fix(pfs): correct CY2026 final-rule FR citation (90 FR 49266) + citation/robustness cleanups (refs #591 #593 #595 #597)

notebooks/cy2026_pfs_proposed_rule.py

@@ -20,7 +20,7 @@ def _(mo):
     (CMS-1832-P, 90 FR 32352, July 16, 2025) proposed the **first-ever
     split conversion factor** — separate dollar amounts for clinicians who
     qualify as Advanced APM Participants (QPs) versus everyone else — and
-    the Final Rule (90 FR 101174) changed several of those numbers again
+    the Final Rule (90 FR 49266) changed several of those numbers again
     before taking effect.
 
     **Mechanism:** MACRA 2015 (section 1848(d)(20) of the Act) sets two
@@ -142,7 +142,7 @@ def _(alt, cf_walk, mo):
             `pfs.rules` registry yet.*
 
             **Sources:** CY2025 — 89 FR 98452. CY2026 NPRM — 90 FR 32352
-            (CMS-1832-P). CY2026 Final Rule — 90 FR 101174.
+            (CMS-1832-P). CY2026 Final Rule — 90 FR 49266.
             """),
         ]
     )
@@ -261,14 +261,15 @@ def _(mo):
     mo.md("""
     ## 2. RVU-Level Deltas — Which HCPCS Codes Move the Most
 
-    `pfs.rvu_proposed` (14,169 rows, CMS-1832-P) is compared against
+    `pfs.rvu_proposed` (14,169 rows as of this ingest, CMS-1832-P) is
-    `pfs.rvu` for CY2025 (pre-rule baseline) and CY2026 (what actually got
+    compared against `pfs.rvu` for CY2025 (pre-rule baseline) and CY2026
-    finalized). Both `pfs.rvu` and `pfs.rvu_proposed` are deduplicated to
+    (what actually got finalized). Both `pfs.rvu` and `pfs.rvu_proposed`
-    one row per HCPCS base code (no modifier), `status_code = 'A'`
+    are deduplicated to one row per HCPCS base code (no modifier),
-    (actively priced), and — for the proposed table specifically — a
+    `status_code = 'A'` (actively priced), and — for the proposed table
-    non-null non-facility PE RVU, since ~29% of `pfs.rvu_proposed` rows
+    specifically — a non-null non-facility PE RVU, since ~29% of
-    carry a null non-facility *or* facility PE RVU (CMS's Addendum B only
+    `pfs.rvu_proposed` rows (as of this ingest) carry a null
-    populates the setting a code is actually priced in).
+    non-facility *or* facility PE RVU (CMS's Addendum B only populates
+    the setting a code is actually priced in).
     """)
     return
 
@@ -737,8 +738,8 @@ def _(alt, mo, qp_differential):
         [
             differential_chart,
             mo.md("""
-            **Sources:** RVUs — `pfs.rvu`, CY2026 (90 FR 101174). CFs —
+            **Sources:** RVUs — `pfs.rvu`, CY2026 (90 FR 49266). CFs —
-            `pfs.rules.RULES[2026]` (final, 90 FR 101174) and
+            `pfs.rules.RULES[2026]` (final, 90 FR 49266) and
             `RULES[2026].proposed` (NPRM, 90 FR 32352). National unadjusted
             means GPCI = 1.0 — actual payment varies by locality.
             """),

src/llm/source.py

@@ -125,7 +125,7 @@ def _attachment_text(store: Store, item_key: str) -> str:
     return "\n\n".join(parts)
 
 
-_TAG_RE = re.compile(r"<[^>]+>")
+_TAG_RE = re.compile(r"</?[A-Za-z][^>]*>")
 
 
 def _strip_html_wrapper(text: str) -> str:
@@ -138,12 +138,15 @@ def _strip_html_wrapper(text: str) -> str:
 
     The FR text itself also carries inline HTML — e.g. Cloudflare's
     obfuscated ``<a href="/cdn-cgi/l/email-protection#...">`` mailto
-    anchors and page-break anchors. Those are stripped with a plain
+    anchors and page-break anchors. Those are stripped with a
-    ``<[^>]+>`` regex (this is a controlled, known corpus — not
+    ``</?[A-Za-z][^>]*>`` regex (this is a controlled, known corpus —
-    attacker-supplied HTML — so no HTML-parsing dependency is needed),
+    not attacker-supplied HTML — so no HTML-parsing dependency is
-    keeping the anchor's inner text. HTML entities (``&amp;`` etc.) are
+    needed), keeping the anchor's inner text. The tag must start with a
-    then unescaped. Plain-text markers like ``[[Page NNNNN]]`` contain no
+    letter (optionally preceded by ``/``) so free-text comparison
-    angle brackets and pass through untouched.
+    fragments like ``p<0.05`` or ``x>2`` are never mistaken for markup.
+    HTML entities (``&amp;`` etc.) are then unescaped. Plain-text
+    markers like ``[[Page NNNNN]]`` contain no angle brackets and pass
+    through untouched.
     """
     idx = text.find("<pre>")
     if idx == -1:
@@ -173,7 +176,9 @@ def _rule_text(store: Store, item_key: str) -> str:
     for (storage_path,) in rows:
         path = Path(storage_path)
         if path.suffix.lower() == ".txt" and path.exists():
-            return _strip_html_wrapper(path.read_text())
+            return _strip_html_wrapper(
+                path.read_text(encoding="utf-8", errors="replace")
+            )
     return _attachment_text(store, item_key)
 
