pubid 2.0.1 → 2.0.3

data/lib/pubid/ieee/parser.rb

@@ -1,1261 +1,11 @@
 # frozen_string_literal: true
 
-require "parslet"
-
 module Pubid
   module Ieee
     # Parser class for IEEE identifiers
     # Single Responsibility: Parsing IEEE identifier syntax
     # Note: IEEE is extremely complex with many edge cases
-    class Parser < ::Pubid::Parser::Grammar
-      # Basic building blocks
-      rule(:space) { str(" ") }
-      rule(:space?) { space.maybe }
-      rule(:dash) { str("-") }
-      rule(:dash?) { dash.maybe }
-      rule(:dot) { str(".") }
-      rule(:slash) { str("/") }
-      rule(:comma) { str(", ") }
-      rule(:digit) { match("[0-9]") }
-      rule(:digits) { digit.repeat(1) }
-      rule(:letter) { match("[A-Za-z]") }
-      rule(:upper) { match("[A-Z]") }
-      rule(:lower) { match("[a-z]") }
-
-      # Year pattern (4 digits starting with 19 or 20), optionally followed by letter(s)
-      # e.g. 2012, 201x, 2010a
-      rule(:year_digits) do
-        (str("19") | str("20")) >> digit.repeat(2,
-                                                2) >> lower.repeat(0,
-                                                                   2) >> digits.absent?
-      end
-
-      # Month patterns - numeric format (01-12)
-      rule(:month_numeric) do
-        (str("0") >> match("[1-9]")) | # 01-09
-          (str("1") >> match("[0-2]")) # 10-12
-      end
-
-      # Comprehensive date parsing
-      # Format 1: "September 2018" or "Sept 2018" (text month + year)
-      rule(:date_with_month_text) do
-        month_name.as(:month) >> space >> year_digits.as(:year)
-      end
-
-      # Format 2: "2018-09" (year-numeric month)
-      rule(:date_with_month_numeric) do
-        year_digits.as(:year) >> dash >> month_numeric.as(:month)
-      end
-
-      # Format 3: Just year "2018"
-      rule(:date_year_only) do
-        year_digits.as(:year)
-      end
-
-      # Combined date rule - longest match first
-      rule(:date_standalone) do
-        date_with_month_text | date_with_month_numeric | date_year_only
-      end
-
-      # Trailing print/reaffirm date "<sep>Month YYYY" — captured under the
-      # distinct :trailing_month/:trailing_year keys so it never collides with a
-      # base -YYYY identity year (the builder promotes it only when no base year
-      # exists). Accepts BOTH a month name ("May 2014") and a month-first numeric
-      # month ("05 2014"). The two variants are unambiguous: month_name never
-      # starts with a digit, and month_numeric (01-12) never matches a bare
-      # 19xx/20xx year — so a bare trailing year still falls through to callers'
-      # own bare-year clause.
-      rule(:trailing_month_year) do
-        # The numeric branch accepts `-` as well as space before the year:
-        # preprocessing (the `(\d)\s+(\d{4})` gsub) rewrites a trailing
-        # "05 2014" → "05-2014", so a numeric month reaches the grammar
-        # dash-joined; the month-name branch escapes that gsub (the char before
-        # the space is a letter) and stays space-joined.
-        ((comma | space) >> month_name.as(:trailing_month) >> space >> year_digits.as(:trailing_year)) |
-          ((comma | space) >> month_numeric.as(:trailing_month) >> (space | dash) >> year_digits.as(:trailing_year))
-      end
-
-      # Month patterns
-      rule(:month_name) do
-        # Period-suffixed abbreviations (longest first)
-        str("Sept.") | str("Oct.") | str("Nov.") | str("Dec.") |
-          str("Jan.") | str("Feb.") | str("Mar.") | str("Apr.") |
-          str("Jun.") | str("Jul.") | str("Aug.") |
-          # Full month names
-          str("January") | str("February") | str("March") | str("April") |
-          str("May") | str("June") | str("July") | str("August") |
-          str("September") | str("October") | str("November") | str("December") |
-          # Non-period abbreviations
-          str("Jan") | str("Feb") | str("Mar") | str("Apr") | str("Jun") |
-          str("Jul") | str("Aug") | str("Sep") | str("Sept") | str("Oct") | str("Nov") | str("Dec")
-      end
-
-      # Organizations
-      rule(:organization) do
-        str("IEEE") | str("AIEE") | str("ANSI") | str("ASA") |
-          # ANS (American Nuclear Society) — a third co-publisher on nuclear
-          # standards ("ANSI/IEEE/ANS 7.4-3-2-1982"). Listed AFTER ANSI so it
-          # never shadows the longer token.
