feat: image download, resize-to-fit, baseline JPEG conversion

Made-with: Cursor

src/images.py

@@ -0,0 +1,61 @@
+import hashlib
+import logging
+import os
+from io import BytesIO
+
+import requests
+from PIL import Image as PILImage
+
+import config
+
+logger = logging.getLogger(__name__)
+
+
+def _url_hash(url: str) -> str:
+    return hashlib.sha256(url.encode()).hexdigest()[:16]
+
+
+def _resize_to_fit(img: PILImage.Image) -> PILImage.Image:
+    w, h = img.size
+    if w >= h:
+        max_w, max_h = config.IMAGE_MAX_LANDSCAPE
+    else:
+        max_w, max_h = config.IMAGE_MAX_PORTRAIT
+
+    scale = min(max_w / w, max_h / h)
+
+    new_w = int(w * scale)
+    new_h = int(h * scale)
+
+    if new_w == w and new_h == h:
+        return img
+
+    return img.resize((new_w, new_h), PILImage.Resampling.LANCZOS)
+
+
+def process_image(url: str, output_dir: str) -> tuple[str, int, int]:
+    """Download an image, resize it, save as baseline JPEG.
+
+    Returns (local_path, width, height). Deduplicates by URL hash.
+    """
+    os.makedirs(output_dir, exist_ok=True)
+    filename = f"{_url_hash(url)}.jpg"
+    local_path = os.path.join(output_dir, filename)
+
+    if os.path.exists(local_path):
+        img = PILImage.open(local_path)
+        return local_path, img.width, img.height
+
+    response = requests.get(url, timeout=30)
+    response.raise_for_status()
+
+    img = PILImage.open(BytesIO(response.content))
+
+    if img.mode in ("RGBA", "P", "LA"):
+        img = img.convert("RGB")
+
+    img = _resize_to_fit(img)
+
+    img.save(local_path, format="JPEG", progressive=False, quality=85)
+
+    return local_path, img.width, img.height