Pagination (200/page), batched+checkpointed ingest (batch_size, resume-safe), quality max_files slicing, upload page; Caddy routes for prefect/photo-filter

quality_scan.py

@@ -86,9 +86,37 @@ def _move(p: Path, dest_root: Path, src_root: Path):
     shutil.move(str(p), str(dest))
 
 
+
+
+def _slice_dir(base_dir: str, max_files: int) -> str:
+    """Copy first N images into a temp dir for CleanVision to audit."""
+    import shutil
+    import tempfile
+    from pathlib import Path
+
+    src = Path(base_dir)
+    tmp = Path(tempfile.mkdtemp(prefix="cvslice_"))
+    exts = {".jpg", ".jpeg", ".png", ".webp", ".gif", ".heic", ".tif", ".bmp"}
+    n = 0
+    for p in src.rglob("*"):
+        if p.is_file() and p.suffix.lower() in exts:
+            shutil.copy2(p, tmp / p.name)
+            n += 1
+            if n >= max_files:
+                break
+    print(f"_slice_dir: copied {n} files to {tmp}")
+    return str(tmp)
+
+
 @flow(name="photo-quality-scan")
-def quality_scan(base_dir: str, move: bool = False, notify: bool = True):
-    """Audit image quality with CleanVision; classify into keep/review/delete."""
+def quality_scan(base_dir: str, move: bool = False, notify: bool = True, max_files: int = None):
+    """Audit image quality with CleanVision; classify into keep/review/delete.
+
+    max_files: if set, only audit the first N image files (slices huge folders
+    into reviewable chunks — prevents OOM on 50K-file trees).
+    """
+    if max_files:  # 0/None = unlimited
+        base_dir = _slice_dir(base_dir, max_files)
     audit = audit_folder(base_dir)
     print(f"Issue summary: {audit['summary']}")
     result = classify_and_sort(base_dir, audit["per_image"], move=move)