Add ruff and pre-commit config and fix issues

ISI-MIP · Aug 30, 2023 · 32cbd6f · 32cbd6f
1 parent 1fc068e
commit 32cbd6f
Show file tree

Hide file tree

Showing 22 changed files with 177 additions and 60 deletions.
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -0,0 +1,18 @@
+exclude: \/migrations\/
+
+repos:
+  - repo: meta
+    hooks:
+      - id: check-hooks-apply
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.4.0
+    hooks:
+      - id: check-ast
+      - id: end-of-file-fixer
+      - id: trailing-whitespace
+      - id: debug-statements
+  - repo: https://github.com/charliermarsh/ruff-pre-commit
+    rev: v0.0.284
+    hooks:
+      - id: ruff
+        args: [--fix, --exit-non-zero-on-fix]
diff --git a/README.md b/README.md
@@ -7,7 +7,7 @@ django-datacite
 [![pytest Workflow Status](https://github.com/ISI-MIP/django-datacite/actions/workflows/pytest.yml/badge.svg)](https://github.com/ISI-MIP/django-datacite/actions/workflows/pytest.yml)
 [![Coverage Status](https://coveralls.io/repos/github/ISI-MIP/django-datacite/badge.svg?branch=master)](https://coveralls.io/github/ISI-MIP/django-datacite?branch=master)
 
-A Django app to properly model the [DataCite Metadata Schema](https://schema.datacite.org/) in a relational database, with full integration into the Django admin interface. The app is *slightly* opinionated in a way to make it better usable as a DOI registration database. Names (`creators` and `contributors`) and identifiers (for the resources, but also for `alternativeIdentifiers` and `relatedIdentifiers`) are stored in separate database tables and can be reused for different resources. 
+A Django app to properly model the [DataCite Metadata Schema](https://schema.datacite.org/) in a relational database, with full integration into the Django admin interface. The app is *slightly* opinionated in a way to make it better usable as a DOI registration database. Names (`creators` and `contributors`) and identifiers (for the resources, but also for `alternativeIdentifiers` and `relatedIdentifiers`) are stored in separate database tables and can be reused for different resources.
 
 This app provides a set of [models](django_datacite/models.py), [admin classes](django_datacite/admin.py), [views](django_datacite/views.py), and utility functions to export and import DataCite files. All options and default settings can be customized in the Django `settings` using the same syntax as in [django_datacite/settings.py](django_datacite/settings.py). Identifiers have an additional `citation` field to store the human readable citation for reuse in other tools.
 

diff --git a/conftest.py b/conftest.py
@@ -1,6 +1,7 @@
 from pathlib import Path
 
 import pytest
+
 from django.conf import settings
 from django.core.management import call_command
 

diff --git a/django_datacite/admin.py b/django_datacite/admin.py
@@ -1,6 +1,5 @@
 import json
 
-import requests
 from django import forms
 from django.contrib import admin
 from django.core.exceptions import ValidationError
@@ -10,13 +9,31 @@
 from django.urls import path
 from django.utils.translation import gettext as _
 
+import requests
+
 from .exports import export_resource
 from .imports import import_resource
-from .models import (AlternateIdentifier, Contributor, Creator, Date,
-                     Description, FundingReference, GeoLocation,
-                     GeoLocationBox, GeoLocationPoint, GeoLocationPolygon,
-                     Identifier, Name, NameIdentifier, RelatedIdentifier,
-                     RelatedItem, Resource, Rights, Subject, Title)
+from .models import (
+    AlternateIdentifier,
+    Contributor,
+    Creator,
+    Date,
+    Description,
+    FundingReference,
+    GeoLocation,
+    GeoLocationBox,
+    GeoLocationPoint,
+    GeoLocationPolygon,
+    Identifier,
+    Name,
+    NameIdentifier,
+    RelatedIdentifier,
+    RelatedItem,
+    Resource,
+    Rights,
+    Subject,
+    Title,
+)
 from .renderers import XMLRenderer
 from .utils import render_bibtex
 
@@ -184,25 +201,25 @@ def clean(self):
         elif cleaned_data.get('file'):
             try:
                 cleaned_data['data'] = json.load(cleaned_data['file'])
-            except json.JSONDecodeError:
-                raise ValidationError(_('Please provide a valid JSON.'))
+            except json.JSONDecodeError as e:
+                raise ValidationError(_('Please provide a valid JSON.')) from e
 
         elif cleaned_data.get('url'):
             response = requests.get(cleaned_data['url'])
             response.raise_for_status()
 
             try:
                 cleaned_data['data'] = response.json()
-            except json.JSONDecodeError:
-                raise ValidationError(_('Please provide a valid JSON.'))
+            except json.JSONDecodeError as e:
+                raise ValidationError(_('Please provide a valid JSON.')) from e
 
         else:
             raise ValidationError(_('Please provide a file OR a URL.'))
 
