Skip to content

Fix improve oai pmh metadata quality for agris ap compliance #1425

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
gitnnolabs merged 4 commits into from
Apr 2, 2026
Merged

Fix improve oai pmh metadata quality for agris ap compliance #1425

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
b09a0da
fix(compile): fix OAI-PMH record template for AGRIS AP compliance
Rossi-Luciano Apr 1, 2026
959b4ca
fix(search_indexes): fix prepare_license and prepare_creator
Rossi-Luciano Apr 1, 2026
adc426f
fix(models): fix License.url to correctly reconstruct CC license URL
Rossi-Luciano Apr 1, 2026
261d196
Habilita o filtro
Rossi-Luciano Apr 1, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 16 additions & 9 deletions article/search_indexes.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -447,13 +447,18 @@ def prepare_titles(self, obj):
return set([title.plain_text for title in obj.titles.all()])

def prepare_creator(self, obj):
"""The list of authors is the contrib_persons on the models."""
if obj.contrib_persons.exists():
return set([
person.names
for person in obj.contrib_persons.all()
if person.names
])
if not obj.contrib_persons.exists():
return None
result = set()
for person in obj.contrib_persons.all():
if person.last_name:
suffix = f" {person.suffix}" if person.suffix else ""
given = f", {person.given_names}" if person.given_names else ""
result.add(f"{person.last_name}{suffix}{given}")
elif person.names:
# fallback: nome completo sem inversão
result.add(person.names)
return result or None

def prepare_collab(self, obj):
"""This is the instituional author."""
Expand Down Expand Up @@ -519,8 +524,9 @@ def prepare_identifier(self, obj):
return idents

def prepare_license(self, obj):
if obj.license and obj.license.license_type:
return [obj.license.license_type]
if obj.license and obj.license.url:
return [obj.license.url]
return None

def prepare_sources(self, obj):
# property no article.
Expand All @@ -535,4 +541,5 @@ def get_model(self):

def index_queryset(self, using=None):
return self.get_model().objects.filter(is_classic_public=True)
# return self.get_model().objects.all()

160 changes: 99 additions & 61 deletions article/templates/search/indexes/article/article_compile.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,95 +10,134 @@
<element name="none">
{% for title in object.titles.all %}
<field name="value">{{ title.plain_text }}</field>
{% endfor%}
{% endfor %}
</element>
</element>
<element name="creator">
<element name="none">

<!-- C1: creator via contrib_persons com afiliação -->
<element name="contributor">
<element name="author">
{% for person in object.contrib_persons.all %}
{% if person.orcid %}
<field name="value" orcid="{{person.orcid}}">{{person.names}}</field>
{% else %}
<field name="value">{{person.names}}</field>
{% if person.last_name %}
<field name="value" {% if person.orcid %}orcid="{{ person.orcid }}"{% endif %}>{{ person.last_name }}{% if person.suffix %} {{ person.suffix }}{% endif %}{% if person.given_names %}, {{ person.given_names }}{% endif %}</field>
{% elif person.names %}
<field name="value" {% if person.orcid %}orcid="{{ person.orcid }}"{% endif %}>{{ person.names }}</field>
{% endif %}
{% endfor %}
</element>
</element>

<element name="subject">
<element name="none">
{% for kwd in object.keywords.all %}
<field name="value">{{ kwd.text }}</field>
{% endfor %}
</element>
</element>

<!-- C2: abstract via plain_text em vez de __str__ -->
<element name="description">
<element name="none">
<element name="abstract">
{% for abs in object.abstracts.all %}
<field name="value">{{ abs }}</field>
<element>
<field name="value">{{ abs.plain_text }}</field>
</element>
{% endfor %}
<field name="value"/>
</element>
</element>
<element name="date">
<element name="none">
<field name="value">{{ object.updated|date:"c" }}</field>
</element>
</element>
<element name="type">
<element name="none">
<field name="value">{{ object.article_type }}</field>
</element>
</element>
<element name="identifier">
<element name="none">

{% for doi in object.doi.all %}
<field name="value">{{ doi.value }}</field>
{% endfor %}

<!-- C3: data de publicação em vez de updated -->
<element name="date">
<element name="issued">
<element>
<field name="value">{{ object.pub_date }}</field>
</element>
</element>
</element>

<element name="type">
<element name="none">
<field name="value">{{ object.article_type }}</field>
</element>
</element>

<!-- C4: identificadores com sub-elementos por tipo -->
<element name="identifier">
<element name="none">
<field name="value">{{ object.pid_v2 }}</field>
<field name="value">{{ object.pid_v3 }}</field>

</element>
{% for doi in object.doi.all %}
<element name="doi">
<element>
<field name="value">{{ doi.value }}</field>
</element>
</element>
{% endfor %}
<element name="uri">
{% for collection in object.collections %}
{% for lang in object.languages.all %}
<field name="value">http://{{collection.domain}}/scielo.php?script=sci_arttext&amp;pid={{object.pid_v2}}&amp;tlng={{lang.code2}}</field>
<element>
<field name="value">https://{{ collection.domain }}/scielo.php?script=sci_arttext&amp;pid={{ object.pid_v2 }}&amp;tlng={{ lang.code2 }}</field>
</element>
{% endfor %}
{% endfor %}

