fix: rekey 10,236 invalid bib.sqlite keys to Zotero charset, convert CCLF cites

Two changes:

1. Rekeyed all bib.sqlite items/attachments with chars outside the
   Zotero charset (23456789ABCDEFGHIJKLMNPQRSTUVWXYZ). Chars 0, 1, O, L
   are not valid in Zotero keys. 10,236 items + 1,881 attachments fixed.

2. Converted all 30 CCLF IP Section freeform citations in
   aco/express/cclf.py to structured :pincite: directives:
     Before: CCLF IP Section 3.1 "Matching MBIs" (p.14):
     After:  :pincite:`VTWVB384 §3.1 p.14` — Matching MBIs:

Updated tests to verify pincite parsing instead of meta: tag extraction
from the now-converted docstrings.

fixes #302

src/aco/express/cclf.py

@@ -85,7 +85,7 @@ def stg_beneficiary_xref(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 3.1 "Matching MBIs" (p.14):
+    :pincite:`VTWVB384 §3.1 p.14` — Matching MBIs:
 
       "A beneficiary's MBI is unique to that beneficiary but may change
       over time. CCLFs 1 through 9, A, and B, as they are provided each
@@ -94,7 +94,7 @@ def stg_beneficiary_xref(
       can be used each month to cross reference records associated with the
       current MBI to records associated with previous MBIs."
 
-    CCLF IP Section 5.1.1 "Creation of the Most Recent MBI field" (p.19):
+    :pincite:`VTWVB384 §5.1.1 p.19` — Creation of the Most Recent MBI field:
 
       "The beneficiary XREF file (CCLF9) contains a complete history of
       MBIs ever used to identify a beneficiary, for beneficiaries who have
@@ -129,7 +129,7 @@ def stg_institutional_header(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.1 "Part A Claims Header File" (p.8-9):
+    :pincite:`VTWVB384 §2.2.1 p.8-9` — Part A Claims Header File:
 
       "The Part A Claims Header File (CCLF1) contains header-level claim
       information for institutional services covered by Part A and some
@@ -141,7 +141,7 @@ def stg_institutional_header(
     MBI resolution: BENE_MBI_ID is joined to CCLF9._stg_beneficiary_xref
     on PRVS_NUM to resolve to CURRENT_BENE_MBI_ID via COALESCE.
 
-    CCLF IP Section 5.3.1 "Calculating Total Part A and B Expenditures" (p.23):
+    :pincite:`VTWVB384 §5.3.1 p.23` — Calculating Total Part A and B Expenditures:
 
       "Identify the canceled claims in the Part A Header file. These claims
       are identified by CLM_ADJSMT_TYPE_CD=1. Change the sign of the
@@ -179,7 +179,7 @@ def int_institutional_header_adr(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 4.4 "Debit/Credit Method" (p.19):
+    :pincite:`VTWVB384 §4.4 p.19` — Debit/Credit Method:
 
       "The Debit/Credit method gives a full account/history of claims
       processed over time. The variable CLM_ADJSMT_TYPE_CD identifies:
@@ -187,12 +187,12 @@ def int_institutional_header_adr(
         1 = Cancellation Claim
         2 = Adjustment Claim (adjustment to an original claim)"
 
-    CCLF IP Section 5.1.2 "Natural Key for Part A" (p.20):
+    :pincite:`VTWVB384 §5.1.2 p.20` — Natural Key for Part A:
 
       Part A natural key: CLM_BLG_PRVDR_OSCAR_NUM, CLM_FROM_DT,
       CLM_THRU_DT, Most Recent MBI (MR_MBI).
 
