2022.5.29_1

Full Changelog: 2022.5.27_1...2022.5.29_1

2022.5.27_1

Full Changelog: 2022.5.26_3...2022.5.27_1

2022.5.26_3

Full Changelog: 2022.5.26_2...2022.5.26_3

2022.5.26_2

Full Changelog: 2022.5.26_1...2022.5.26_2

2022.5.26_1

Full Changelog: 2022.5.25_1...2022.5.26_1

2022.5.25_1

Full Changelog: 2022.5.24_1...2022.5.25_1

2022.5.24_1

What's Changed

• tests: add tests for query utils by @stefanomunarini in #184
• add tests GHA + fix failing tests by @stefanomunarini in #194

Full Changelog: 2022.5.13_1...2022.5.24_1

2022.5.13_1

There were many changes since the last featured release.

Bibliographic item display:

• Item’s type is not shown anymore, doctype is shown instead if the latter is provided in source data.
• Document identifiers (docid entries) do not display type, as it’s mostly useless and confusing to visitors. Primary document identifiers are shown in bold.
• An error during bibliographic item export could result in a weird situation where exported file contains error message. Now, in case of an error, nothing is downloaded and the message is shown in site GUI after a redirect.
• If Relaton field contains a formatted string with format specified as text/html, HTML markup now has expected effect and does not get escaped. (We assume Relaton data sources are trusted.)
• If source data has issues (validation errors), an exclamation mark is shown in the sourcing block.
• External links (Relaton’s link field) are now clickable, if they are URLs.

Datatracker integration:

• Site visitor who logs in with Datatracker is redirected back to the page they were at, rather than to landing page.
• User data (namely, name) is now fetched and shown in the logged-in block correctly.
• Bibliographic item export links are replaced with Datatracker login prompt, if user is not authenticated.
• Fetching latest I-D version metadata will no longer fail if Datatracker API returns no submission elements (ietf-ribose@a09d3be).
• OIDC now uses auth.ietf.org instead of datatracker.ietf.org.

xml2rfc-style path behavior:

• When resolving bibxml2 or bibxml-ids paths, Datatracker API errors no longer cause HTTP 500 responses.
• Fixed automatic resolution logic of some paths within rfcsubseries, ieee, nist datasets.
• Fixed versioned & unversioned path patterns for Internet Drafts; using mismatching pattern now results in HTTP 404.
• anchor can now be provided as GET parameter, and will cause the anchor attribute on the top-level element (be that <reference> or <referencegroup> to be set appropriately.

BibXML export:

• Contributors with publisher role are now serialized as authors.
• <reference> elements within a <referencegroup> now also have anchor attributes.
• target attribute is now correct, and is provided on <referencegroup> elements as well.

Other changes:

• Service version in the header now links to source repository, if environment is configured correctly.
• A summary of service’s purpose on landing page and an About page with more detailed overview were added.
• json_repr is now a valid search query format (though it is not exposed in the GUI).

For operations and development:

• Automatically provisioned Grafana dashboards now use better queries for more sensible bibitem/home page hit graphs.
• Documentation, including OpenAPI specifications, had been significantly corrected and expanded.
• DisallowedHost events are no longer logged.
• bibxml-data-* repositories are no longer indexed, saving indexing time. XML is now generated on-the-fly.
• Relaton data model and serialization capabilities were moved into a separate package (relaton on PyPI).
• Test coverage vastly improved.

2022.2.16_5

The following changes and improvements were made since the last published release:

• Fetching Internet Drafts from Datatracker has been implemented.
• External sources were generalized and made pluggable. Both Crossref and Datatracker integrations are now registered as external sources for their respective document types.
• xml2rfc path resolution for Internet Drafts (bibxml3, bibxml-ids) was improved.
  • It now makes use of Datatracker as external source. Document metadata obtained from Datatracker is compared with what was obtained from indexed sources, and if Datatracker’s revision is newer (or indexed sources have nothing) then Datatracker’s response is used. (The extra requests make bibxml3 xml2rfc paths slightly slower.)
  • It now correctly handles variants without the draft- prefix.
• Bibliographic item retrieval is now more predictable.
  • Bibliographic items that represent the same document are more reliably grouped (now this uses primary identifiers provided by Relaton), both in search results and individual details page.
  • Bibliographic items in search results are now ordered by latest date (e.g., revision or publication), from newest to oldest.
  • API behavior in cases where multiple bibliographic items were found representing requested document now uses the latest document, instead of merged result (which returned a structure of questionable validity given service’s use case, this affected for example Internet Drafts).
• Source indexing performance was improved by avoiding retrieving repository history when cloning.
• xml2rfc path handlers no longer increment access metric when used by internal xml2rfc path resolution checker tool.
• RFP requirement coverage section has been expanded, and parts of documentation now cross-reference respective RFP requirements.

2022.2.11_7

This release implements a few improvements:

• Dockerfile has been updated so that HTML documentation is built together with the service, and made accessible via /static/docs/index.html. (In case of current dev instance, the URL is http://dev.bibxml.org/static/docs/index.html.)
• The bibliographic item details page shows export links for all registered export formats, rather than just bibxml.
• When clicking on a particular document ID in search results listing, the anchor will be used to automatically scroll to the bibliographic item with that identifier, if there are multiple items shown on the page.
• Management section that lists available indexable datasets shows progress information without having to navigate to individual dataset’s page.