Initial commit — PlantGuideScraper project

backend/app/scrapers/base.py

@@ -0,0 +1,57 @@
+from abc import ABC, abstractmethod
+from typing import Dict, Any, Optional
+import logging
+
+from sqlalchemy.orm import Session
+
+from app.models import Species, ApiKey
+
+
+class BaseScraper(ABC):
+    """Base class for all image scrapers."""
+
+    name: str = "base"
+    requires_api_key: bool = True
+
+    @abstractmethod
+    def scrape_species(
+        self,
+        species: Species,
+        db: Session,
+        logger: Optional[logging.Logger] = None
+    ) -> Dict[str, int]:
+        """
+        Scrape images for a species.
+
+        Args:
+            species: The species to scrape images for
+            db: Database session
+            logger: Optional logger for debugging
+
+        Returns:
+            Dict with 'downloaded' and 'rejected' counts
+        """
+        pass
+
+    @abstractmethod
+    def test_connection(self, api_key: ApiKey) -> str:
+        """
+        Test API connection.
+
+        Args:
+            api_key: The API key configuration
+
+        Returns:
+            Success message
+
+        Raises:
+            Exception if connection fails
+        """
+        pass
+
+    def get_api_key(self, db: Session) -> ApiKey:
+        """Get API key for this scraper."""
+        return db.query(ApiKey).filter(
+            ApiKey.source == self.name,
+            ApiKey.enabled == True
+        ).first()