Publish curated VacatureRadar source snapshot

apps/sources/adapters/base.py

@@ -0,0 +1,52 @@
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from typing import Any, Protocol
+
+
+@dataclass(slots=True)
+class FieldEvidence:
+    field_name: str
+    method: str
+    confidence: float
+    evidence: str = ""
+
+
+@dataclass(slots=True)
+class ExtractedJob:
+    url: str
+    title: str
+    employer_name: str = ""
+    external_id: str = ""
+    location_text: str = ""
+    region: str = ""
+    postal_code: str = ""
+    country: str = ""
+    description_html: str = ""
+    description_text: str = ""
+    language: str = ""
+    date_posted: str = ""
+    valid_through: str = ""
+    employment_types: list[str] = field(default_factory=list)
+    workplace_type: str = ""
+    compensation: dict[str, Any] = field(default_factory=dict)
+    skills_required: list[str] = field(default_factory=list)
+    skills_preferred: list[str] = field(default_factory=list)
+    raw: dict[str, Any] = field(default_factory=dict)
+    evidence: list[FieldEvidence] = field(default_factory=list)
+
+
+@dataclass(slots=True)
+class ExtractionResult:
+    jobs: list[ExtractedJob]
+    parser_key: str
+    parser_version: str
+    confidence: float
+    warnings: list[str] = field(default_factory=list)
+
+
+class SourceAdapter(Protocol):
+    parser_key: str
+    parser_version: str
+
+    def extract(self, content: str, *, url: str) -> ExtractionResult: ...