Dec 2011 20:52:32 +0000 (15:52 -0500)]
Alert addresses for patron registration : UI
Add support for comparing user addresses to alert addresses. When an
address is found, the address in question is styled (the header row
turns red) and the configured alert message is shown along the
top-right, where other warnings appear.
Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Bill Erickson [Tue, 6 Dec 2011 21:09:16 +0000 (16:09 -0500)]
Alert addresses for patron registration : DB/IDL
DB and IDL components to support configuration of alert addresses. The
target use case for alert addresses is in the patron registration
interface. When a patron's address matches an alert address, the staff
should be notified.
Alert address fields are treated as regular expressions.
Signed-off-by: Bill Erickson <berick@esilibrary.com> Signed-off-by: Thomas Berezansky <tsbere@mvlc.org>
Dan Scott [Mon, 12 Dec 2011 05:29:47 +0000 (00:29 -0500)]
RPAC: Remove one formatting row from results
One of the table rows was being inserted into every result simply to
provide a border that is better handled via CSS on the existing useful
elements. Begone.
Dan Scott [Mon, 12 Dec 2011 05:06:39 +0000 (00:06 -0500)]
TPAC: results use floated DIVs instead of table layout
Rather than a table containing a single row for the rather gross purpose
of placing the facet sidebar on the left, a little bit of a divider
between the facets and the results, and the results on the right, just
use a floated div and a relatively positioned div for the two pertinent
chunks of content. Bonus: one touch of CSS moves the facet sidebar to
the right instead of the left.
Dan Scott [Mon, 12 Dec 2011 04:37:02 +0000 (23:37 -0500)]
TPAC: Record details - inline style to classes
Continue the march towards no inline styles in the record details page.
Includes a small tweak for electronic resource display so that the
unordered list gains a bit of space on the left-hand side from the image
location.
Dan Scott [Mon, 12 Dec 2011 00:51:10 +0000 (19:51 -0500)]
TPAC: Include Chilifresh conditionally
Rather than including Chilifresh JavaScript and table cells
unconditionally, only include the Chilifresh calls if an account has
been defined in the Apache environment.
Dan Scott [Sun, 11 Dec 2011 22:40:39 +0000 (17:40 -0500)]
TPAC: Fix styling in advanced searchbar
Per Peter Lux's comment, we needed to fix the styling in the advanced
search bar. Marking the DIV id with "search-box" makes it consistent
with the searchbar in other interfaces.
Peter Lux [Fri, 9 Dec 2011 21:12:00 +0000 (16:12 -0500)]
TPAC: Link to Basic Search from advanced page
The searchbar in basic search, results, record details offers an option
to get to advanced search - but advanced search doesn't offer a way to
get directly back to a basic search. Now it does; however, some CSS is
required to make it look consistent.
Signed-off-by: Peter Lux <plux@upei.ca> Signed-off-by: Dan Scott <dscott@laurentian.ca>
Art Rhyno [Fri, 9 Dec 2011 21:36:28 +0000 (16:36 -0500)]
Added CUFTS support in ResolverResolver
This is based on the CUFTS implementation for UPEI. This is
very closely modeled on the SFX implementation.
opensrf.xml.example has been modified to include a <resolver_type>
element for ResolverResolver; by default, SFX will be used and this
shouldn't break anything else. If you use CUFTS rather than SFX, you
can specify 'cufts'.
Signed-off-by: Art Rhyno <artrhyno@uwindsor.ca> Signed-off-by: Dan Scott <dscott@laurentian.ca>
Dan Scott [Fri, 9 Dec 2011 20:12:53 +0000 (15:12 -0500)]
TPAC: Fix record summary e-resource and copy display
Electronic resources were wrapped with an ugly black line and no
differentiation from the outstanding text. Give them a light background
and remove the black line. Also, use an unordered list if we have more
than one URI to display.
Move the copy summary below the "Copies" heading that immediately
follows it for a more logical grouping of elements.
Dan Scott [Fri, 9 Dec 2011 17:05:14 +0000 (12:05 -0500)]
TPAC: Remove more explicit widths from results page
Refactor the facet sidebar such that headers consist of a floated span
and a div with no specific width; absent a specific width, the button
will float over top of text that might expand due to translation.
Also remove the explicit table cell width for the results counter column
and move some inline styles over to CSS. More to be done in results
table but we are on our way.
Dan Scott [Fri, 9 Dec 2011 16:09:24 +0000 (11:09 -0500)]
TPAC: Use padding instead of explicit width for tabs
Explicit widths work well for a combination of a specific font sizes and
a known string; however, in the context of translation, strings may get
longer than the explicit width allows. Using padding gives us more
flexibility to accommodate different font sizes and variable length
strings; the uniformity of the tab size will be sacrificed but perhaps
that is not such a bad thing.
Serials: be less eager to offer season dropdown in holding code mini-wizard
The recent commit (ebce79a288b in master) to make the holding code
mini-wizard offer a dropdown for season offers it too often, sometimes
in completely inappropriate cases.
