fix(pixiv): capture ugoira frame timings in the post record (ordering bug)

The core writes the post record BEFORE extract_media, but the ugoira frame
delays were only memoized DURING extract_media — so write_post_record never saw
them and ugoira_frames was always empty in the record. Extract a memoized
_ugoira_meta (frames + zip url share ONE /v1/ugoira/metadata call regardless of
order) and inject client.fetch_ugoira_frames into the downloader (mirrors
Patreon's content_fetcher) so write_post_record populates the frames itself.
Zero extra API calls — the fetch is shared/memoized with extract_media. A
recapture now backfills the timings onto existing ugoira posts.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM

tests/test_pixiv_client.py

@@ -200,6 +200,38 @@ def test_extract_media_ugoira_zip_swap(client, page1, monkeypatch):
     assert post["_work"]["_ugoira_frames"] == frames
 
 
+def test_fetch_ugoira_frames_memoizes_and_shares_one_call(client, page1, monkeypatch):
+    # fetch_ugoira_frames (called by write_post_record, which the core runs
+    # BEFORE extract_media) populates frames; extract_media then REUSES the
+    # memoized metadata — exactly one /v1/ugoira/metadata call total.
+    frames = [{"file": "000000.jpg", "delay": 90}]
+    calls = []
+
+    def fake_call(endpoint, params):
+        calls.append(endpoint)
+        return {"ugoira_metadata": {
+            "zip_urls": {"medium": (
+                "https://i.pximg.net/img-zip-ugoira/x/333_ugoira600x600.zip"
+            )},
+            "frames": frames,
+        }}
+
+    monkeypatch.setattr(client, "_call", fake_call)
+    post = _post_for(client, page1, 333)
+    client.fetch_ugoira_frames(post)
+    assert post["_work"]["_ugoira_frames"] == frames
+    items = client.extract_media(post, {})          # reuses memoized meta
+    assert items[0].media_id == "ugoira"
+    assert calls == ["/v1/ugoira/metadata"]          # ONE fetch, not two
+
+
+def test_fetch_ugoira_frames_noop_for_non_ugoira(client, page1, monkeypatch):
+    monkeypatch.setattr(client, "_call", lambda e, p: pytest.fail("should not fetch"))
+    post = _post_for(client, page1, 222)             # a single-page illust
+    client.fetch_ugoira_frames(post)
+    assert "_ugoira_frames" not in post["_work"]
+
+
 def test_extract_media_ugoira_metadata_failure_downgrades(
     client, page1, monkeypatch
 ):