 
 # Inlines
 
-class NoExtraInlineMixin(object):
+class NoExtraInlineMixin:
     extra = 0
 
 
@@ -334,33 +351,39 @@ def get_queryset(self, request):
 
     def get_urls(self):
         return [
-            path('<int:pk>/export/<str:format>/', self.admin_site.admin_view(self.datacite_resource_export),
+            path('<int:pk>/export/<str:format>/',
+                 self.admin_site.admin_view(self.datacite_resource_export),
                  name='datacite_resource_export'),
-            path('import/', self.admin_site.admin_view(self.datacite_resource_import),
+            path('import/',
+                 self.admin_site.admin_view(self.datacite_resource_import),
                  name='datacite_resource_import'),
-            path('<int:pk>/import/', self.admin_site.admin_view(self.datacite_resource_import),
+            path('<int:pk>/import/',
+                 self.admin_site.admin_view(self.datacite_resource_import),
                  name='datacite_resource_import'),
-            path('<int:pk>/copy/', self.admin_site.admin_view(self.datacite_resource_copy),
+            path('<int:pk>/copy/',
+                 self.admin_site.admin_view(self.datacite_resource_copy),
                  name='datacite_resource_copy'),
-            path('<int:pk>/validate/', self.admin_site.admin_view(self.datacite_resource_validate),
+            path('<int:pk>/validate/',
+                 self.admin_site.admin_view(self.datacite_resource_validate),
                  name='datacite_resource_validate'),
-            ] + super().get_urls()
+            *super().get_urls()
+        ]
 
     def datacite_resource_export(self, request, pk=None, format=None):
         resource = get_object_or_404(Resource, id=pk)
 
         if format == 'json':
             resource_json = json.dumps(export_resource(resource), indent=2)
             response = HttpResponse(resource_json, content_type="application/json")
-            response['Content-Disposition'] = 'filename="{}.json"'.format(resource.identifier)
+            response['Content-Disposition'] = f'filename="{resource.identifier}.json"'
         elif format == 'xml':
             resource_xml = XMLRenderer().render(export_resource(resource))
             response = HttpResponse(resource_xml, content_type="application/xml")
-            response['Content-Disposition'] = 'filename="{}.xml"'.format(resource.identifier)
+            response['Content-Disposition'] = f'filename="{resource.identifier}.xml"'
         elif format == 'bibtex':
             resource_bibtex = render_bibtex(resource)
             response = HttpResponse(resource_bibtex, content_type='application/x-bibtex')
-            response['Content-Disposition'] = 'filename="{}.bib"'.format(resource.identifier)
+            response['Content-Disposition'] = f'filename="{resource.identifier}.bib"'
         else:
             raise Http404
 

diff --git a/django_datacite/exports.py b/django_datacite/exports.py
@@ -50,7 +50,8 @@ def export_resource(resource):
     # contributors
     contributors = resource.contributor_set.all()
     if contributors:
-        data['contributors'] = [export_name(contributor.name, contributor.contributor_type) for contributor in contributors]
+        data['contributors'] = [export_name(contributor.name, contributor.contributor_type)
+                                for contributor in contributors]
 
     # dates
     dates = resource.dates.all()
@@ -72,7 +73,8 @@ def export_resource(resource):
     # related identifiers
     related_identifiers = resource.relatedidentifier_set.all()
     if related_identifiers:
-        data['relatedIdentifiers'] = [export_related_identifiers(related_identifier) for related_identifier in related_identifiers]
+        data['relatedIdentifiers'] = [export_related_identifiers(related_identifier)
+                                      for related_identifier in related_identifiers]
 
