fix: downloads stop leaving files on disk that the library doesn't use

- An archive whose images were extracted was kept at its download path
  as well as in the attachments store. Both download paths now remove
  the downloaded copy.
- A downloaded file the import filters reject (too small, transparent,
  single colour, unreadable) was left on disk with no record. A later
  walk found it, imported it again and rejected it again, every time.
  Rejected files are now removed; the seen ledger keeps them from being
  downloaded again.
- Both paths share one policy, Importer.discard_after_attach, which
  never removes a file an ImageRecord points at.
- An external-link fetch that fails, or whose worker was killed, left
  partial files (megatools' .megatmp.*) in the link's dir, and a killed
  fetch left its link 'downloading' for good. A fetch now clears
  unrecorded leftovers before starting and after failing. The sweep
  returns a link to retry once it has been 'downloading' past the
  task's hard limit, using a new external_link.started_at column
  (migration 0117).
- The orphaned temp-file sweep now recognises .megatmp.* as well as
  .part and .partial.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

tests/test_download_service.py

@@ -193,6 +193,69 @@ async def test_download_source_attaches_written_files(
     assert imported_when_marked == [2]
 
 
+@pytest.mark.asyncio
+async def test_import_step_keeps_only_the_files_the_library_needs_on_disk(
+    db, db_sync, tmp_path, seed_artist_and_source,
+):
+    """After the import step, a downloaded image the library took stays where it
+    is, while an archive whose images were extracted (the archive is in the
+    attachments store) and an image the import filters rejected are removed, so
+    nothing is left on disk without a record for a later walk to import again."""
+    import zipfile
+
+    from PIL import Image
+
+    from backend.app.models import PostAttachment
+    from backend.app.services.download_service import DownloadService
+    from backend.app.services.importer import Importer
+
+    _artist, source = seed_artist_and_source
+    images_root = tmp_path / "images"
+    post_dir = images_root / "alice" / "patreon" / "post"
+
+    kept = post_dir / "a.jpg"
+    _make_jpg(kept, split="h")
+    member = tmp_path / "member.jpg"
+    _make_jpg(member, split="v")
+    archive = post_dir / "pack.zip"
+    with zipfile.ZipFile(archive, "w") as z:
+        z.writestr("art.jpg", member.read_bytes())
+    tiny = post_dir / "tiny.png"
+    Image.new("RGB", (16, 16), (200, 30, 30)).save(tiny)  # below min_width
+
+    result = _make_fake_dl_result(
+        success=True, written_paths=[str(kept), str(archive), str(tiny)],
+        files_downloaded=3,
+    )
+    sync_settings = db_sync.execute(
+        select(ImportSettings).where(ImportSettings.id == 1)
+    ).scalar_one()
+    sync_settings.phash_threshold = 0
+    sync_settings.min_width = 32
+    importer = Importer(
+        session=db_sync, images_root=images_root, import_root=images_root,
+        thumbnailer=Thumbnailer(images_root=images_root), settings=sync_settings,
+    )
+    crypto = CredentialCrypto(tmp_path / "key.b64", bootstrap_ok=True)
+    svc = DownloadService(
+        async_session=db, sync_session=db_sync, gdl=_fake_gdl_with_result(result),
+        importer=importer, cred_service=CredentialService(db, crypto),
+    )
+    _stub_patreon_ingester(svc, result)
+    await svc.download_source(source.id)
+
+    assert kept.exists()
+    assert not archive.exists()
+    assert not tiny.exists()
+    # The archive and its image are both still in the library.
+    assert db_sync.execute(
+        select(func.count(PostAttachment.id)).where(
+            PostAttachment.original_filename == "pack.zip"
+        )
+    ).scalar_one() == 1
+    assert db_sync.execute(select(func.count(ImageRecord.id))).scalar_one() == 2
+
+
 @pytest.mark.asyncio
 async def test_download_source_skipped_when_disabled(
     db, db_sync, tmp_path, seed_artist_and_source,