 

src/pfs/rules/__init__.py

@@ -418,7 +418,13 @@ RULES: dict[int, RuleYear] = {
         budget_neutrality_adjustor=1.0049,  # +0.49% BN adjustment
         clinical_labor_rate_update=True,
         telehealth_originating_site_fee=30.38,
-        federal_register_citation="90 FR 101174",
+        # Corrected from "90 FR 101174" (mistranscribed; no such page in
+        # this rule). The Final Rule (RIN 0938-AV50, Vol. 90 No. 212,
+        # Nov 5, 2025) runs 90 FR 49266-50481; per this registry's
+        # convention (see other years above) the citation is the rule's
+        # first page. The CF statement itself ($33.5675 QP / $33.4009
+        # non-QP) is at the pincite 90 FR 49961.
+        federal_register_citation="90 FR 49266",
         # NPRM values (pre-final) per :pincite:`2KVJ2HKX` — CMS-1832-P,
         # 90 FR 32352, transcribed from data/fr_downloads/2025-13271.txt.
         proposed=ProposedRule(

src/qpp/__init__.py

@@ -258,7 +258,10 @@ class QppYear(BaseModel):
 #     the same NPRM) — it reads as stale boilerplate describing the
 #     pre-2023 -> 2023 statutory step-up, not a change CMS-1832-P
 #     itself proposes.
-#   - Final Rule 90 FR 49928 (line 69796-69798) drops the "increase
+#   - Final Rule 90 FR 49980 (line 69796-69798; corrected from a
+#     mistranscribed "90 FR 49928" — that page marker is at line 64990,
+#     these lines fall between the [[Page 49980]] marker at line 69792
+#     and the [[Page 49981]] marker at line 69913) drops the "increase
 #     from" framing entirely: "The thresholds to achieve QP status in
 #     the 2026 QP Performance Period (2028 payment year) are set to
 #     75 percent for the payment amount, and 50 percent for the
@@ -373,7 +376,12 @@ QPP: dict[int, QppYear] = {
         risk_standards=_risk_standards(),
         apm_incentive_pct=None,
         qp_cf_applies=True,
-        citation="90 FR 49928",
+        # Corrected from "90 FR 49928" alone: the QP threshold statement
+        # itself is at 90 FR 49980 (see docstring pincite above); 49928
+        # is retained alongside it since it's independently the source
+        # for the Partial QP payment-amount confirmation (docstring
+        # pincite, line 65021-65026).
+        citation="90 FR 49928, 49980",
         # NPRM proposals per :pincite:`2KVJ2HKX` — CMS-1832-P, 90 FR
         # 32352, transcribed from data/fr_downloads/2025-13271.txt;
         # disposition from data/fr_downloads/2025-19787.txt.

tests/llm/test_source.py

@@ -152,6 +152,7 @@ class TestRuleDocs:
             "[[Page 12345]]\n"
             'Contact us at <a href="/cdn-cgi/l/email-protection#x">'
             "someone</a> regarding CMS &amp; PFS.\n"
+            "A finding was significant (p<0.05 and >2 cm).\n"
             "</pre></body></html>"
         )
         store.attach_file(rule_key, txt)
@@ -165,6 +166,7 @@ class TestRuleDocs:
         assert "someone" in text
         assert "CMS & PFS" in text
         assert "[[Page 12345]]" in text
+        assert "p<0.05 and >2 cm" in text
 
     def test_pdf_only_falls_back_to_extract_attachment(self, store, rule_key, tmp_path):
         pdf = tmp_path / "rule.pdf"

tests/pfs/test_rules.py

@@ -149,7 +149,7 @@ class TestSpecificValues:
         assert cy2026.telehealth_originating_site_fee == 30.38
 
     def test_cy2026_federal_register(self):
-        assert cy2026.federal_register_citation == "90 FR 101174"
+        assert cy2026.federal_register_citation == "90 FR 49266"
 
     def test_cy2026_clinical_labor_update(self):
         assert cy2026.clinical_labor_rate_update is True