This project forked from alanorth/dspace
Fork of the official DSpace repository. DSpace powers the CGIAR outputs repository (CGSpace). This repository contains local modifications to the DSpace source code.
Home Page: https://cgspace.cgiar.org
License: Other
For example, ICARDA's cg.subject.icarda has only been populated by batch import and doesn't have a corresponding value-pairs entry in the input forms. This means we can't use it for Listings and Reports (and probably other things!), and there are probably other places where this causes ill effects.
Other examples might be: cg.subject.cip and cg.subject.drylands...
Basic theme done by @swebsheet. Still needs:
Our Linode VPSes have IPv6, but we don't have AAAA records, so IPv6 clients have to use IPv4 instead.
New banner for CIAT community.
We need plural and lower case for second words (except proper nouns), eg:
Also:
This is possibly related to #32, but because the fields are used elsewhere, it might be as simple as adjusting the messages.xml and all occurrences will print correctly.
Sisay had added some but they were based on DSpace 3.x code, and they included a lot of "kitchen sink" changes that had other side effects. We need to do new ones based on the DSpace 4.x dspace/config/spring/api/ discovery files.
We need to be able to store, display, browse, and search by ICARDA subject.
dspace.cfg)?discovery.xmldiscovery.xmlmessages.xmldspace/config/modules/atmire-listings-and-reports.cfgdspace/config/modules/atmire-cua.cfgitem-list.xsl and item-view.xslAfter adding to dspace.cfg you need to re-index, but I think you also need to re-index after modifying discovery.xml. I believe both are handled by dspace index-discovery -b... I suspect we only need to add browse indexes to dspace.cfg, as they are used for "browse by blah", but search indexes there used to be Lucene, and are now handled in discovery.xml.
We are going to be adding lots of new facets to the sidebar and the login box is nearly at the bottom, so let's just stick it at the top.
See #64 for location of sidebar config.
https://github.com/ilri/DSpace/blob/4_x-dev/pom.xml#L38
You can change this into:
<cua.version.number>4.2-3.1.4</cua.version.number>
This version bump will address Peter's request regarding contextualized representation of the most popular pages.
As a follow up to #28, we need to add more sidebar facets for:
Etc etc...
Check which of these are already configured as a searchFilter in dspace/config/spring/api/discovery.xml, then add them as sidebarFacets, ie:
<ref bean="searchFilterCrpsubject" />
Some of our older custom metadata fields are living in the DC namespace, which is ugly, confusing, and just makes me lose sleep at night. For example:
dc.isubject.ilrisubjectdc.crsubject.crpsubjectWe learned to behave better, and started prefixing custom fields with "cg", for example:
cg.subject.ciatcg.subject.cipWe should find all instances of these and kill them one by one, including in dspace.cfg, input-forms, XMLUI themes, spreadsheet templates by the content staff... etc.
As a subset of #32, we need to add these to the Discovery filters:
Allows you to see things like tweets, news articles, etc about your items. See it in action on the sidebar here:
Necessary code merged in DSpace 5.0.
We're somewhere near 50,000 right now... and importing stuff like crazy. Someone brought up a question about what happens when we reach 99999, and I didn't know the answer. I'll have to ask the DSpace Test mailing list.
Some users reported that the search results in DSpace 4.2 display strange characters. For example, searching for "ilri research brief" displays results such as (attached).
Item in the screen shot is: https://cgspace.cgiar.org/handle/10568/51393
I suspect these strange characters are present in the full-text search for that item, which begins:
ILRIRESEARCH
RESEARCHBRIEF
BRIEF N
32O.
I LRI
Date
year
November
2014
6
14
Growing food and feed with less environmental
impact: A dual-crop impact narrative
Did you know?
•
Right now we have both robots.txt and add_header X-Robots-Tag "none"; for DSpace Test... but I don't know which one is respected. Ideally, on DSpace Test we would block access in robots.txt, but we deploy CGSpace from the same code base, so that's tricky to have different settings.
Maybe we could use robots_test.txt, and redirect those in the nginx vhost for DSpace Test?
When using Listings and Reports to generate a bibliography, instead of downloading the results as a PDF / PS / RTF, it prints in the browser window. According to Chrome dev tools the content type is text/plain... but it should be application/pdf for PDF at least.
We need screen real estate on the side bar, so we want to hide the Browse By sidebar block. We can hide them in XMLUI but they will remain functional for when people click the subject terms in item views and lists. Related to #48.
Change the order of facets on the navigation sidebar to be of order most common to least common. Related to #32 for adding more Discovery facets.
We want to keep the same order in the homepage and the main Discovery config.
List of tasks for Drylands CRP subjects, aka cg.subject.drylands:
dspace.cfg)discovery.xmldiscovery.xmlmessages.xmldspace/config/modules/atmire-listings-and-reports.cfgdspace/config/modules/atmire-cua.cfgitem-list.xsl and item-view.xslAfter adding to dspace.cfg and discovery.xml you need to re-index, ie dspace index-discovery -b.
Need to add a new search filter for dc.title.jtitle in the Discovery config, dspace/spring/api/discovery.xml and then add it to the CUA config: dspace/modules/atmire-cua.cfg.
Unfortunately we'll need to re-index Solr after this because of adding a new search filter.
New discovery search filters need SimpleSearch XMLUI strings:
We want to clean up and consolidate the input forms so they are simpler and we don't have 10+ separate input forms with duplicate metadata fields.
There is a rough pull request in #81 which works, but needs one last look to clean up some typos ("IPASTORAL").
Basic theme done by @swebsheet, still needs:
We enabled it in #29 for crpsubject LIVESTOCK+AND+FISH, and it seemed to work... but it never worked for other CRPs, and the RSS feed seemed to stop updating eventually. Also, sometimes there are results in the RSS which shouldn't match!
