feat(patreon): scrape creator-page HTML as a campaign-id resolution fallback

Patreon's /api/campaigns?filter[vanity]= lookup returns empty data for creators
that plainly exist (operator-flagged 2026-06-06 — Atole etc. erroring at the
resolve step). gallery-dl never used that endpoint; it pulls the campaign id out
of the creator page's bootstrap JSON. Add the same as a fallback: when the API
misses, GET the creator page (bare + /c/ vanity paths) and scrape the first
campaign id from any known embedding ("id":"…","type":"campaign" /
"campaign":{"data":{"id" / /api/campaigns/<id> / "campaign_id"). API is still
tried first (cheap, structured); the page scrape only runs on a miss.

Tests: API-empty → page-scrape fallback resolves; _scrape_campaign_id pattern
coverage. Existing API-path tests unchanged (happy paths short-circuit before
the fallback; failure paths hit the guarded scrape and still return None).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

backend/app/services/patreon_resolver.py

@@ -41,6 +41,19 @@ _USER_AGENT = (
 )
 _TIMEOUT_SECONDS = 10.0
 
+# Fallback resolution: Patreon's `/api/campaigns?filter[vanity]=` lookup has
+# proven unreliable (returns empty `data` for creators that clearly exist —
+# operator-flagged 2026-06-06). gallery-dl never used that endpoint; it scrapes
+# the campaign id out of the creator page's bootstrap JSON. We do the same as a
+# fallback: fetch the creator page HTML and pull the first campaign id out of
+# any of these embeddings (ordered most- to least-specific).
+_PAGE_CAMPAIGN_ID_PATTERNS = (
+    re.compile(r'"id":\s*"(\d+)",\s*"type":\s*"campaign"'),
+    re.compile(r'"campaign":\s*\{\s*"data":\s*\{\s*"id":\s*"(\d+)"'),
+    re.compile(r"/api/campaigns/(\d+)"),
+    re.compile(r'"campaign_id":\s*"?(\d+)'),
+)
+
 
 def _load_cookie_jar(cookies_path: str | None) -> http.cookiejar.MozillaCookieJar | None:
     if not cookies_path or not os.path.isfile(cookies_path):
@@ -55,6 +68,15 @@ def _load_cookie_jar(cookies_path: str | None) -> http.cookiejar.MozillaCookieJa
 
 
 def _sync_lookup(vanity: str, cookies_path: str | None) -> str | None:
+    """Resolve a vanity to a campaign id: try the campaigns API first (cheap,
+    structured), then fall back to scraping the creator page (robust against the
+    API's empty-data failures). Returns None only when both miss."""
+    return _lookup_via_api(vanity, cookies_path) or _lookup_via_page(
+        vanity, cookies_path
+    )
+
+
+def _lookup_via_api(vanity: str, cookies_path: str | None) -> str | None:
     jar = _load_cookie_jar(cookies_path)
     headers = {
         "User-Agent": _USER_AGENT,
@@ -102,6 +124,50 @@ def _sync_lookup(vanity: str, cookies_path: str | None) -> str | None:
     return campaign_id
 
 
+def _scrape_campaign_id(html: str) -> str | None:
+    """First campaign id found in creator-page HTML via the known embeddings."""
+    if not isinstance(html, str):
+        return None
+    for pat in _PAGE_CAMPAIGN_ID_PATTERNS:
+        m = pat.search(html)
+        if m:
+            return m.group(1)
+    return None
+
+
+def _lookup_via_page(vanity: str, cookies_path: str | None) -> str | None:
+    """Fallback: GET the creator page and scrape the campaign id from the page
+    bootstrap (gallery-dl's method). Tries both the bare and `/c/` vanity paths
+    Patreon redirects between. Never raises."""
+    jar = _load_cookie_jar(cookies_path)
+    headers = {"User-Agent": _USER_AGENT, "Accept": "text/html"}
+    for page_url in (
+        f"https://www.patreon.com/{vanity}",
+        f"https://www.patreon.com/c/{vanity}",
+    ):
+        try:
+            resp = requests.get(
+                page_url,
+                headers=headers,
+                cookies=jar,
+                timeout=_TIMEOUT_SECONDS,
+                allow_redirects=True,
+            )
+        except requests.RequestException as exc:
+            log.warning("Patreon creator-page fetch failed for %s: %s", page_url, exc)
+            continue
+        if resp.status_code != 200:
+            continue
+        campaign_id = _scrape_campaign_id(resp.text)
+        if campaign_id:
+            log.info(
+                "Resolved Patreon vanity=%s → campaign_id=%s via creator page",
+                vanity, campaign_id,
+            )
+            return campaign_id
+    return None
+
+
 async def resolve_campaign_id(
     vanity: str,
     cookies_path: str | None,

tests/test_patreon_resolver.py

@@ -91,6 +91,41 @@ async def test_loads_cookies_file_when_present(tmp_path):
     assert kwargs.get("cookies") is not None
 
 
+@pytest.mark.asyncio
+async def test_falls_back_to_creator_page_when_api_empty():
+    """When the campaigns API returns empty data (its known failure mode), the
+    resolver scrapes the campaign id out of the creator page (gallery-dl's
+    method)."""
+    api_resp = MagicMock()
+    api_resp.status_code = 200
+    api_resp.json.return_value = {"data": []}
+    page_resp = MagicMock()
+    page_resp.status_code = 200
+    page_resp.text = (
+        '<html><script>window.patreon={"bootstrap":{"campaign":'
+        '{"data":{"id":"7654321","type":"campaign"}}}}</script></html>'
+    )
+
+    def _get(url, **kwargs):
+        return api_resp if "/api/campaigns" in url else page_resp
+
+    with patch(
+        "backend.app.services.patreon_resolver.requests.get", side_effect=_get
+    ):
+        result = await resolve_campaign_id("alice", cookies_path=None)
+    assert result == "7654321"
+
+
+def test_scrape_campaign_id_patterns():
+    from backend.app.services.patreon_resolver import _scrape_campaign_id
+
+    assert _scrape_campaign_id('foo /api/campaigns/424242 bar') == "424242"
+    assert _scrape_campaign_id('"campaign_id":"99"') == "99"
+    assert _scrape_campaign_id('{"id":"5","type":"campaign"}') == "5"
+    assert _scrape_campaign_id("no id here") is None
+    assert _scrape_campaign_id(None) is None
+
+
 # -- resolve_campaign_id_for_source (shared by download + verify) ----------