fix(downloads): a Facebook download titled only "Facebook" takes the post's text

Since 2026-10-28 a signed-in read of a reel gives the placeholder title and
no description, so the adapter had nothing to fall back on and every such
download (and its name) read "Facebook". A placeholder with nothing behind
it now gets one signed-out, metadata-only read of the page. The backfill
retitles the downloads already made, and every name that only repeated the
placeholder; a name its owner gave stays.

backend/tests/test_facebook_titles.py

@@ -0,0 +1,180 @@
+"""A Facebook download's title when the signed-in read gives only a placeholder.
+
+Measured on prod (2026-11-01): every Facebook download since 2026-10-28 is titled "Facebook" — the
+signed-in read (the owner's cookie) now gives that placeholder and NO description, so the adapter had
+no post text to fall back on, and the download name copied the placeholder too. A signed-out read of
+the same reel (yt-dlp 2026.08.19, as on prod) still carries the post's text in its description.
+"""
+
+from app.config import settings
+from app.downloads import adapters, backfill, pipeline
+from app.downloads.models import DownloadJob, MediaAsset
+
+_URL = "https://www.facebook.com/reel/1860682468422402/"
+# The measured signed-out read of that reel.
+_PUBLIC = {
+    "title": "163K reactions · 14K shares | sorry not helping #workhumor #officelife | Work dot exe",
+    "description": "sorry not helping\n\n#workhumor #officelife",
+    "uploader": "Work dot exe",
+}
+
+
+def _fake_ydl(monkeypatch, outcome):
+    """yt-dlp faked at its `YoutubeDL` boundary: records each read's options, answers `outcome`."""
+    seen: list[dict] = []
+
+    class FakeYDL:
+        def __init__(self, opts):
+            seen.append(opts)
+
+        def __enter__(self):
+            return self
+
+        def __exit__(self, *exc):
+            return False
+
+        def extract_info(self, url, download):
+            return outcome
+
+    monkeypatch.setattr(backfill.yt_dlp, "YoutubeDL", FakeYDL)
+    return seen
+
+
+def test_a_placeholder_with_nothing_behind_it_takes_the_signed_out_pages_text(monkeypatch):
+    seen = _fake_ydl(monkeypatch, _PUBLIC)
+    fb = adapters.FacebookAdapter()
+
+    info = backfill.public_title_info({"title": "Facebook", "description": None}, _URL, fb, 7)
+
+    assert fb.title(info) == "sorry not helping"
+    # One metadata-only read, and signed out: no cookie goes with it.
+    assert len(seen) == 1
+    assert seen[0]["skip_download"] is True
+    assert not seen[0].get("cookiefile")
+
+
+def test_a_real_title_reads_nothing_more(monkeypatch):
+    seen = _fake_ydl(monkeypatch, _PUBLIC)
+    info = {"title": "Funny Dark Humor Dad Joke!", "description": None}
+
+    assert backfill.public_title_info(info, _URL, adapters.FacebookAdapter(), 7) is info
+    assert seen == []
+
+
+def test_a_placeholder_the_signed_out_page_cannot_better_stays(monkeypatch):
+    _fake_ydl(monkeypatch, {"title": "Facebook", "description": None})
+    fb = adapters.FacebookAdapter()
+
+    info = backfill.public_title_info({"title": "Facebook", "description": None}, _URL, fb, 7)
+
+    assert fb.title(info) == "Facebook"
+
+
+def test_another_source_is_never_read_again(monkeypatch):
+    seen = _fake_ydl(monkeypatch, _PUBLIC)
+    info = {"title": "Video", "description": None}
+
+    assert backfill.public_title_info(info, _URL, adapters.SourceAdapter(), 7) is info
+    assert seen == []
+
+
+def test_the_backfill_retitles_a_placeholder_download_and_the_names_that_repeated_it(
+    db, make_user, monkeypatch
+):
+    _fake_ydl(monkeypatch, _PUBLIC)
+    owner, other = make_user(), make_user()
+    asset = MediaAsset(
+        source_kind="url",
+        source_ref=_URL,
+        format_sig="fb",
+        status="ready",
+        rel_path="facebook/x.mp4",
+        poster_path="facebook/x.jpg",
+        uploader_url="https://www.facebook.com/1",
+        title="Facebook",
+        ref_count=2,
+    )
+    db.add(asset)
+    db.flush()
+    copied = DownloadJob(
+        user_id=owner.id,
+        asset_id=asset.id,
+        source_kind="url",
+        source_ref=_URL,
+        status="done",
+        queue_pos=0,
+        profile_snapshot={},
+        display_name="Facebook",
+    )
+    renamed = DownloadJob(
+        user_id=other.id,
+        asset_id=asset.id,
+        source_kind="url",
+        source_ref=_URL,
+        status="done",
+        queue_pos=0,
+        profile_snapshot={},
+        display_name="Office jokes",
+    )
+    db.add_all([copied, renamed])
+    db.commit()
+
+    assert backfill.run()["retitled"] == 1
+
+    db.expire_all()
+    assert db.get(MediaAsset, asset.id).title == "sorry not helping"
+    assert db.get(DownloadJob, copied.id).display_name == "sorry not helping"
+    assert db.get(DownloadJob, renamed.id).display_name == "Office jokes"
+    # Run again: nothing left to retitle, and no page read for it.
+    assert backfill.run()["retitled"] == 0
+
+
+def test_a_download_titled_facebook_is_named_from_the_signed_out_page(
+    db, make_user, monkeypatch, tmp_path
+):
+    # The whole download path with yt-dlp and the media tools faked: the signed-in read gives only
+    # the placeholder, and the file and the download's name get the post's text.
+    _fake_ydl(monkeypatch, _PUBLIC)
+    monkeypatch.setattr(settings, "files_root", str(tmp_path))
+    monkeypatch.setattr(pipeline.urlguard, "check_url", lambda url: None)
+    monkeypatch.setattr(pipeline, "_ensure_phone_safe", lambda _j, produced, *_a: produced)
+    monkeypatch.setattr(
+        pipeline.probe, "resolve_media_facts", lambda produced, info: ("h264", "aac", 720, 1280, 9)
+    )
+    monkeypatch.setattr(pipeline.editmod, "thumbnail_to_jpg", lambda *a: None)
+    monkeypatch.setattr(pipeline.editmod, "ensure_poster", lambda *a: None)
+    owner = make_user()
+    asset = MediaAsset(
+        source_kind="url", source_ref=_URL, format_sig="fb", status="downloading", ref_count=1
+    )
+    db.add(asset)
+    db.flush()
+    job = DownloadJob(
+        user_id=owner.id,
+        asset_id=asset.id,
+        source_kind="url",
+        source_ref=_URL,
+        status="running",
+        queue_pos=0,
+        profile_snapshot={},
+    )
+    db.add(job)
+    db.commit()
+
+    def signed_in_read(job_id, asset_id, url, spec, staging, adapter, **kw):
+        f = staging / "1860682468422402.mp4"
+        f.write_bytes(b"v" * 100)
+        return {
+            "id": "1860682468422402",
+            "title": "Facebook",
+            "uploader": "Work dot exe",
+            "requested_downloads": [{"filepath": str(f)}],
+        }
+
+    monkeypatch.setattr(pipeline, "_extract_with_fallback", signed_in_read)
+    pipeline._download(job.id, asset.id, "url", _URL, {}, owner.id)
+
+    db.expire_all()
+    assert db.get(DownloadJob, job.id).status == "done"
+    assert db.get(MediaAsset, asset.id).title == "sorry not helping"
+    assert db.get(DownloadJob, job.id).display_name == "sorry not helping"

