build(deps): bump org.apache.commons:commons-configuration2 from 2.9.0 to 2.10.1 #518

rest.cors.allowed-origins allows us to develop on Angular locally while using a remote DSpace API. We will perhaps disable this in the long run after we deploy DSpace 7.

I ported these from DSpace 6 but they are aesthetically annoying in DSpace 7 when you have more than ~4.

These were copied directly from our DSpace 6 branch.

These are the latest spider agent lists from our DSpace 6.x branch.

Sync settings with DSpace 6. We do not want to log bot requests, and we want to match user agents case insensitively.

Use a "2x" supersample by overriding the ImageMagick density from 72 to 144. This creates higher-quality PDF thumbnails.

We only want to send subscription mails for new items, not modified ones.

These behave differently in DSpace 7 than in DSpace 6 when the IDs are missing (you get infinitely repeating values in the submission form dropdown).

Our subjects are in dcterms.subject and we don't use hierarchical subjects. Also, change the date issued field.

Use dcterms.type for Type and add a few more filters from CGSpace.

Update from latest COUNTER-Robots list, removing from the local ILRI overrides where needed.

It has been ten years since we last increased the thumbnail size on CGSpace. 300px has served us well, but higher-density displays and new uses for thumbnails call for larger sizes. For reference, I checked 300 - 600px to see the implications on file size for one portrait layout PDF: - 212x300px: 31 kilobytes <--- current - 283x400px: 52 kilobytes - 354x500px: 76 kilobytes - 424x600px: 106 kilobytes I think 500px is a good compromise.

We don't use Datacite yet, but this is easy to change now so it can possibly work in the future.

We are not using this yet, but it will surely come in handy.

Copied directly from CGSpace's DSpace 6 repository.

Hopefully easier to maintain during future rebases now.

Just to show that it's possible and get feedback from our submitters.

Update a few Discovery configs to use dcterms.issued instead of dc.date.issued.

I was surprised to see this was disabled, as I already tested it a few months ago.

The DC=ILRI no longer exists in Active Directory as far as I can see from checking a few users.

There seems to be a bug in DSpace 7.6-SNAPSHOT that causes LDAP lo- gins to fail when the groupmap is enabled.

We are using an external controlled vocabulary instead. This hasn't been needed since we got rid of Atmire's Listings and Reports module in DSpace 6.

Change a few more references from dc.subject to dcterms.subject.

Replace a few more references to dc.description.abstract, since we are using dcterms.abstract.

Remove this so that DSpace doesn't automatically try to create a browse option for it. This is a side effect of a new browse feature in DSpace 7.6, where vocabularies in the submission form are auto- matically configured as browses.

The implementation is awkward, but seems to work if we list all our fields one by one. See: https://wiki.lyrasis.org/display/DSDOC7x/Configuration+Reference

Port of DSpace 6.x patch to legacy REST API in DSpace 7. See: DSpace#8343

We should use the same capitalization that we use elsewhere so it is more predictable and consistent.

For any fields that have more than a dozen or so values and don't need free text, we can use the tag input type.

The tag input type is currently buggy, and in any case we should only be using it for lists that have tens or more values.

It's too buggy.

Fields using external vocabularies must be onebox.

Re-order and add CGIAR Impact Platform.

I'm not using these anymore, as I've worked around them by creating a custom component in Angular that links to the Discovery search.

I always run this manually after doing migrations.

We were not using this in DSpace previously, and DSpace 7 has now changed the default to enabled.

Allow super administrators to impersonate other users.

Port the MODS 3.7 crosswalk from our DSpace 6 implementation.

These crosswalks are used to insert metadata into item pages for search bots like Google to consume. I am not sure whether the XHTML crosswalk is actually being used anymore...

This reverts commit 8ed657958dbab884daea84d537553ad17ef5e620. The bug was fixed. See: DSpace#9152

mwood on Slack suggested this for reducing the unhelpful and noisy logs like: 2023-11-01 00:00:07,581 INFO a2151e44-c888-4fbc-a54e-b71edeb5cc45 1b74f2ee-d4cc-473b-962c-65e6f43c9544 org.dspace.app.rest.utils.DSpaceAPIRequestLoggingFilter @ Before request [GET /server/api/core/items/f5e081f4-1c45-4dbf-8cac-2126119f87e5] originated from /

Make sure that we have both ascending and descending whenever there is a date available for sorting. Also, change the default sort order for sortDateIssued to asc as there is already sortDateIssuedDesc. See: DSpace#9104

FAO highlighted that the markup for alternative titles is slightly different. See: https://www.loc.gov/standards/mods/userguide/titleinfo.html

Clarify that these are not just for CGIAR authors, but recommended.