</element>
</element>
<element name="language">
<element name="none">
{% for lang in object.languages.all %}
<field name="value">{{ lang.code2 }}</field>
{% endfor %}
</element>
</element>
<element name="rights">
<element name="none">
{% for lic in object.license.all %}
<field name="value">{{ lic }}</field>
{% endfor %}
</element>
</element>
<element name="source">
<element name="none">
<field name="value">{{ object.source }}</field>
</element>
</element>

<element name="publisher">
<element name="none">
</element>
</element>

<element name="language">
<element name="iso">
{% for lang in object.languages.all %}
<element>
<field name="value">{{ lang.code2 }}</field>
</element>
{% endfor %}
</element>
</element>

<!-- C5: rights via ForeignKey, não .all() -->
<element name="rights">
<element name="none">
{% if object.license and object.license.url %}
<field name="value">{{ object.license.url }}</field>
{% endif %}
</element>
</element>

<!-- C6: source via journal -->
<element name="source">
<element name="none">
{% if object.journal %}
<field name="value">{{ object.journal.title }}{% if object.pub_date_year %}, {{ object.pub_date_year }}{% endif %}</field>
{% endif %}
</element>
</element>

<element name="publisher">
<element name="none">
{% for publisher_name in object.publisher_names %}
<field name="value">{{ publisher_name }}</field>
<field name="value">{{ publisher_name }}</field>
{% endfor %}
</element>
</element>
<element name="format">
<element name="none">
<field name="value">text/html</field>
</element>
</element>
</element>
</element>

<element name="format">
<element name="none">
<field name="value">text/html</field>
</element>
</element>

<element name="relation">
<element name="ispartof">
<element>
{% if object.journal %}
<field name="value">{{ object.journal.title }}</field>
{% endif %}
</element>
</element>
</element>

</element>
<element name="bundles"/>
<element name="others">
Expand All @@ -111,4 +150,3 @@
<field name="name"/>
</element>
</metadata>

36 changes: 22 additions & 14 deletions core/models.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -399,7 +399,7 @@ def get_or_create(
institution=institution,
user=user,
)

@classmethod
def create(cls, raw_text, initial_date=None, final_date=None, user=None, institution=None):
history = cls()
Expand Down Expand Up @@ -446,7 +446,7 @@ def data(self):
class OrganizationNameMixin(models.Model):
"""
Mixin that provides organization name and acronym fields.

Fields:
name: The organization's full name
acronym: The organization's acronym
Expand Down Expand Up @@ -483,7 +483,7 @@ def __str__(self):
class VisualIdentityMixin(models.Model):
"""
Mixin that provides visual identity fields for organizations.

Fields:
logo: The organization's logo image
url: The organization's website URL
Expand Down Expand Up @@ -645,10 +645,18 @@ def create_or_update(cls, user, license_type=None, version=None):
return cls.get(license_type=license_type, version=version)
except cls.DoesNotExist:
return cls.create(user, license_type, version)

@property
def url(self):
return f"https://creativecommons.org/licenses/{self.license_type}/{self.version or '4.0'}/"
"""Reconstrói a URL CC a partir do license_type no formato 'CC {CODE} {VERSION}'"""
if not self.license_type:
return None
parts = self.license_type.split()
if len(parts) == 3 and parts[0].upper() == "CC":
code = parts[1].lower() # "BY" → "by", "BY-NC" → "by-nc"
version = parts[2] # "4.0"
return f"https://creativecommons.org/licenses/{code}/{version}/"
return None

@property
def data(self):
Expand Down Expand Up @@ -743,7 +751,7 @@ def create_or_update(cls, user, license, language=None, url=None, license_p=None
def parse_url(url):
"""
Parse Creative Commons license URL.

Exemplos de URLs:
- https://creativecommons.org/licenses/by/4.0/
- https://creativecommons.org/licenses/by-nc/3.0/br/
Expand All @@ -754,32 +762,32 @@ def parse_url(url):

url = url.lower().rstrip("/")
url_parts = [p for p in url.split("/") if p]

if not url_parts:
return {}

license_types = dict(choices.LICENSE_TYPES)

for i, part in enumerate(url_parts):
if part not in license_types:
continue

license_type = part
remaining = url_parts[i + 1:]
license_version = None
license_language = None

if remaining:
version_candidate = remaining[0]
if all(c.isdigit() or c == "." for c in version_candidate):
license_version = version_candidate
remaining = remaining[1:]

if remaining:
lang_candidate = remaining[0]
if lang_candidate.isalpha() and 2 <= len(lang_candidate) <= 3:
license_language = lang_candidate

return {
"license_type": license_type,
"license_version": license_version,
Expand Down Expand Up @@ -853,7 +861,7 @@ def initial_date_isoformat(self):
if self.initial_date:
return self.initial_date.isoformat()
return None

@property
def final_date_isoformat(self):
if self.final_date:
Expand Down Expand Up @@ -1305,7 +1313,7 @@ def create(cls, pid, collection, data=None, user=None, url=None, processing_date
obj.creator = user
obj.save()
return obj

@classmethod
def create_or_update(cls, pid, collection, data=None, user=None, url=None, status=None, processing_date=None, force_update=None, new_record=None):
try:
Expand Down
Loading