The configuration for the sort_by parameter in dspace.cfg is currently:
webui.itemlist.sort-option.1 = title:dc.title:title
webui.itemlist.sort-option.2 = dateissued:dc.date.issued:date
webui.itemlist.sort-option.3 = dateaccessioned:dc.date.accessioned:date
webui.itemlist.sort-option.4 = type:dc.type.output:text
Need to double check:
websvc.opensearch.Remove nonsense of LDAP and just say "CGIAR users" or something. The string to translate is in dspace/modules/xmlui/src/main/webapp/i18n/messages.xml.
During research for #34 I found that controlled vocabularies allow you to split up input-forms.xml by keeping ILRI subjects, for example, in a separate file. The side effect is that this functionality implies a hierarchical taxonomy, like ILRI::Subjects::AGRICULTURE and is stored / displayed in the user interface as such! That requires us to change a lot of the places where we are printing / aggregating subjects in Discovery and XMLUI item lists...
But there is power in this approach! The hierarchy is obviously powerful, and sites like the World Bank's Open Knowledge use it, see okr.sector etc:
https://openknowledge.worldbank.org/handle/10986/21529?show=full
This, together with the proper Discovery config is very useful:
Item lists and displays should show ILRI subjects. Related to #68.
Code is in: https://github.com/alanorth/DSpace/tree/4_x-dev-tc
New Discovery sidebar facets are missing some XMLUI strings, which show up when you use the "View more" from the sidebar facet. So far I've seen:
xmlui.Discovery.AbstractSearch.type_status
xmlui.Discovery.AbstractSearch.type_outputtype
CCAFS is moving into Phase II and needs to update some of their metadata taxonomy. This requires:
input-forms.xml for the new terminologyIn the Livestock and Fish XMLUI community item lists display "Subject Focus" with the crpsubject instead of ilrisubject.
Some time during DSpace 1.7 or 1.8, we were seeing bizarre caching effects in XMLUI themes, like themes using CSS from one theme, but a banner from another. We initially suspected client-side browser caching but eventually found that the Cocoon pipeline was somehow messing up.
The fix at the time was to use a non-caching pipeline for XMLUI themes. It has been nearly two years since we made this fix, and we need to re-evaluate whether we still want this.
Since DSpace 4.0 the automatically generated thumbnails are much better quality than the manual ones we had been creating. For items where there is an auto-generated thumbnail we should remove the manual ones.
Related to #58 for CIAT, but now I've found some items in ILRI's theme which do this too. For example: https://cgspace.cgiar.org/handle/10568/59787.
We should probably just apply this across all XMLUI themes as it fixes the bug in Chrome and gives more proportionate space to the columns (20% / 80%).
We can use OpenSearch backend in DSpace to provide custom RSS feeds based on OpenSearch queries, ie for certain subject terms like CRPs. This is useful because we can feed them to FeedBurner to get alerts when new ones are added.
List of tasks for CIP subjects, aka cg.subject.cip:
dspace.cfg)discovery.xmldiscovery.xmlmessages.xmldspace/config/modules/atmire-listings-and-reports.cfgdspace/config/modules/atmire-cua.cfgitem-list.xsl and item-view.xslAfter adding to dspace.cfg and discovery.xml you need to re-index, ie dspace index-discovery -b.
I need to look at other XMLUI themes and see if this is just an issue with CIAT or what. Item in screenshot is: https://cgspace.cgiar.org/handle/10568/54615.
Our current nginx configuration bends over backwards to serve https://dspace.ilri.org and https://mahider.ilri.org.
We only need to handle those links for as long as they keep coming, so should actually be pro-active and de-register links containing those domains from search indexes (ie in Google Webmaster tools).
Will reduce round trip time for TLS sessions, see commit and comments here:
https://github.com/alanorth/ansible_blog/commit/6ccfdb99fa316f88c7e3e3278915e2e0c7e3b187
I didn't notice it before, but discovery searches like this expose new strings we need to translate: https://cgspace.cgiar.org/search-filter?field=ilrisubject
At least:
xmlui.Discovery.AbstractSearch.type_ilrisubjectFor some reason CGSpace doesn't have a repository identifier, but DSpace Test does! See:
Bring back the recent submissions on the homepage. They were disabled a few months ago because it was kinda messy, but it's the only place where you can see all CG centers working together, so we want to show the latest from the entire CGSpace.
It should be somewhere in dspace/config/spring/api/discovery.xml (look in git history for that file, you'll find it).
At the planning meeting in December we discussed the post-4.x roadmap. We agreed we'd try to start testing DSpace 5.x in March and hopefully deploy in June. No testing has happened yet.
We need to:
5_x-dev branch based on upstream's 5_x4_x-prod on top of 5_x-devInstead of old "Browse by" searches:
https://cgspace.cgiar.org/browse?type=ilrisubject
We should probably be using Discovery searches:
https://cgspace.cgiar.org/search-filter?field=ilrisubject
Is there anything "Browse by" can do that Discovery can't? Technically they are both using the same SOLR (discovery) backend.
The Drylands CRP doesn't have an XMLUI theme yet, but when it does, here is the information we need:
We fixed OAI support in #57, but now, testing CGSpace's OAI with the base-search validation tool fails with the following single fatal error:
ERROR: No incremental harvesting (day granularity) of ListRecords: Harvest for reference date 2015-02-14 returned no records. Please make sure that the from and until parameters are evaluated.
I've seen several messages on the dspace-tech mailing list about this, and will have to investigate their solutions. //cc @Jayakananthan
input-forms.xml is one big, ugly file right now. It has many templates for many institutes, which makes it hard to manage, and always conflicts with mainline DSpace during upgrades.
We need to be able to split the institution-specific controlled vocabularies into separate files.
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.