     # size
     if resource.size:
@@ -113,7 +115,8 @@ def export_resource(resource):
     # funding reference
     funding_references = resource.fundingreference_set.all()
     if funding_references:
-        data['fundingReferences'] = [export_funding_references(funding_reference) for funding_reference in funding_references]
+        data['fundingReferences'] = [export_funding_references(funding_reference)
+                                     for funding_reference in funding_references]
 
     # related items
     related_items = resource.relateditem_set.all()
@@ -261,7 +264,7 @@ def export_geo_location(geo_location):
             }
         })
     if geo_location_polygons:
-        data['geoLocationPolygons'] = [geo_location_polygon for geo_location_polygon in geo_location_polygons]
+        data['geoLocationPolygons'] = list(geo_location_polygons)
 
     return data
 
@@ -354,6 +357,7 @@ def export_related_item(related_item):
     # contributors
     contributors = related_item.item.contributor_set.all()
     if contributors:
-        data['contributors'] = [export_name(contributor.name, contributor.contributor_type) for contributor in contributors]
+        data['contributors'] = [export_name(contributor.name, contributor.contributor_type)
+                                for contributor in contributors]
 
     return data
diff --git a/django_datacite/imports.py b/django_datacite/imports.py
@@ -3,11 +3,27 @@
 
 from django.utils.dateparse import parse_date
 
-from .models import (AlternateIdentifier, Contributor, Creator, Date,
-                     Description, FundingReference, GeoLocation,
-                     GeoLocationBox, GeoLocationPoint, GeoLocationPolygon,
-                     Identifier, Name, NameIdentifier, RelatedIdentifier,
-                     RelatedItem, Resource, Rights, Subject, Title)
+from .models import (
+    AlternateIdentifier,
+    Contributor,
+    Creator,
+    Date,
+    Description,
+    FundingReference,
+    GeoLocation,
+    GeoLocationBox,
+    GeoLocationPoint,
+    GeoLocationPolygon,
+    Identifier,
+    Name,
+    NameIdentifier,
+    RelatedIdentifier,
+    RelatedItem,
+    Resource,
+    Rights,
+    Subject,
+    Title,
+)
 from .utils import get_settings
 
 logger = logging.getLogger(__name__)
@@ -170,7 +186,8 @@ def import_resource(resource_instance, data):
                         'order': order
                     }
                 )
-                logger.info('AlternateIdentifier="%s" %s', alternate_identifier_instance, 'created' if created else 'updated')
+                logger.info('AlternateIdentifier="%s" %s',
+                            alternate_identifier_instance, 'created' if created else 'updated')
 
     # relatedIdentifiers
     related_identifier_nodes = data.get('relatedIdentifiers')
@@ -200,7 +217,8 @@ def import_resource(resource_instance, data):
                             'resource_type_general': resource_type_general
                         }
                     )
-                    logger.info('RelatedIdentifier="%s" %s', related_identifier_instance, 'created' if created else 'updated')
+                    logger.info('RelatedIdentifier="%s" %s',
+                                related_identifier_instance, 'created' if created else 'updated')
 
     # rightsList
     right_list_node = data.get('rightsList')
@@ -251,7 +269,8 @@ def import_resource(resource_instance, data):
                         'award_title': funding_reference_node.get('awardTitle', '')
                     }
                 )
-                logger.info('FundingReference="%s" %s', funding_reference_instance, 'created' if created else 'updated')
+                logger.info('FundingReference="%s" %s',
+                            funding_reference_instance, 'created' if created else 'updated')
 
     # related_items
     related_item_nodes = data.get('relatedItems')

diff --git a/django_datacite/models.py b/django_datacite/models.py
@@ -200,8 +200,8 @@ def citation(self):
     def get_absolute_url(self):
         try:
             return reverse('django_datacite:resource', args=[self.identifier.identifier])
-        except (NoReverseMatch, AttributeError):
-            raise Http404
+        except (NoReverseMatch, AttributeError) as e:
+            raise Http404 from e
 
     @staticmethod
     def get_default_public():
@@ -718,7 +718,8 @@ class GeoLocationBox(models.Model):
     )
 
     def __str__(self):
-        return '{west_bound_longitude}, {east_bound_longitude}, {south_bound_latitude}, {north_bound_latitude}'.format(**vars(self))
+        return '{west_bound_longitude}, {east_bound_longitude}, ' \
+               '{south_bound_latitude}, {north_bound_latitude}'.format(**vars(self))
 
 
 class GeoLocationPolygon(models.Model):

diff --git a/django_datacite/renderers.py b/django_datacite/renderers.py
@@ -3,7 +3,7 @@
 from xml.sax.saxutils import XMLGenerator
 