-    CCLF IP Section 5.2.1 "Related Claims in Part A Header File" (p.21-22):
+    :pincite:`VTWVB384 §5.2.1 p.21-22` — Related Claims in Part A Header File:
 
       "This group of claims consists of records that are all related to one
       another — they all involve a claim for services provided to a
@@ -235,7 +235,7 @@ def stg_revenue_center(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.2 "Part A Claim Revenue Center Detail File" (p.9):
+    :pincite:`VTWVB384 §2.2.2 p.9` — Part A Claim Revenue Center Detail File:
 
       "The Part A Claims Revenue Center Detail File (CCLF2) contains
       line-item level detail for each claim from the Part A Claims Header
@@ -243,7 +243,7 @@ def stg_revenue_center(
       service was received, and for outpatient claims the payment amount
       and allowed charge amount for individual services."
 
-    CCLF IP Section 3.5 "Part A Header vs Revenue Center Expenditures" (p.16):
+    :pincite:`VTWVB384 §3.5 p.16` — Part A Header vs Revenue Center Expenditures:
 
       "Both the Part A Header file (CCLF1) and the Part A Revenue Center
       file (CCLF2) contain a payment field (CLM_PMT_AMT and
@@ -282,7 +282,7 @@ def int_diagnosis_pivot(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.4 "Part A Diagnosis Code File" (p.9):
+    :pincite:`VTWVB384 §2.2.4 p.9` — Part A Diagnosis Code File:
 
       "The Part A Diagnosis Code File (CCLF4) contains the diagnosis code
       for the principal diagnosis as well as all secondary diagnoses that
@@ -336,7 +336,7 @@ def int_procedure_pivot(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.3 "Part A Procedure Code File" (p.9):
+    :pincite:`VTWVB384 §2.2.3 p.9` — Part A Procedure Code File:
 
       "The Part A Procedure Code File (CCLF3) contains detailed information
       regarding claims from the Part A Claims Header File, such as the type
@@ -402,7 +402,7 @@ def int_institutional_medical_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2 "Part A Claims Data" (p.8):
+    :pincite:`VTWVB384 §2.2 p.8` — Part A Claims Data:
 
       "Part A claim data files contain claims submitted by facilities such
       as hospitals, SNFs, HHAs, rehabilitation facilities, and dialysis
@@ -560,7 +560,7 @@ def stg_physician_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.6 "Part B Physician File" (p.10):
+    :pincite:`VTWVB384 §2.2.6 p.10` — Part B Physician File:
 
       "The Part B Physician File (CCLF5) consists of claim-line records but
       includes both claim-level and line-level information. At the claim
@@ -570,7 +570,7 @@ def stg_physician_claim(
       amount, allowed charge amount, line-level diagnosis code, units of
       service, primary payer, provider TIN, and rendering NPI number."
 
-    CCLF IP Section 5.3.1 (p.23):
+    :pincite:`VTWVB384 §5.3.1 p.23` — Section 5.3.1:
 
       "Identify all the canceled records (line items) in the Part B
       Physician file. The canceled line items are identified by
@@ -603,16 +603,16 @@ def int_physician_claim_adr(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 5.1.2 "Natural Key for Part B Physician/DME" (p.20):
+    :pincite:`VTWVB384 §5.1.2 p.20` — Natural Key for Part B Physician/DME:
 
       Part B natural key: CLM_CNTL_NUM, Most Recent MBI (MR_MBI).
 
-    CCLF IP Section 5.2.2 "Related Claims in the Part B Physician File" (p.22):
+    :pincite:`VTWVB384 §5.2.2 p.22` — Related Claims in the Part B Physician File:
 
       "In the Part B Physician File, you will find original claims,
       cancellation claims, and adjustment claims."
 
