Cache the category tree walk (fixes ~1s File Management page load)

_build_category_index() was the one tree-scanning function never
wrapped in the app's existing 5-minute cache - it ran fresh on every
File Management page load and Sort Unsorted scan (not a manual "this
may take a moment" action like Duplicate Courses or Storage Usage), so
the cost was invisible until it was already slow. Measured ~1.0-1.1s
consistently against the live NAS-mounted library at 161 courses,
confirmed via profiling that a single call makes dozens of iterdir()
round-trips - each one a network hop over SMB. Now shares the same
cache_get_or_compute pattern and invalidate_cache() call sites as
get_all_course_dirs(), so no new invalidation logic was needed.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

README.md

@@ -137,8 +137,12 @@ silently stay blank instead of erroring.
   ones that recur across many categories (a prolific creator's name, etc.)
   so they can't outvote a genuinely specific word just by sharing more of
   them.
-  - *Refresh Library*: manually bypasses the 5-minute filesystem-scan cache,
+  - *Refresh Library*: manually bypasses the 5-minute filesystem-scan cache
-  for when files were added/removed directly on disk.
+  (this now includes the category tree Sort Unsorted/Manage Library's
+  picker builds - previously rebuilt on every page load, a full,
+  uncached walk of the whole library that got noticeably slow over a
+  NAS-mounted (SMB) library as the course count grew), for when files
+  were added/removed directly on disk.
   - *Bulk Rename*: find & replace across every course/folder name in the
   library at once, with a per-match preview and the ability to drop
   individual matches before applying. Three match modes: plain text

VERSION

@@ -1 +1 @@
-2026-08-24 18:40 UTC — search, bulk actions, undo, storage usage
+2026-08-24 21:41 UTC — cache category tree walk (fixes ~1s picker lag)

offlineu_core.py

@@ -1445,7 +1445,18 @@ def _build_category_index() -> List[Dict[str, Any]]:
     courses and their non-lecture contents never show up as if they were
     categories to file things under. The Unsorted folder itself is
     excluded - it's the source, never a valid destination.
+
+    This walk visits every directory in the library and, over a
+    NAS-mounted (SMB) library, each of those is a network round-trip -
+    the same reasoning behind get_all_course_dirs()'s cache, so this
+    result is cached the same way (same 5-minute TTL, same
+    invalidate_cache() call sites already busting it - no separate
+    invalidation needed): otherwise this ran fresh on every File
+    Management page load and Sort Unsorted scan, neither of which is a
+    manual "this may take a moment" action like Duplicate Courses or
+    Storage Usage, so the cost was invisible until it was already slow.
     """
+    def compute() -> List[Dict[str, Any]]:
         library_root = Path(get_library_root())
         try:
             unsorted_root = (library_root / UNSORTED_FOLDER_NAME).resolve()
@@ -1494,6 +1505,8 @@ def _build_category_index() -> List[Dict[str, Any]]:
         walk(library_root, [])
         return categories
 
+    return cache_get_or_compute('category_index', compute)
+
 
 def _bonus_token_frequency(categories: List[Dict[str, Any]]) -> Dict[str, int]:
     """How many distinct categories each bonus token shows up in, across