It is not repeatable in our current production DSpace 6 instance so I don't know why it is repeatable here.

This does not need to be repeatable, and indeed is not in our prod- uction DSpace 6 instance.

Because DSpace applies these case-insensitively, this means that we excluded real hits from Mozilla user agents for the past two weeks.

Requested by IFPRI.

From the cgspace-java-helpers.

Copied from the 6_x-prod branch.

It is not helpful at all because the values are uncontrolled.

Sorted in vim with `sort i` and formatted with tidy: $ tidy -xml -utf8 -m -iq -w 0 dspace/config/controlled-vocabularies/cg-creator-identifier.xml

This solves an issue when a user belongs to too many groups, which causes the Solr query to have too many "OR" clauses when searching for communities and collections they have access to. We only need the increased value in the search core, but one comment notes that this setting modifies an internal Lucene variable globally, so it can get reset depending on the order cores are loaded in. The error in the DSpace log is: too many boolean clauses.

This is based on the current top ~100 publishers in CGSpace in addition to all CGIAR centers, Initiatives, and Impact platforms.

Top 1,000 or so existing affiliations on CGSpace, plus all current Initiatives and Impact Platforms.

The Fruit and Vegetables Initiative started using their long name on 2023-08. I updated it everywhere but here!

Author requested we update his name format on CGSpace.

Python script interacting with the database can be read only when they are running in dry run mode.

Sorted in vim with `sort i` and formatted with tidy: $ tidy -xml -utf8 -m -iq -w 0 dspace/config/controlled-vocabularies/cg-creator-identifier.xml

This got lost in the transition to DSpace 7.

Put Initiatives and Impact Areas on the same line.

Move English, French, and Spanish to the top of the controlled list in the submission form.

Remove some incorrect ones based on clustering in OpenRefine and checking their profiles on orcid.org.

For now only IFPRI is using these.

Use type-bind to only show ISSN for journal articles and data papers.

Minor tweak to language.

We decided to re-work the way we tag datasets: - If a dataset has a DOI, it should be entered in the DOI field - Otherwise, it should be entered in the normal URL field - The "cg.identifier.dataurl" field will be for "related" datasets

IFPRI sent me a list of their authors so I combined it with ours.

There is a bug with onebox fields that use controlled vocabularies and are not repeatable.

Remove a few incorrect and duplicate affiliations.

I don't know why these are here, I fixed them twice already grrr.

The name has been United Nations Children's Fund since 1953. :) See: https://www.unicef.org/about-unicef/frequently-asked-questions#3

Use a list comprehension to filter out rows that are missing a url or filename.

IWMI sent me some new ones so I updated our list. This time I used GNU `sort` instead of `sort i` in vim because GNU sort handles ac- cents better. Formatted with tidy: $ tidy -xml -utf8 -m -iq -w 0 dspace/config/controlled-vocabularies/cg-creator-identifier.xml

Bumps org.apache.commons:commons-configuration2 from 2.9.0 to 2.10.1. --- updated-dependencies: - dependency-name: org.apache.commons:commons-configuration2 dependency-type: direct:production ... Signed-off-by: dependabot[bot] <[email protected]>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

build(deps): bump org.apache.commons:commons-configuration2 from 2.9.0 to 2.10.1 #518

build(deps): bump org.apache.commons:commons-configuration2 from 2.9.0 to 2.10.1 #518

Commits on Mar 7, 2024

Commits on Mar 8, 2024

Commits on Mar 20, 2024

Commits on Mar 21, 2024

build(deps): bump org.apache.commons:commons-configuration2 from 2.9.0 to 2.10.1 #518

Are you sure you want to change the base?

build(deps): bump org.apache.commons:commons-configuration2 from 2.9.0 to 2.10.1 #518

Commits on Mar 7, 2024

Commits on Mar 8, 2024

Commits on Mar 20, 2024

Commits on Mar 21, 2024