-          str("ANS") |
-          str("IEC") | str("ISO") | str("ASTM") | str("CSA") | str("ASME") |
-          str("NACE") | str("NSF") | str("ASHRAE") | str("NCTA") | str("AESC") |
-          str("EIA") | # NEW Session 224: Add EIA support
-          # Historical / foreign co-publishers seen in relaton-data-ieee
-          # (bucket 4): AMPP, USAS (both standalone), and the IEEE sub-board /
-          # partner co-publishers USEMCSC, EAB, MPAI.
-          str("USEMCSC") | str("AMPP") | str("USAS") | str("EAB") | str("MPAI")
-      end
-
-      # Complex organization prefixes (Category 5: ANSI Complex)
-      rule(:complex_org_prefix) do
-        str("ANSI/IEEE-ANS") | str("ANSI/IEEE") | str("ANSI")
-      end
-
-      # Characteristic IEEE number patterns (without prefix)
-      # These patterns are distinctly IEEE even without "IEEE Std" prefix
-      rule(:characteristic_ieee_number) do
-        # C37.xxx series (power systems) - C followed by 2 digits, dot, more digits
-        (str("C") >> digit.repeat(2,
-                                  2) >> dot >> digits >> match("[a-z]").repeat.maybe) |
-          # 802.xxx series (networking) - 802 followed by dot, digits, optional letter suffix
-          (str("802") >> dot >> digits >> match("[a-z]").repeat.maybe) |
-          # P followed by digits (draft projects)
-          (str("P") >> digits.repeat(1))
-      end
-
-      rule(:publisher) do
-        complex_org_prefix.as(:publisher) | organization.as(:publisher)
-      end
-
-      rule(:copublisher) do
-        # Three-way copublisher strings (treat as single unit, longest first)
-        str("/ISO/IEC").as(:copublisher) |
-          str("/IEC/ISO").as(:copublisher) |
-          # Two-way copublishers (original pattern)
-          (slash >> space? >> organization.as(:copublisher))
-      end
-
-      # Conformance document patterns (/Conformance01-2003, /Conformance02-2014)
-      # Allow optional space before slash for malformed inputs
-      rule(:conformance) do
-        (space? >> slash >> str("Conformance") >> match("[0-9]").repeat(1).as(:conf_number) >> dash >> year_digits.as(:conf_year)).as(:conformance)
-      end
-
-      # ASHRAE joint publication patterns (/ASHRAE Guideline 21-2012)
-      # Also handles /ASHRAE 21 without "Guideline"
-      rule(:ashrae_copub) do
-        (slash >> str("ASHRAE") >> space >>
-         (str("Guideline") >> space).maybe >>
-         digits.as(:ashrae_number) >>
-         (dash >> year_digits.as(:ashrae_year)).maybe).as(:ashrae_copub)
-      end
-
-      # IEEE cross-reference patterns (/C62.22.1-1996)
-      # References another IEEE standard from a specific series (e.g., C62, C37, C57)
-      rule(:ieee_crossref) do
-        (slash >> str("C") >> digits >> dot >> digits >> dot >> digits >> dash >> year_digits).as(:ieee_crossref)
-      end
-
-      # IPCEA co-designation suffix (/IPCEA P-46-426-1962). Captured verbatim
-      # (leading slash included) so it round-trips through the `crossref`
-      # attribute. Used only by the S-designation rule below.
-      rule(:ipcea_copub) do
-        (slash >> str("IPCEA") >> space >>
-         match('[A-Za-z0-9.\-]').repeat(1)).as(:ipcea_copub)
-      end
-
-      # Historical IEEE/IPCEA co-published cable designation (e.g. S-135).
