From f168bece00b2820f1bc76cc7e854dc38abd6096e Mon Sep 17 00:00:00 2001
From: pax <paxxe@protonmail.com>
Date: Thu, 9 Apr 2026 19:53:20 -0500
Subject: [PATCH] category_fetcher: fix _do_ensure to try batch API when not
 yet probed
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

_do_ensure only tried the batch API when _batch_api_works was True,
but after removing the search-time prefetch (where the probe used
to run), _batch_api_works stayed None forever. Gelbooru's only
viable path IS the batch API (its post-view HTML has no tag links),
so clicks on Gelbooru posts produced zero categories.

Fix: _do_ensure now tries the batch API when _batch_api_works is
not False (i.e., both True and None). When None, the call doubles
as an inline probe: if the batch produced categories, save True;
if nothing useful came back, save False and fall to HTML.

This is simpler than the old prefetch_batch probe because it runs
on ONE post at a time — no batch/HTML mixing concerns, no "single
path per invocation" rule. The probe result is persisted to DB so
it only fires once per site ever.

Dispatch matrix in _do_ensure:
  _batch_api_works True  + auth → batch API (Gelbooru proper)
  _batch_api_works None  + auth → batch as probe → True or False
  _batch_api_works False        → HTML scrape (Rule34)
  no auth                       → HTML scrape (Safebooru.org)
  transient error               → stays None, retry next click

Verified all three sites from clean cache: Gelbooru 55/56+49/50
(batch), Rule34 40/40+38/38 (HTML), Safebooru.org 47/47+47/47
(HTML).
---
 booru_viewer/core/api/category_fetcher.py | 35 ++++++++++++++++++-----
 1 file changed, 28 insertions(+), 7 deletions(-)

diff --git a/booru_viewer/core/api/category_fetcher.py b/booru_viewer/core/api/category_fetcher.py
index 3319f0f..7e3a81b 100644
--- a/booru_viewer/core/api/category_fetcher.py
+++ b/booru_viewer/core/api/category_fetcher.py
@@ -332,13 +332,34 @@ class CategoryFetcher:
             self._inflight.pop(post.id, None)
 
     async def _do_ensure(self, post: "Post") -> None:
-        """Inner dispatch for ensure_categories."""
-        # Batch API path (for single-post ensure, e.g. click or save)
-        if self._batch_api_works is True and self._batch_api_available():
-            await self.fetch_via_tag_api([post])
-            if post.tag_categories:
-                return
-        # HTML fallback
+        """Inner dispatch for ensure_categories.
+
+        Tries the batch API when it's known to work (True) OR not yet
+        probed (None). The result doubles as an inline probe: if the
+        batch produced categories, it works (save True); if it
+        returned nothing useful, it's broken (save False). Falls
+        through to HTML scrape as the universal fallback.
+        """
+        if self._batch_api_works is not False and self._batch_api_available():
+            try:
+                await self.fetch_via_tag_api([post])
+            except Exception as e:
+                log.debug("Batch API ensure failed (transient): %s", e)
+                # Leave _batch_api_works at None → retry next call
+            else:
+                if post.tag_categories:
+                    if self._batch_api_works is None:
+                        self._batch_api_works = True
+                        self._save_probe_result(True)
+                    return
+                # Batch returned nothing → broken API (Rule34) or
+                # the specific post has only unknown tags (very rare).
+                if self._batch_api_works is None:
+                    self._batch_api_works = False
+                    self._save_probe_result(False)
+        # HTML scrape fallback (works on Rule34/Safebooru.org/Moebooru,
+        # returns empty on Gelbooru proper which is fine because the
+        # batch path above covers Gelbooru)
         await self.fetch_post(post)
 
     # ----- dispatch: prefetch (batch, fire-and-forget) -----