-    CCLF IP Section 3.2 "Dropping Denied Claims" (p.15):
+    :pincite:`VTWVB384 §3.2 p.15` — Dropping Denied Claims:
 
       "For Part B Physician/DME claims, some individual line-items can be
       denied, whereas other line-items are not denied. Part B claims need
@@ -652,7 +652,7 @@ def int_physician_medical_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.6 "Part B Physician File" (p.10):
+    :pincite:`VTWVB384 §2.2.6 p.10` — Part B Physician File:
 
       CCLF5 contains both claim-level and line-level data in a single
       file. Unlike Part A (which has separate header/line/dx/px files),
@@ -763,7 +763,7 @@ def stg_dme_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.7 "Part B DME File" (p.10):
+    :pincite:`VTWVB384 §2.2.7 p.10` — Part B DME File:
 
       "The Part B DME File (CCLF6) consists of claim-line records but
       includes both claim-level and line-level information. Claim-level
@@ -773,7 +773,7 @@ def stg_dme_claim(
       amount, ordering NPI number, and 'paid to' NPI number."
 
     Cancellation adjustment applied to CLM_LINE_CVRD_PD_AMT and
-    CLM_LINE_ALOWD_CHRG_AMT per IP Section 5.3.1.
+    CLM_LINE_ALOWD_CHRG_AMT per :pincite:`VTWVB384 §5.3.1` —
     """
     xref = cclf___stg_beneficiary_xref
     return cclf__cclf6.join(
@@ -799,7 +799,7 @@ def int_dme_claim_adr(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 5.2.3 "Related Claims in the Part B DME File" (p.23):
+    :pincite:`VTWVB384 §5.2.3 p.23` — Related Claims in the Part B DME File:
 
       "In the Part B DME File, you will find original claims, cancellation
       claims, and adjustment claims. A variety of related claims are found
@@ -840,7 +840,7 @@ def int_dme_medical_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2.7 "Part B DME File" (p.10):
+    :pincite:`VTWVB384 §2.2.7 p.10` — Part B DME File:
 
     Key field mappings from CCLF6 (IP Table 19):
       - CUR_CLM_UNIQ_ID -> claim_id
@@ -940,7 +940,7 @@ def medical_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.2 "Part A Claims Data" + Section 2.2.6-2.2.7 (p.8-10):
+    :pincite:`VTWVB384 §2.2 p.8-10` — Part A Claims Data:
 
       The three claim types are combined into a single medical_claim table:
 
@@ -954,7 +954,7 @@ def medical_claim(
       - Professional/DME (CCLF6): Durable Medical Equipment, Prosthetics,
         Orthotics, and Supplies. claim_type = 'professional'.
 
-    CCLF IP Section 5.4 "Part A vs Part B Claims" (p.27):
+    :pincite:`VTWVB384 §5.4 p.27` — Part A vs Part B Claims:
 
       "The Part A claims files will include Medicare provider payments for
       some services covered under both Part A and Part B. To distinguish,
@@ -984,7 +984,7 @@ def stg_pharmacy_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.3 "Part D Claims Data" (p.10):
+    :pincite:`VTWVB384 §2.3 p.10` — Part D Claims Data:
 
       "The Part D File (CCLF7) contains prescription drug information at
       the beneficiary level. Data elements include the NDC, quantity
@@ -1015,7 +1015,7 @@ def int_pharmacy_claim_adr(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 3.3 "Part D Data Limitations" (p.15):
+    :pincite:`VTWVB384 §3.3 p.15` — Part D Data Limitations:
 
       "The Part D claims contained in the CCLF are 'final action' claims,
       unlike the other claims-related files which are debit/credit claims.
@@ -1025,7 +1025,7 @@ def int_pharmacy_claim_adr(
       for any given set of Related Claims the most recent claim and
       delete/ignore all of the previous related claims for that event."
 