-      # The "S-<digits>" number has a dash between the letter series and the
-      # digits, which the shared `number` rule deliberately rejects (that
-      # tightening is what keeps a bare "IEEE S" from parsing). So this one-off
-      # family gets its own rule rather than loosening `number` and risking the
-      # 12k-row corpus. Requiring `str("S") >> dash >> digits` keeps "IEEE S"
-      # (no dash+digit) rejected. Handles: "IEEE Std S-135", "IEEE S-135",
-      # bare "S-135", the "/IPCEA …" slash co-designation, and the
-      # "(IPCEA …)" parenthetical variant.
-      rule(:s_designation) do
-        (publisher >> space).maybe >>
-          (type_word.as(:type) >> space?).maybe >>
-          (str("S") >> dash >> digits).as(:s_number) >>
-          ipcea_copub.maybe >>
-          parenthetical.maybe
-      end
-
-      # Document number - support letters and digits, with optional prefix P
-      # Complex multi-part numbers like P11073-10404-10419 should be fully captured
-      # But simple cases like "623-1976" should not consume the dash before year
-      rule(:number) do
-        (str("P").maybe >>
-         # The numeric core must contain at least one digit: an optional letter
-         # prefix (C, S, …) then a required `digits` run, then any digit/letter
-         # tail. The old `(digits | upper).repeat(1)` wrongly accepted an
-         # all-letter token like a bare "S" (a prefix with no number — e.g.
-         # "IEEE S"), which is not a valid IEEE identifier.
-         upper.repeat(0) >> digits >> (digits | upper).repeat(0) >>
-         # Only consume dash+digits if followed by another dash+digits (multi-part pattern)
-         # OR if the digits don't look like a year (not 4 digits starting with 19/20)
-         # This prevents consuming "623-1976" as a number but allows "P11073-10404-10419"
-         (dash >> digits >> year_digits.absent? >> (dash >> digits).repeat).maybe >>
-         lower.maybe).as(:number)
-      end
-
-      # Type - handle "No." and "No" (case-insensitive), longest first
-      rule(:type_word) do
-        str("Draft Std") | str("STD") | str("Standard") |
-          str("Std No.") | str("Std") | # Add "Std No." before "Std"
-          str("PTC") | # ASME Performance Test Code
-          (match("[Nn]") >> str("o.")) | (match("[Nn]") >> str("o")) |
-          str("No")
-      end
-
-      # Part and subpart - handle both dot and dash separators
-      rule(:part) do
-        (dot | dash) >> match("[0-9A-Za-z]").repeat(1).as(:part)
-      end
-
-      rule(:subpart) do
-        (dot | dash | str("_")) >>
-          ((str("REV") | str("Rev")).maybe >> match("[0-9a-z]").repeat(1) >>
-           (dot >> digits).maybe).as(:subpart)
-      end
-
-      # Year component - updated to use comprehensive date parsing
-      rule(:year) do
-        (dot | dash) >> date_standalone >> str("(E)").maybe
-      end
-
-      # Draft patterns
-      rule(:draft_status) do
-        # "Active" (bucket 3) joins the generic draft-status path so
-        # "IEEE Active Std P… /D…" parses like the "Unapproved" forms, without
-        # touching ieee_approved_draft_identifier (which would break the
-        # issue-#209 unapproved-drops-Std rendering). "Active Approved" completes
-        # the two-word matrix alongside "Active Unapproved" (draft-grammar
-        # coverage: the "<status> Draft P…/D…" prefix forms). Longest token first.
-        (str("Active Unapproved") | str("Active Approved") | str("Unapproved") |
-         str("Approved") | str("Active")) >> space
-      end
-
-      rule(:draft_prefix) do
-        space? >> (str("/") | str("_") | dash | space)
-      end
-
-      rule(:draft_version) do
-        # Enhanced to handle multiple draft notation patterns
-        # D is optional to handle /08 style drafts (e.g., IEEE P1052/08)
-        # A draft never begins with "R-" — that is the revision suffix
-        # (revision_suffix rule); guard so the D-less path doesn't swallow a
-        # bare "/R-<id>" (e.g. the no-draft "P1722/R-1") as a draft.
