Skip to content

Releases: duke-libraries/ezid-client

v1.11.0

15 May 20:10
f52815c
Compare
Choose a tag to compare
  • Updates Datacite schema to v4.5 (#98)

v1.10.0

23 Apr 18:41
v1.10.0
b2d7531
Compare
Choose a tag to compare
Version 1.10.0

v1.9.4

18 Nov 14:57
85d3bf5
Compare
Choose a tag to compare
  • #95 - Updates GitHub workflows
  • Adds Dockerfile

v1.9.3

17 Nov 19:13
8bf8b24
Compare
Choose a tag to compare

#93 - Fixes tests failing under Ruby 3.0. Thanks, @ahamelers!

v1.9.2

17 Dec 13:30
dfcf7f4
Compare
Choose a tag to compare

v1.9.0

09 Apr 12:52
24cc925
Compare
Choose a tag to compare
  • Migrates CI from Travis to GitHub actions.
  • Adds code to catch unexpected HTML responses and to retry server error and unexpected response error 2 times before re-raising.

v1.8.0

29 Jun 02:17
cae2aca
Compare
Choose a tag to compare
  • Deprecated settings related to SSL, since SSL now required.
  • Added Datacite compatibility (contributed by @jgondron):
    • Added a new MetadataTransformDatacite class that transforms a hashes datacite.* keys to the xml required by the datacite api.
    • Added an additional step in Metadata#to_anvl. Before the metadata is converted to anvl, it will perform an additional transformation of the metadata hash to make it compatible with datacite. A future improvement might be to do IoC here so that we can inject requested transforms and ignore ones an app doesn't care about.
    • Added an inverse transformation method for transforming metadata from the datacite api to keys required by the Metadata class
    • Added an additional step in Metadata#replace to transform the remote datacite metadata into the correct format
    • Adding a default DOI identifier type even when one is not available in the existing metadata. I've found that if you add this, the api's will generate the identifier xml for you. This is mostly for the case when you want to mint and give it the metadata in one op, since you won't have the identifier yet.
    • Datacite chokes on newlines in the xml. Likely to do with how the anvl gets parsed. Changed the transform to remove all new lines from the xml and fixed specs/fixtures to match.
    • Also fixed some issues with the readme's testing env variables

v1.7.1

20 Apr 20:57
Compare
Choose a tag to compare

Patches Ezid::Configuration to set @use_ssl to true by default, since EZID will require SSL starting April 30, 2017. In practice, unless the default port of 443 is changed, the client would use SSL by default prior to this change.

In the future, the use_ssl configuration setting will probably be removed.

v1.7.0

20 Apr 20:40
Compare
Choose a tag to compare
  • Ezid::Identifier#metadata is now merged local and remote metadata.
  • Adds Ezid::Identifier.load(id, metadata)
  • Ezid::BatchEnumerator renamed to Ezid::Batch without API change.

v1.6.0

28 Feb 20:38
Compare
Choose a tag to compare
  • NEW: Ezid::BatchEnumerator. Contructor receives a format (only :anvl currently supported) and file path to batch download (uncompressed). #each method yields Ezid::Identifier instance for each record in the batch.
  • Adds compression option to Ezid::BatchDownload (closes #74).
  • Ezid::BatchDownload#download_file returns file path (closes #73).