-    CCLF IP Section 5.2.4 "Related Claims in the Part D File" (p.23):
+    :pincite:`VTWVB384 §5.2.4 p.23` — Related Claims in the Part D File:
 
       "In the Part D File, you will find original claims, cancellation
       claims, and adjustment claims."
@@ -1057,7 +1057,7 @@ def pharmacy_claim(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.3 "Part D Claims Data" (p.10):
+    :pincite:`VTWVB384 §2.3 p.10` — Part D Claims Data:
 
     Key field mappings from CCLF7 (IP Table 20):
       - CUR_CLM_UNIQ_ID -> claim_id
@@ -1137,14 +1137,14 @@ def stg_beneficiary_demographics(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.4.1 "Beneficiary Demographics File" (p.10-12):
+    :pincite:`VTWVB384 §2.4.1 p.10-12` — Beneficiary Demographics File:
 
       "This file contains the beneficiary's current MBI, first/middle/last
       name, ZIP code, date of birth, sex, race, age, Medicare Status Code,
       dual eligibility status, hospice begin/end dates, and date of death
       if a decedent."
 
-    CCLF IP Section 3.1 (p.14):
+    :pincite:`VTWVB384 §3.1 p.14` — Section 3.1:
 
       MBI resolution via CCLF9 applied before deduplication to ensure a
       single row per resolved beneficiary identity.
@@ -1181,7 +1181,7 @@ def eligibility(
 
     CCLF IP Reference
     ~~~~~~~~~~~~~~~~~~
-    CCLF IP Section 2.4.1 "Beneficiary Demographics File" (p.10):
+    :pincite:`VTWVB384 §2.4.1 p.10` — Beneficiary Demographics File:
 
     Key field mappings from CCLF8 (IP Table 21):
       - BENE_MBI_ID (resolved) -> person_id, member_id

src/bib/pincite.py

@@ -48,7 +48,7 @@ if TYPE_CHECKING:
 # Zotero key charset: 23456789ABCDEFGHIJKLMNPQRSTUVWXYZ (no 0, 1, O, L)
 _PINCITE_RE = re.compile(
     r":pincite:`"
-    r"([23456789A-HJ-NP-Z]{8})"  # group 1: item key
+    r"([23456789A-HJ-NP-Z]{8})"  # group 1: item key (strict Zotero charset)
     r"(?:\s+(.+?))?"  # group 2: locator (optional)
     r"`"
     r"(?:\s*(?:--|—)\s*(.+))?"  # group 3: note (optional)

tests/bib/test_meta.py

@@ -113,12 +113,15 @@ class TestExtractMetaTags:
         assert extract_meta_tags(fn) == []
 
     def test_real_cclf_function(self):
+        """CCLF functions now use :pincite: directives instead of bare CCLF IP Section refs."""
         from aco.express.cclf import stg_beneficiary_xref
+        from bib.pincite import parse_pincites
 
-        tags = extract_meta_tags(stg_beneficiary_xref)
+        doc = stg_beneficiary_xref.__doc__ or ""
-        labels = {t.label for t in tags}
+        pincites = parse_pincites(doc, "aco.express.cclf.stg_beneficiary_xref")
-        assert "meta:cclf-ip-s3.1" in labels
+        assert len(pincites) >= 2
-        assert "meta:cclf-ip-s5.1.1" in labels
+        keys = {p.item_key for p in pincites}
+        assert "VTWVB384" in keys
 
 
 # ── extract_meta_details ──────────────────────────────────────────
@@ -269,6 +272,7 @@ class TestTagItemsFromPipeline:
 
         store = Store(":memory:")
         item = Source(
+            key="VTWVB384",
             title="CCLF Information Packet",
             tags=["source:cms-cclf-ip", "module:aco"],
         )
@@ -282,12 +286,10 @@ class TestTagItemsFromPipeline:
         tagged = {k: v for k, v in report.items() if v}
         assert len(tagged) > 0
 
-        # Verify the item now has fn: and meta: tags
+        # Verify the item now has fn: tags (meta: tags replaced by :pincite: directives)
         updated = store.get(key)
         fn_tags = [t for t in updated.tags if t.startswith("fn:")]
-        meta_tags = [t for t in updated.tags if t.startswith("meta:")]
         assert len(fn_tags) > 0
-        assert len(meta_tags) > 0
 
 
 # ── generate_docstring_refs ───────────────────────────────────────