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

tests/test_images.py

@@ -0,0 +1,80 @@
+import os
+import pytest
+from io import BytesIO
+from unittest.mock import patch, MagicMock
+from PIL import Image as PILImage
+from src.images import process_image, _resize_to_fit
+
+
+def _make_test_image(width, height, fmt="JPEG"):
+    img = PILImage.new("RGB", (width, height), color="red")
+    buf = BytesIO()
+    img.save(buf, format=fmt)
+    buf.seek(0)
+    return buf.read()
+
+
+def test_resize_landscape_downscale():
+    img = PILImage.new("RGB", (1600, 900))
+    result = _resize_to_fit(img)
+    assert result.width <= 800
+    assert result.height <= 480
+    assert result.width / result.height == pytest.approx(1600 / 900, rel=0.02)
+
+
+def test_resize_portrait_downscale():
+    img = PILImage.new("RGB", (600, 1200))
+    result = _resize_to_fit(img)
+    assert result.width <= 480
+    assert result.height <= 800
+    assert result.width / result.height == pytest.approx(600 / 1200, rel=0.02)
+
+
+def test_resize_small_image_upscales():
+    img = PILImage.new("RGB", (200, 100))
+    result = _resize_to_fit(img)
+    assert result.width > 200
+    assert result.width <= 800
+    assert result.height <= 480
+
+
+def test_resize_already_fits():
+    img = PILImage.new("RGB", (800, 480))
+    result = _resize_to_fit(img)
+    assert result.size == (800, 480)
+
+
+def test_process_image_downloads_and_saves(tmp_path):
+    image_bytes = _make_test_image(1024, 768)
+
+    mock_response = MagicMock()
+    mock_response.content = image_bytes
+    mock_response.raise_for_status = MagicMock()
+
+    with patch("src.images.requests.get", return_value=mock_response):
+        path, w, h = process_image(
+            "https://example.com/photo.jpg", str(tmp_path)
+        )
+
+    assert os.path.exists(path)
+    assert path.endswith(".jpg")
+    assert w <= 800
+    assert h <= 480
+
+    saved = PILImage.open(path)
+    assert saved.format == "JPEG"
+
+
+def test_process_image_dedup(tmp_path):
+    image_bytes = _make_test_image(500, 300)
+
+    mock_response = MagicMock()
+    mock_response.content = image_bytes
+    mock_response.raise_for_status = MagicMock()
+
+    with patch("src.images.requests.get", return_value=mock_response) as mock_get:
+        path1, _, _ = process_image("https://example.com/same.jpg", str(tmp_path))
+        path2, _, _ = process_image("https://example.com/same.jpg", str(tmp_path))
+
+    assert path1 == path2
+    assert mock_get.call_count == 1