fix(places): a locality collision must not break the sitemap

Sitemap regeneration failed on staging. 'richmond' in the curated locality
list collides with the GIAS town Richmond in North Yorkshire (37 schools), the
registry raised, and the admin endpoint 500d — taking down sitemap generation
for all 25,000 school pages over one bad row of curated data.

The guard now skips the colliding locality and logs an error. Skipping still
achieves what the guard was for — a locality never silently shadows a town —
without letting curated data break the site. That matters beyond this bug:
GIAS town names change with no code change here, so a raise could fire
spontaneously in production later.

Also removes four localities that were London boroughs rather than districts.
Hackney, Islington, Greenwich and Ealing are local authorities with 104, 72,
108 and 115 schools and already have authority pages; a locality defined by
two or three outcodes would have been a partial near-duplicate of one — the
thin-content failure the two-namespace design exists to avoid. A test now
guards the whole borough list.

Validated against the live corpus: 15 localities, no town collisions, no
authority duplicates, all 15 clear the threshold.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mWQnpye9F299NVRCCSRvj

backend/localities.py

@@ -20,6 +20,18 @@ A locality whose outcodes hold fewer than MIN_SCHOOLS schools is not
 published, so a typo produces no page rather than an empty one. Places that
 fail that check are logged at startup, because a locality you meant to publish
 quietly not appearing is the failure worth hearing about.
+
+Two rules for anything added here.
+
+**Sub-borough districts only.** A London borough is a local authority and
+already has a page at /schools/authority/[la] covering all of its schools; a
+locality defined by two or three outcodes would be a partial, near-duplicate
+subset of it. Hackney, Islington, Greenwich and Ealing were all in the first
+draft for that reason and have been removed.
+
+**The slug must not match a GIAS town.** "Richmond" did — GIAS has a Richmond
+in North Yorkshire with 37 schools — so the London one could never publish.
+The registry skips any locality that collides and logs it.
 """
 
 # slug -> (display name, outcodes)
@@ -30,16 +42,11 @@ LOCALITY_OUTCODES: dict[str, tuple[str, tuple[str, ...]]] = {
     "shoreditch":    ("Shoreditch",    ("EC2A", "E1")),
     "peckham":       ("Peckham",       ("SE15",)),
     "brixton":       ("Brixton",       ("SW2", "SW9")),
-    "hackney":       ("Hackney",       ("E5", "E8", "E9")),
-    "islington":     ("Islington",     ("N1", "N5", "N7")),
     "camden-town":   ("Camden Town",   ("NW1",)),
-    "greenwich":     ("Greenwich",     ("SE10",)),
     "wimbledon":     ("Wimbledon",     ("SW19",)),
     "putney":        ("Putney",        ("SW15",)),
     "fulham":        ("Fulham",        ("SW6",)),
     "chiswick":      ("Chiswick",      ("W4",)),
-    "ealing":        ("Ealing",        ("W5", "W13")),
-    "richmond":      ("Richmond",      ("TW9", "TW10")),
     "stratford":     ("Stratford",     ("E15",)),
     "walthamstow":   ("Walthamstow",   ("E17",)),
     "tooting":       ("Tooting",       ("SW17",)),