backend/tests/test_failure_lines.py

@@ -368,7 +368,7 @@ def test_the_metadata_backfill_names_the_asset_and_not_the_download(sink, monkey
             )
 
     monkeypatch.setattr(backfill.yt_dlp, "YoutubeDL", _Refusing)
-    assert backfill._extract_meta(url, adapters.YouTubeAdapter(), 11) is None
+    assert backfill.extract_meta(url, adapters.YouTubeAdapter(), 11) is None
 
     lines = [o for o in sink.objects() if "meta extract failed" in o["msg"]]
     assert lines and all("dQw4w9WgXcQ" not in json.dumps(o) for o in lines)
@@ -393,7 +393,7 @@ def test_the_metadata_backfill_takes_out_a_generic_name_whatever_it_holds(sink,
             raise yt_dlp.utils.DownloadError("ERROR: [generic] Our Trip: Day 1: Unable to download")
 
     monkeypatch.setattr(backfill.yt_dlp, "YoutubeDL", _Refusing)
-    assert backfill._extract_meta(url, adapters.SourceAdapter(), 11) is None
+    assert backfill.extract_meta(url, adapters.SourceAdapter(), 11) is None
 
     lines = [o for o in sink.objects() if "meta extract failed" in o["msg"]]
     assert lines and all("Day 1" not in json.dumps(o) for o in lines)

backend/tests/test_upload.py

@@ -596,6 +596,7 @@ DISPLAY_WRITE_SITES = {
     ("downloads/service.py", "enqueue_clip"): {"display_name or asset.title[:DISPLAY_TEXT_MAX]"},
     ("downloads/pipeline.py", "_download"): {"meta.title[:DISPLAY_TEXT_MAX]"},
     ("downloads/pipeline.py", "_process_clip"): {"named[:DISPLAY_TEXT_MAX]"},
+    ("downloads/backfill.py", "_retitle"): {"named[:DISPLAY_TEXT_MAX]"},
     ("worker.py", "_finish_from_cache"): {"asset.title[:DISPLAY_TEXT_MAX]"},
     ("downloads/upload.py", "ingest_upload"): {"display_text(display_name)"},
     ("routes/auth.py", "callback"): {"display_text(userinfo.get('name'))"},