[crawler]

crawler_platform/app/core/ontology/__init__.py

@@ -0,0 +1,2 @@
+"""Ontology definitions and mapping helpers."""
+

crawler_platform/app/core/ontology/definitions.py

@@ -0,0 +1,130 @@
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+
+
+@dataclass(slots=True)
+class Ontology:
+    domain: str
+    entity_types: list[str]
+    predicates: list[str]
+    attributes: list[str]
+    aliases: dict[str, str] = field(default_factory=dict)
+
+
+COMMON_ONTOLOGY = Ontology(
+    domain="common",
+    entity_types=[
+        "Source",
+        "Page",
+        "Entity",
+        "Attribute",
+        "Relation",
+        "Claim",
+        "Evidence",
+        "Extraction",
+        "Confidence",
+        "UpdatePolicy",
+    ],
+    predicates=[
+        "mentions",
+        "hasAttribute",
+        "relatedTo",
+        "sameAs",
+        "soldBy",
+        "hasPrice",
+    ],
+    attributes=["name", "source_url", "updated_at", "confidence"],
+)
+
+
+DOMAIN_ONTOLOGIES: dict[str, Ontology] = {
+    "perfume": Ontology(
+        domain="perfume",
+        entity_types=[
+            "Perfume",
+            "Brand",
+            "Note",
+            "Accord",
+            "Mood",
+            "Season",
+            "Occasion",
+            "Review",
+            "Price",
+            "ProductPage",
+        ],
+        predicates=[
+            "hasBrand",
+            "hasTopNote",
+            "hasMiddleNote",
+            "hasBaseNote",
+            "hasAccord",
+            "evokesMood",
+            "suitableForSeason",
+            "suitableForOccasion",
+            "similarTo",
+            "soldBy",
+            "hasPrice",
+            "hasReviewKeyword",
+        ],
+        attributes=[
+            "name",
+            "brand",
+            "gender_bias",
+            "longevity",
+            "sillage",
+            "price_range",
+            "popularity_score",
+            "review_count",
+            "source_url",
+            "updated_at",
+        ],
+        aliases={
+            "top_notes": "hasTopNote",
+            "middle_notes": "hasMiddleNote",
+            "heart_notes": "hasMiddleNote",
+            "base_notes": "hasBaseNote",
+            "accords": "hasAccord",
+            "mood_tags": "evokesMood",
+            "season_tags": "suitableForSeason",
+            "occasion_tags": "suitableForOccasion",
+            "review_keywords": "hasReviewKeyword",
+            "price": "hasPrice",
+        },
+    ),
+    "tea": Ontology(
+        domain="tea",
+        entity_types=["Tea", "Ingredient", "Flavor", "Effect", "CaffeineLevel", "MoodState"],
+        predicates=["hasIngredient", "hasFlavor", "hasEffect", "suitableForCondition"],
+        attributes=["name", "origin", "caffeine_level", "price_range"],
+    ),
+    "coffee": Ontology(
+        domain="coffee",
+        entity_types=["CoffeeBean", "Origin", "RoastLevel", "FlavorNote", "BrewMethod"],
+        predicates=["hasOrigin", "hasRoastLevel", "hasFlavorNote", "recommendedForBrewMethod"],
+        attributes=["name", "origin", "roast_level", "process", "price_range"],
+    ),
+    "candle": Ontology(
+        domain="candle",
+        entity_types=["ScentProduct", "ScentNote", "SpaceType", "Mood", "Season"],
+        predicates=["suitableForSpace", "evokesMood", "hasScentNote"],
+        attributes=["name", "burn_time", "volume", "price_range"],
+    ),
+    "supplement": Ontology(
+        domain="supplement",
+        entity_types=["Supplement", "Ingredient", "HealthGoal", "Symptom", "Dosage"],
+        predicates=["hasIngredient", "supportsGoal", "recommendedForCondition"],
+        attributes=["name", "dosage", "warnings", "price_range"],
+    ),
+    "gift": Ontology(
+        domain="gift",
+        entity_types=["GiftProduct", "RecipientType", "Relationship", "Occasion", "PersonalityTag"],
+        predicates=["suitableForRecipient", "suitableForOccasion", "matchesPersonality"],
+        attributes=["name", "price_range", "availability", "gift_wrap_available"],
+    ),
+}
+
+
+def ontology_for_domain(domain: str) -> Ontology:
+    return DOMAIN_ONTOLOGIES.get(domain, COMMON_ONTOLOGY)
+

crawler_platform/app/core/ontology/mapper.py

@@ -0,0 +1,27 @@
+from __future__ import annotations
+
+from typing import Any
+
+
+def normalize_predicate(predicate: str, ontology: dict[str, Any] | None) -> str:
+    if not ontology:
+        return predicate
+    aliases = ontology.get("aliases", {})
+    return aliases.get(predicate, predicate)
+
+
+def is_allowed_predicate(predicate: str, ontology: dict[str, Any] | None) -> bool:
+    if not ontology or not ontology.get("predicates"):
+        return True
+    return predicate in ontology["predicates"]
+
+
+def ontology_to_dict(domain_ontology) -> dict[str, Any]:
+    return {
+        "domain": domain_ontology.domain,
+        "entity_types": domain_ontology.entity_types,
+        "predicates": domain_ontology.predicates,
+        "attributes": domain_ontology.attributes,
+        "aliases": domain_ontology.aliases,
+    }
+