-        (str("R") >> dash).absent? >>
-          (str("D") >> str("IS").absent?).maybe >> # Avoid matching "DIS" (ISO stage)
-          (
-            # Pattern: D3.1 (decimal with 1-2 digits on each side) - MOST COMMON, put first
-            # Also handles trailing letter: D7.3A, D2.0E
-            (match("[0-9]").repeat(1,
-                                   2) >> dot >> match("[0-9]").repeat(1,
-                                                                      2) >> lower.maybe) |
-            # Pattern: D.XX (decimal starting with dot) - e.g., D.19
-            (dot >> digits) |
-            # Pattern: DX+X (plus sign) - e.g., D1+1
-            (digits >> str("+") >> digits) |
-            # Pattern: DXXXXeYY or DXXXX.eYY (complex) - e.g., D2012.e27
-            (digits >> dot.maybe >> str("e") >> digits) |
-            # Pattern: D-X or DX or DX-d or DX_letter (original patterns)
-            # Handles: D12, D3.0, D043Rev18, suffixes like D15Sept
-            (str("-").maybe >> match("[0-9A-Za-z]").repeat(1) >> (str("-d") | (str("_") >> match("[0-9A-Za-z]").repeat(0))).maybe)
-          ).as(:draft_version)
-      end
-
-      rule(:draft_date) do
-        # Enhanced to handle: ", Sept 2008" or " Sept 2008" or ", Month Year"
-        ((comma | space) >> month_name.as(:month) >> space >> year_digits.as(:year)) |
-          # Space-separated bare year after the draft designator
-          # ("IEEE Draft Std P14764/D1 2004, Nov 2004" - the "D1 2004" draft
-          # carries its own year, with the print date still trailing).
-          (space >> year_digits.as(:year)) |
-          # Numeric-month form ", 05 2007" / " 05 2007" — no text month name. The
-          # year separator may be a dash: preprocessing (parser.rb ~1278) rewrites
-          # a trailing " <digits> <year>" to "<digits>-<year>", so "05 2007"
-          # reaches the grammar as "05-2007".
-          ((comma | space) >> month_numeric.as(:month) >> (space | dash) >> year_digits.as(:year)) |
-          (((space? >> comma >> space?) | space) >> month_name.as(:month) >>
-          (
-            # "Month DD, Year" (day then comma) and "Month DD Year" (day, no
-            # comma) — the day/year separator may be a dash for the same
-            # preprocessing reason ("July 15 2012" → "July 15-2012").
-            ((space >> digits.as(:day)) >> ((comma | space | dash) >> year_digits.as(:year))) |
-            ((space >> digits.as(:day)).maybe >> comma >> year_digits.as(:year)) |
-            (comma >> space? >> year_digits.as(:year)) |
-            (space >> year_digits.as(:year))
-          ))
-      end
-
-      # FDIS and similar ISO stage codes without D prefix (Pattern 3)
-      # These appear after / in IEEE P identifiers but don't have the D prefix
-      # Examples: IEEE P15939/FDIS, IEEE P1234/CDV
-      rule(:fdraft) do
-        (slash >>
-         (str("FDIS") | str("CDV") | str("CD") | str("WD") | str("PWI") | str("NP")) >>
-         (
-           ((comma | space) >> month_name.as(:month) >> space >> year_digits.as(:year)) | # Month Year
-           ((comma | space) >> year_digits.as(:year)) # Year only (e.g., /FDIS, 2016)
-         ).maybe >>
-         parenthetical.maybe).as(:fdraft)
-      end
-
-      rule(:draft) do
-        (draft_prefix >> draft_version.repeat(1, 2) >>
-         (dot >> digits.as(:revision)).maybe >>
-         # The compound both-systems form (docs/IEEE-DRAFT-STAGES.md
-         # §1.3): "=DDIS.3" - IEEE draft ordinal = draft of the ISO/IEC
-         # stage, with its iteration. The glued "=DDIS3" / "=DDIS-3"
-         # spellings are accepted aliases of "=DDIS.3".