 
-class XMLRenderer(object):
+class XMLRenderer:
 
     def __init__(self):
         self.stream = io.StringIO()
@@ -19,7 +19,7 @@ def render(self, data):
     def render_node(self, tag, attrs, value):
         if value is not None:
             # remove None values from attrs
-            attrs = dict((k, v) for k, v in attrs.items() if v is not None)
+            attrs = {k: v for k, v in attrs.items() if v is not None}
 
             self.xml.startElement(tag, attrs)
             self.xml.characters(str(value))

diff --git a/django_datacite/settings.py b/django_datacite/settings.py
@@ -183,11 +183,16 @@
 
 DATACITE_DEFAULT_RIGHTS_IDENTIFIER = 'CC0-1.0'
 DATACITE_RIGHTS_IDENTIFIERS = (
-    ('CC0-1.0', _('CC0 1.0 Universal Public Domain Dedication')),
-    ('CC-BY-4.0', _('Creative Commons Attribution 4.0 International (CC BY 4.0)')),
-    ('CC-BY-SA-4.0', _('Creative Commons Attribution Share Alike 4.0 International (CC BY-SA 4.0)')),
-    ('CC-BY-NC-4.0', _('Creative Commons Attribution Non Commercial 4.0 International (CC BY-NC 4.0)')),
-    ('CC-BY-NC-SA-4.0', _('Creative Commons Attribution Non Commercial Share Alike 4.0 International (CC BY-NC-SA 4.0)')),
+    ('CC0-1.0',
+        _('CC0 1.0 Universal Public Domain Dedication')),
+    ('CC-BY-4.0',
+        _('Creative Commons Attribution 4.0 International (CC BY 4.0)')),
+    ('CC-BY-SA-4.0',
+        _('Creative Commons Attribution Share Alike 4.0 International (CC BY-SA 4.0)')),
+    ('CC-BY-NC-4.0',
+        _('Creative Commons Attribution Non Commercial 4.0 International (CC BY-NC 4.0)')),
+    ('CC-BY-NC-SA-4.0',
+        _('Creative Commons Attribution Non Commercial Share Alike 4.0 International (CC BY-NC-SA 4.0)')),
 )
 
 DATACITE_RIGHTS_IDENTIFIER_URIS = {

diff --git a/django_datacite/tests/test_admin.py b/django_datacite/tests/test_admin.py
@@ -1,5 +1,5 @@
-import re
 import json
+import re
 
 from django.conf import settings
 from django.urls import reverse

diff --git a/django_datacite/tests/test_import.py b/django_datacite/tests/test_import.py
@@ -3,7 +3,7 @@
 from django.conf import settings
 
 from django_datacite.imports import import_resource
-from django_datacite.models import Resource, Name, Identifier, Subject
+from django_datacite.models import Identifier, Name, Resource, Subject
 
 resource_id = 1
 

diff --git a/django_datacite/tests/test_models.py b/django_datacite/tests/test_models.py
@@ -2,8 +2,7 @@
 
 from django.http import Http404
 
-from django_datacite.models import Resource, NameIdentifier, GeoLocation
-
+from django_datacite.models import GeoLocation, NameIdentifier, Resource
 
 resource_id = 1
 name_identifier_id = 1

diff --git a/django_datacite/urls.py b/django_datacite/urls.py
@@ -1,6 +1,6 @@
 from django.urls import re_path
 
-from django_datacite.views import resource, resource_json, resource_xml, resource_bibtex
+from django_datacite.views import resource, resource_bibtex, resource_json, resource_xml
 
 app_name = 'django_datacite'
 

diff --git a/django_datacite/validators.py b/django_datacite/validators.py
@@ -33,5 +33,5 @@ def validate_polygon_points(value):
                 ]
             }
         })
-    except jsonschema.exceptions.ValidationError:
-        raise ValidationError('JSON is not of the form [[lon, lat], ...] with at least 4 items.')
+    except jsonschema.exceptions.ValidationError as e:
+        raise ValidationError('JSON is not of the form [[lon, lat], ...] with at least 4 items.') from e