Hacktoberfest 2026 : les issues que les mainteneurs ont marquées pour octobre, ouvertes et accessibles aux débutants. Parcourir les issues Hacktoberfest

A set[str] field of a fragment with validate_on_load=False loads the stored list as a list, not a set

Fermée
  • Quelque chose ne va pas ?
#129 0 commentaires 0 réactions 0 personnes assignées Voir sur GitHub

Les mainteneurs répondent en général sous 1 jour

Une pull request liée a déjà été fusionnée.

  • #204 par @jasperblues — fusionnée

Évaluation

Difficulté
4/5
Temps estimé
1-2 jours
Accessibilité débutants
67/100
Type d'issue
Bug
Clarté
Clairement spécifiée
Activité
Active
Stack technique
python
Domaine
backend

Piste de recherche

Start with drivine/mapper/graph_object_mapper.py and inspect _properties, _collected, and _instance; the issue describes how validate_on_load=False skips coercion. Port the listed unit and integration cases from SetFields.spec.ts and StatelessGraphObjectManager.sets.spec.ts, covering fragments, views, relationships, paths, nulls, and absent values. Done means declared set fields load as sets without changing absent or null values, and loaded sets save back as lists.

Rédigé par le modèle d'indexation à partir du texte de l'issue.

Description

bug

Port of liberation-data/drivine#280.

What happens here now: a set[str] field of a fragment declared with validate_on_load=False loads the list the store holds as a list, not a set. A validated fragment is fine: Pydantic's model_validate coerces the list to the field's set.

GraphObjectMapper._properties (drivine/mapper/graph_object_mapper.py:84-114) copies each stored value as it is, and only an empty_when_absent field passes through _collected (line 111-113). fragment() then builds the instance with _instance(..., model.validate_on_load) (line 45), and _instance (line 122-123) uses model_construct when validation is off, which coerces nothing. So the same field loads as set() when the node holds nothing for it (@EmptyWhenAbsent), and as a list when it holds items. A tuple[...] field has the same gap. Relationship and path collections already go through _collected with their kind, validated or not (lines 53-62).

Checked against main (987f07f):

class U(NodeFragment, labels=["U"], validate_on_load=False):
    id: Annotated[str, NodeId()]
    codes: set[str] = set()

mapper.fragment(registry.fragment(U), {"props": {"id": "1", "codes": ["x", "y"]}, "labels": ["U"]}).codes
# ['x', 'y'], a list; the same class without validate_on_load=False gives {'x', 'y'}

What the fix in drivine does: the registry records which fields hold a Set (FieldMapping.sets), and the mapper turns the list a store holds for such a field into a Set, in fragments, views, the fragments and views they hold, relationship fragments and paths. A field the node holds nothing for, or holds null, is left as it is unless it has @EmptyWhenAbsent. A {} a set was held as before sets were stored as lists loads as an empty set.

Suggested fix here: have _properties pass a stored list through _collected for any field whose FieldShape is a set or tuple, so model_construct gets the declared collection too.

Tests to port: drivine's test/1.Unit/mapper/SetFields.spec.ts (a load gives a set; absent stays absent; null stays null; in a view, a nested view, a relationship fragment and a path) and test/2.Integration/manager/StatelessGraphObjectManager.sets.spec.ts (store as a list, load as a set, save the loaded set back as a list, on every store), each with validate_on_load=False.

Langage dominant
Python
Étoiles
1
Forks
0
Merge moyen
1 h 21 min
PR mergées (30 j)
128

Préparer son environnement

Ce projet ne fournit ni conteneur de développement, ni Dockerfile, ni guide de contribution : l'installation est à votre charge. Commencez par son README, et consultez notre guide de la première contribution pour les étapes générales.

Par où commencer

  1. Lisez l'issue en entier, puis le guide de contribution du projet.
  2. Signalez en commentaire que vous la prenez — cela évite que deux personnes fassent le même travail.
  3. Forkez le dépôt et travaillez sur une branche.
  4. Ouvrez une pull request qui référence le numéro de l'issue.

Autres issues de liberation-data/drivine-python

Toutes les issues de liberation-data/drivine-python

Issues similaires

Plus d'issues Python

Recevez les nouvelles issues par e-mail

Un résumé court des issues GitHub adaptées aux débutants.