feat(download_imageset): process ZIP archives

download_imageset.py

@@ -9,8 +9,8 @@ import os
 import re
 import struct
 import subprocess
-import sys
 import urllib.parse
+import zipfile
 from functools import partial
 from pathlib import Path
 from typing import Callable
@@ -413,37 +413,58 @@ def download_cmd(
 
 
 @cli.command("process")
-@click.argument("dirs", nargs=-1, required=True)
-def process_cmd(dirs: tuple[str, ...]) -> None:
-    for d in dirs:
-        p = Path(d)
+@click.argument("paths", nargs=-1, required=True, type=click.Path(path_type=Path))
+def process_cmd(paths: tuple[Path, ...]) -> None:
+    """Process directories and ZIP files, extracting ZIPs into sibling directories."""
+    for p in paths:
+        archive_path: Path | None = None
         if not p.exists():
-            print(f"skipping {p}: does not exist", file=sys.stderr)
+            logging.warning(f"skipping {p}: does not exist")
+            continue
+        if p.is_file() and p.suffix.lower() == ".zip":
+            output_dir = p.with_suffix("")
+            if output_dir.exists() and not output_dir.is_dir():
+                logging.warning(f"skipping {p}: {output_dir} is not a directory")
+                continue
+            try:
+                with zipfile.ZipFile(file=p) as archive:
+                    logging.info(f"extracting {p} → {output_dir}")
+                    archive.extractall(path=output_dir)
+            except zipfile.BadZipFile:
+                logging.warning(f"skipping {p}: invalid ZIP file")
+                continue
+            archive_path = p
+            p = output_dir
+        if not p.is_dir():
+            logging.warning(f"skipping {p}: expected a directory or ZIP file")
             continue
         if _needs_resize(p):
             resized_dir = p / "resized"
             resized_dir.mkdir(exist_ok=True)
             originals = [f for f in p.rglob("*") if is_image(f)]
-            print(f"resizing {len(originals)} image(s) in {p} → {resized_dir}")
+            logging.info(f"resizing {len(originals)} image(s) in {p} → {resized_dir}")
             for img in originals:
                 run_command(
                     [
                         "imgz",
                         "resize",
-                        "-o",
+                        "--output",
                         str(resized_dir / f"{img.stem}.resized.jpg"),
                         str(img),
                     ]
                 )
             if originals:
-                print(f"trashing {len(originals)} original(s)")
+                logging.info(f"trashing {len(originals)} original(s)")
                 move_to_trash(originals)
         else:
-            print(f"skipping {p}: already within size limits")
+            logging.info(f"skipping {p}: already within size limits")
 
         rename_children(p, flatten=True)
         remove_empty_dirs(p)
+        if archive_path is not None:
+            move_to_trash(file_paths=[archive_path])
 
 
 if __name__ == "__main__":
+    logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s: %(message)s")
     cli()