This commit gets it right, and incidentally fixes two totally broken
links elsewhere in alternate serials control view interfaces.
Signed-off-by: Lebbeous Fogle-Weekley <lebbeous@esilibrary.com> Signed-off-by: Dan Wells <dbw2@calvin.edu>
Dan Scott [Fri, 2 Dec 2011 04:01:24 +0000 (23:01 -0500)]
Propagate search scope through advanced search interfaces
If a user had a chosen search scope (via the CGI "loc" parameter), it
would be lost when switching to the advanced search interfaces. Now we
progagate the search scope through the advanced search interfaces.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Bill Erickson <berick@esilibrary.com>
Dan Scott [Fri, 25 Nov 2011 18:20:32 +0000 (13:20 -0500)]
LP893315: Do not return delected call numbers in holdings_xml
Per LP893315, deleted call numbers were obscuring the actual active URIs
associated with a given bib record in a given scope at the default limit
of 5 results per record in unapi.holdings_xml.
Filtering out the deleted call numbers from the results seems like a
good idea, as we're generally interested in only showing the active
results (at least as in-db unapi is currently used). This also addresses
the problem of not showing active URIs when there are active URIs to
display.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Bill Erickson <berick@esilibrary.com>
Dan Wells [Mon, 5 Dec 2011 16:44:41 +0000 (11:44 -0500)]
Code comments and tweaks for lost processing
This commit is intended to do the following:
* Fix case where MISSING status is not retained for items
being sent 'home' (as it was in previous versions)
* Be more explicit about when we unset the LOST status and when
we do not in checkin_handle_lost()
* Prevent update of the copy in checkin_handle_lost() unless we
actually change the status
* Restore log messages for special handling of LOST/MISSING
checkins away from 'home'
* Provide additional code comments to clarify intended behavior
Note: Given the current restructure, MISSING item behavior is no
longer affected by 'lost_immediately_available' setting. That
seemed more bug than feature, and can be implemented later
(perhaps as a separate setting) if required.
Signed-off-by: Dan Wells <dbw2@calvin.edu> Signed-off-by: Jason Stephenson <jstephenson@mvlc.org>
Jason Stephenson [Tue, 29 Nov 2011 14:22:29 +0000 (09:22 -0500)]
Check OILS_SETTING_LOST_IMMEDIATELY_AVAILABLE in checkin_handle_lost.
Check the value of OILS_SETTING_LOST_IMMEDIATELY_AVAILABLE before
changing the copy status in checkin_handle_lost. This makes the setting
actually do something.
Add a couple of blank lines to aid in readability.
Signed-off-by: Jason Stephenson <jstephenson@mvlc.org> Signed-off-by: Dan Wells <dbw2@calvin.edu>
Jason Stephenson [Thu, 17 Nov 2011 20:15:09 +0000 (15:15 -0500)]
Fix LP851000.
Remove about 14 lines from OpenILS/Application/Circ/Circulate.pm
in the checkin_handle_circ subroutine that check whether or not
the checkin happens at the copy's circ_lib and the value of the
circ.lost_immediately_available org unit setting.
This change causes the handle_lost subroutine to run regardless of
where the checkin takes place for a more consistent experience.
Whether or not billings are voided, overdues restored, etc. is still
determined by the copy's circ_lib.
Given how the hold targeter and transit code works, it does not appear
the the circ.lost_immediately_available setting could ever do what it
was intended to do. Given these changes to the code it is now unnecessary.
Signed-off-by: Jason Stephenson <jstephenson@mvlc.org> Signed-off-by: Dan Wells <dbw2@calvin.edu>
Dan Scott [Fri, 25 Nov 2011 17:30:24 +0000 (12:30 -0500)]
TPAC: Protect against missing 901c in legacy MARCXML
Bad MARCXML to which Evergreen has been unable to add a 901c happens -
particularly with legacy MARCXML. Currently, TPAC crashes hard with an
ugly error if a 901c is not found when attempting to display the record
in search results. In comparison, the JSPAC displays the record without
a problem.
This commit logs a warning and enables the TPAC to continue somewhat
gracefully (the problem record is simply skipped in the search results).
A more robust solution may be to teach unapi.bre to add the record ID to
the results and to have the TPAC pull from that, rather than relying on
the MARCXML itself to contain the record ID.
Signed-off-by: Dan Scott <dscott@laurentian.ca> Signed-off-by: Bill Erickson <berick@esilibrary.com>
Dan Scott [Fri, 2 Dec 2011 04:06:50 +0000 (23:06 -0500)]
LP881774: Missing schema on function in 2.0-2.1 upgrade
Although the search path is altered to include the evergreen schema at
the start of the upgrade, one of the functions in the evergreen schema
does not resolve when invoked without the explicit schema later in the
upgrade script. We address that superficial (but failure provoking / time
wasting) problem by qualifying the function with the evergreen schema in
this commit, but we should also find out why the altered search path
does not seem to be functioning in the upgrade script.