-         (str("=") >> (str("D") >> (str("FDIS") | str("CDV") | str("PWI") |
-            str("DIS") | str("WD") | str("NP") | str("CD")) |
-           (str("FDIS") | str("CDV") | str("PWI") | str("DIS") | str("WD") |
-            str("NP") | str("CD"))).as(:draft_iso_stage) >>
-           (dot | dash).maybe >> digits.as(:draft_iso_iteration).maybe).maybe >>
-         draft_date.maybe).as(:draft)
-      end
-
-      # Trailing revision suffix "/R-<id>". Normalization funnels every revision
-      # spelling here — IEEE's native inline "Rev<n>" (repositioned) and
-      # relaton's synthetic "/R-<x>" (kept in place) — so a single trailing rule
-      # captures them all. The id is alphanumeric ("2", "18", "i").
-      rule(:revision_suffix) do
-        slash >> str("R") >> dash >> match('[0-9A-Za-z]').repeat(1).as(:revision)
-      end
-
-      # Edition - enhanced to support IEC formats like "Edition 1.0 2015-03"
-      #
-      # The edition's own year is captured as :edition_year (NOT :year). The
-      # generic Std branch carries a base "-YYYY" slot AND a separate trailing
-      # `edition.maybe`; a bare :year here would collide with the base :year
-      # when both fire, so Parslet warns "Duplicate subtrees … keys: [:year]"
-      # and drops the base identity year (the same failure #299 fixed for the
-      # trailing month/year clause). The builder promotes :edition_year to the
-      # identity year only when there is no base year.
-      rule(:edition) do
-        (comma >> year_digits.as(:edition_year) >> str(" Edition")) |
-          ((space | dash) >> str("Edition ") >>
-           (digits >> dot >> digits).as(:edition) >>
-           # Year separator: a space, " - ", or a bare dash (preprocessing
-           # rewrites "Edition 3.0 2016" -> "Edition 3.0-2016", the shape the
-           # normalized "/E-<n>-YYYY" suffix produces — nil-residue item 1).
-           (str(" - ") | space | dash) >>
-           year_digits.as(:edition_year) >>
-           (dash >> digit.repeat(2, 2).as(:edition_month)).maybe) # Capture -MM as edition_month
-      end
-
-      # Part/subpart/year combinations
-      rule(:part_subpart_year) do
-        (part >> subpart.repeat(1, 2) >> year) |
-          (part >> subpart >> year) |
-          (part >> year) |
-          (part >> subpart) |
-          year |
-          part
-      end
-
-      # Corrigendum
-      rule(:corrigendum) do
-        # Enhanced: Accept space as separator, make separators more flexible
-        # Also accept "Corrigendum" as alternative to "Cor"
-        ((str("_") | slash | dash | space) >>
-         (str("Corrigendum") | str("Cor")) >>
-         (dash | dot | space).maybe >> # More flexible separator after "Cor"
-         space? >> # Add space handling after separator
-         digits.as(:cor_number).maybe >>
-         ((dash | str(":") | space) >> year_digits.as(:cor_year)).maybe).as(:corrigendum)
-      end
-
-      # Amendment — mirrors the corrigendum rule's separator flexibility so
-      # IEEE-format "/Amd 2-2004" and "/Amd2-2004" parse the same way as
-      # their Cor counterparts (issue #210).
-      rule(:amendment) do
-        ((str("_") | slash | dash | space) >>
-         (str("Amendment") | str("Amd")) >>
-         (dash | dot | space).maybe >>
-         space? >>
-         digits.as(:amd_number).maybe >>
-         ((dash | str(":") | space) >> year_digits.as(:amd_year)).maybe).as(:amendment)
-      end
-
-      # Interpretation notation (/INT)
-      # Enhanced to support optional year suffix: /INT-1991, /INT 1991
-      rule(:interpretation) do
-        (slash >> str("INT") >> ((dash | str(":") | space) >> year_digits.as(:int_year)).maybe).as(:interpretation)
-      end
-
-      # Reaffirmed - enhanced to support (R1992) format without space
-      rule(:reaffirmed) do
-        (
-          # Format: "Reaffirmed 1992"
-          (str("Reaffirmed ") >> year_digits.as(:year)) |
-          # Format: "(R1992)" - parentheses with R prefix (with or without space before)
-          (space.maybe >> str("(R") >> year_digits.as(:year) >> str(")"))
-        ).as(:reaffirmed)
-      end
-
-      # Redline suffix at the very end. Accepts relaton's canonical " Redline"
-      # (space, no dash) and pubid's older " - Redline" (space-dash-space),
-      # case-insensitive. Captured (presence only) so the builder sets a
-      # `redline: true` flag the renderer restores — a redline is a distinct
-      # document and must not collapse to its base standard.
