qa/_text_norm: canonical stem_for_match (dedup query.py + corpus.py)

Phase 1 step 2 of #53. Hoist the trailing-s stemmer to a single home
in arborist.qa._text_norm so query.py, corpus.py, and source_roles.py
all use the same implementation.

Before:
  - query.py:_stem_token_for_match — `len>4 and endswith('s') and not
    endswith('ss')` (apostrophes assumed pre-stripped by _TITLE_TOKEN_RE)
  - corpus.py:apply_title_boost._stem (inline) — same length/suffix
    check PLUS apostrophe strip ("'", "’")

Behaviorally compatible when inputs are pre-stripped, but the dual
implementations were a drift risk waiting to bite. The apostrophe-safe
version (corpus.py's) is the canonical now — handles raw title text
without an upstream sanitizer, no behavior change for the pre-stripped
call sites.

Re-exports in query.py + import-update in corpus.py + source_roles.py
keep every existing caller working. 257 query/corpus/sidecar/wallet/
bucket/claim_lattice tests pass.

Defers lifting _title_query_tokens (and its 5-fold variant helpers —
hyphen, numeral, accent, honorific, brit) for later: those carry
years of bench-tuned hot-loop optimization and a TITLE_TOKEN_POLICY
slug threaded into run-DAG provenance. source_roles still lazy-imports
_title_query_tokens from query.py; no change there.

arborist/qa/_text_norm.py

@@ -82,3 +82,23 @@ def tokenize_text(text: str) -> list[str]:
             continue
         out.append(t)
     return out
+
+
+def stem_for_match(t: str) -> str:
+    """Light suffix-strip for query-token vs title/body matching.
+
+    Two normalizations rolled into one trailing-s strip:
+      - possessive  "superman's" → "supermans" → "superman"
+                    (apostrophe stripped first; ASCII and curly quotes
+                    handled — fold safely on tokens that pre-strippers
+                    don't reach, e.g. raw title text)
+      - plural      "powers" → "power", "girlfriends" → "girlfriend"
+
+    Conservative: only fires on tokens > 4 chars (preserves "is", "as",
+    "us") and skips ``ss``-enders ("class", "moss"). Idempotent —
+    stemming an already-stemmed token is a no-op.
+    """
+    t = t.replace("'", "").replace("’", "")
+    if len(t) > 4 and t.endswith("s") and not t.endswith("ss"):
+        return t[:-1]
+    return t

arborist/qa/corpus.py

@@ -89,15 +89,9 @@ def apply_title_boost(
         return hits
     from arborist.qa._text_norm import (
         _WORD_RE, STOPWORDS,
-        fold_accents, numeral_expand, tokenize_text,
+        fold_accents, numeral_expand, stem_for_match as _stem, tokenize_text,
     )
 
-    def _stem(t: str) -> str:
-        t = t.replace("'", "").replace("’", "")
-        if len(t) > 4 and t.endswith("s") and not t.endswith("ss"):
-            return t[:-1]
-        return t
-
     query_stems = numeral_expand({_stem(t) for t in tokenize_text(query)})
     if not query_stems:
         return hits

arborist/qa/query.py

@@ -1048,23 +1048,10 @@ def _docs_with_core_keyword_match(
     return rows
 
 
-def _stem_token_for_match(t: str) -> str:
-    """Light suffix-strip for query-token vs body matching.
-
-    Two normalizations:
-        possessive  ``"superman's" -> "supermans" -> "superman"`` (the apostrophe
-                    is already gone via _TITLE_TOKEN_RE; we drop the trailing
-                    ``s`` here so the lookup matches plain ``superman`` in body).
-        plural      ``"powers" -> "power"``, ``"girlfriends" -> "girlfriend"``
-                    so plural questions match singular source mentions.
-
-    Both are the same operation: strip trailing ``s`` for tokens > 4 chars.
-    Conservative on short tokens (``"is"``, ``"as"``, ``"us"`` would lose
-    meaning) and on tokens that don't end in ``s`` (no-op).
-    """
-    if len(t) > 4 and t.endswith("s") and not t.endswith("ss"):
-        return t[:-1]
-    return t
+# Canonical stemmer lives in _text_norm.py so query.py + corpus.py +
+# source_roles.py share one implementation. Re-exported under the old
+# name so existing call sites (and tests) keep working.
+from arborist.qa._text_norm import stem_for_match as _stem_token_for_match  # noqa: E402
 
 
 def _body_count_with_stem(body: str, t: str) -> int:

arborist/qa/source_roles.py

@@ -100,12 +100,13 @@ def classify_source_role(
         return "sequel_background_source"
     if any(k in t for k in _SECONDARY_TITLE_MARKERS):
         return "secondary_context_source"
-    # Lazy import — `_title_query_tokens` + `_stem_token_for_match`
-    # live in query.py today; they move to _text_norm.py in a later
-    # Phase 1 step. Import-cycle-safe by deferring to call time.
-    from arborist.qa.query import _title_query_tokens, _stem_token_for_match
+    # _title_query_tokens still lives in query.py (full fold stack is
+    # not lifted yet); the stemmer is now canonical in _text_norm.
+    # Lazy import keeps the import path acyclic.
+    from arborist.qa.query import _title_query_tokens
+    from arborist.qa._text_norm import stem_for_match
     title_tokens = _title_query_tokens(t.replace("_", " "))
-    title_stems = {_stem_token_for_match(tok) for tok in title_tokens}
+    title_stems = {stem_for_match(tok) for tok in title_tokens}
     if qtokens_stem and len(title_stems & qtokens_stem) >= max(
         1, len(qtokens_stem) - 1
     ):