Changelog¶
v0.1.17¶
Updates regarding the bibxml
serializer:
When deciding whether to output
<referencegroup>
as root element, rely on presence of “includes”-type relations alone, rather than also require absence of titles.Fix an issue with serializing Internet Drafts to XML
v0.1.16¶
Allow multiple
BibliographicItem.version
entries
v0.1.15¶
Make
BibliographicItem.docid
a list (and required property)
v0.1.14¶
Corrected a problem with abstract conversion (
bibxml
serializer)
v0.1.11¶
Fixed CodeCov badge syntax in README/PyPI description.
v0.1.10¶
bibxml
serializer now takes abstract’sformat
into account. For abstracts withapplication/x-jats-xml
format (typically, coming from Crossref DOI data), paragraphs are now extracted as<t>
sub-elements.
v0.1.9¶
Added
prefix
andaddition
fields toPersonName
model.bibxml
serializer now attempts to fill infullname
attribute on an<author>
even ifPersonName.completename
is absent (formatting a string using whatever parts of the name are available).
Anchors and identifier scope¶
This starts a switch to
anchor
attribute values formatted based on primary identifiers, rather than using identifiers withscope
property set toanchor
(which will be phased out).So far this is only implemented for Internet Drafts (see
serializers.bibxml.anchor.format_internet_draft_anchor()
, whichget_suitable_anchor()
now delegates to if adocid
withtype
matching “internet-draft” is detected on the item.
v0.1.8¶
Added
BibliographicItem.version
field and the correspondingVersionInfo
class. It follows LutaML models in thatdraft
is expected to be either a string or not defined, not an array of strings like the RNC grammar suggests. This may be subject to change.
v0.1.7¶
Fixed re-imports.
v0.1.6¶
Added re-imports for models classes in
relaton.models
for convenience.Added changelog.