-      rule(:redline) do
-        (space >> (dash >> space).maybe >>
-         (str("Redline") | str("REDLINE") | str("redline"))).as(:redline)
-      end
-
-      # Book nickname (e.g., "[The Orange Book]", "[IEEE Gold Book]")
-      rule(:book_nickname) do
-        space >> str("[") >> match("[^\\]]").repeat(1).as(:nickname) >> str("]")
-      end
-
-      # Relationship type keywords for Pattern 4 identifiers
-      rule(:relationship_revision_of) do
-        str("Revision of ") | str("Revison of ")
-      end
-      rule(:relationship_amendment_to) { str("Amendment to ") }
-      rule(:relationship_corrigendum_to) do
-        str("Corrigendum to ") | str("Corrigenda to ")
-      end
-      rule(:relationship_incorporates) do
-        str("incorporates ") | str("Incorporating ") | str("Incorporates ")
-      end
-      rule(:relationship_adoption_of) { str("Adoption of ") }
-      rule(:relationship_supplement_to) { str("Supplement to ") }
-      rule(:relationship_draft_amendment) do
-        str("Draft Amendment to ") | str("DRAFT Amendment to ")
-      end
-      rule(:relationship_draft_revision) { str("Draft Revision of ") }
-      rule(:relationship_reaffirmation) { str("Reaffirmation of ") }
-      rule(:relationship_redesignation) do
-        str("Redesignation of ") | str("redesignated as ")
-      end
-      rule(:relationship_supersedes) { str("Supersedes ") | str("Supercedes ") }
-      rule(:relationship_previously_designated) do
-        str("Previously designated as ")
-      end
-      rule(:relationship_includes) { str("Includes ") } # NEW Session 171
-
-      # Combined relationship type (longest match first)
-      rule(:relationship_type) do
-        relationship_draft_amendment.as(:draft_amendment_to) |
-          relationship_draft_revision.as(:draft_revision_of) |
-          relationship_previously_designated.as(:previously_designated_as) |
-          relationship_reaffirmation.as(:reaffirmation_of) |
-          relationship_redesignation.as(:redesignation_of) |
-          relationship_supersedes.as(:supersedes) |
-          relationship_includes.as(:includes) | # NEW Session 171
-          relationship_revision_of.as(:revision_of) |
-          relationship_amendment_to.as(:amendment_to) |
-          relationship_corrigendum_to.as(:corrigendum_to) |
-          relationship_incorporates.as(:incorporates) |
-          relationship_adoption_of.as(:adoption_of) |
-          relationship_supplement_to.as(:supplement_to)
-      end
-
-      # Identifier string (for parsing list of related identifiers)
-      # Captures text until delimiter: comma, closing paren, "and", " / ", "; ", "as amended by"
-      # Uses absent? to ensure we stop at these delimiters
-      rule(:identifier_string) do
-        (
-          str(", and ").absent? >>
-          str(" and ").absent? >>
-          str(", ").absent? >>
-          str(" as amended by ").absent? >>
-          # Stop at any "/", ";", or "-" that introduces another relationship
-          # (look-ahead: separator + relationship_type keyword). This lets the
-          # slash inside "IEEE Std 525-2007/Cor 1-2015" stay part of the
-          # identifier, while the slash before "Incorporates ..." splits.
-          relationship_break.absent? >>
-          str(")").absent? >>
-          match(".")
-        ).repeat(1)
-      end
-
-      # Identifier list (comma and "and" separated)
-      rule(:identifier_list) do
-        identifier_string.as(:id) >>
-          (
-            (str(", and ") | str(" and ") | str(", ")) >>
-            identifier_string.as(:id)
-          ).repeat
-      end
-
-      # "as amended by" clause with identifier list
-      rule(:as_amended_by_clause) do
-        # Variant 1: "as amended by IEEE's X, Y, Z"
-        (str(" as amended by IEEE's ") >> identifier_list.as(:amendments)) |
-          # Variant 2: "as amended by X, Y, Z" (standard)
-          (str(" as amended by ") >> identifier_list.as(:amendments)) |
-          # Variant 3: "and its approved amendments" (no specific list)
-          str(" and its approved amendments").as(:approved_amendments)
-      end
-
-      # A character sequence that may separate two relationships inside the
-      # parenthetical: a "/", ";", or "-" with optional surrounding spaces.
-      # Standalone it is permissive — the actual decision to split is gated by
-      # `relationship_break`, which requires another relationship_type to
-      # follow. This way the slash in "IEEE Std 525-2007/Cor 1-2015" stays
-      # part of the related identifier.
-      rule(:relationship_separator) do
-        (space.maybe >> str("/") >> space.maybe) |
-          (space.maybe >> str(";") >> space.maybe) |
-          (space.maybe >> str("-") >> space.maybe)
-      end
-
-      # Look-ahead: a separator followed by another relationship_type keyword.
-      # Used as an absent? guard in identifier_string so the parser stops
-      # consuming characters right before a new relationship begins.
-      rule(:relationship_break) do
-        relationship_separator >> space.maybe >> relationship_type
-      end
-
-      # Relationship clause (handles all relationship types)
-      rule(:relationship_clause) do
-        space.maybe >> str("(") >>
-          relationship_type.as(:relationship_type) >>
-          identifier_list.as(:related_ids) >>
-          as_amended_by_clause.maybe >>
-          # Additional relationships separated by "/", ";", or "-" (optionally
-          # surrounded by spaces). The separator is only honored when followed
-          # by another relationship_type — identifier-internal slashes like
-          # "/Cor 1-2015" are not mistaken for relationship breaks because
-          # identifier_string stops at relationship_break ahead.
-          (
-            relationship_separator >>
-            relationship_type.as(:relationship_type) >>
-            identifier_list.as(:related_ids) >>
-            as_amended_by_clause.maybe
-          ).repeat.as(:additional_rels) >>
-          str(")")
-      end
-
-      # Title portion separated by colon (Category 8)
-      rule(:title_portion) do
-        str(":") >> space >> match('[^\n]').repeat(1).as(:title)
-      end
-
-      # Approved Draft suffix (Category 7)
-      rule(:approved_draft_suffix) do
-        (space >> str("- (Approved Draft)")) | (space >> str("(Approved Draft)"))
-      end
-
-      # Additional parameters (inside parentheses)
-      rule(:additional_parameters) do
-        (space.maybe >> str("(") >> space.maybe >> # Space before/after '(' optional
-         # ("( Revision of …)" - a crawl space inside the paren must not push
-         # the narrative to the catch-all and into the render).
-         (reaffirmed |
-          # Handle "Revision of IEEE Std ..." with optional space after Std
-          (str("Revision of IEEE Std ") >> space.maybe >> match("[^)]").repeat(1).as(:revision_of)) |
-          # Handle typo "Revison of IEEE Std ..." with optional space after Std
-          (str("Revison of IEEE Std ") >> space.maybe >> match("[^)]").repeat(1).as(:revision_of)) |
-          # Handle "Revision to IEEE Std ..." with optional space after Std
-          (str("Revision to IEEE Std ") >> space.maybe >> match("[^)]").repeat(1).as(:revision_of)) |
-          # Handle "Revison to IEEE Std ..." with optional space after Std
-          (str("Revison to IEEE Std ") >> space.maybe >> match("[^)]").repeat(1).as(:revision_of)) |
-          # Amendment patterns (case-insensitive DRAFT)
-          ((str("DRAFT") | str("Draft") | str("draft")) >> str(" Amendment to ") >> match("[^)]").repeat(1).as(:draft_amendment_to)) |
-          (str("Amendment to IEEE Std ") >> space.maybe >> match("[^)]").repeat(1).as(:amendment_to)) |
-          # Adoption patterns
-          (str("Adoption of ") >> match("[^)]").repeat(1).as(:adoption)) |
-          # Other specific patterns
-          (str("Notebooks") >> space? >> match("[^,\\)]").repeat(1).as(:notebooks)) |
-          (str("Standard Newspaper(s)") >> space? >> match("[^,\\)]").repeat(1).as(:standard_newspapers)) |
-          # Catch-all for any other parenthetical content (MUST BE LAST)
-          match("[^)]").repeat(1).as(:parenthetical_content)
-         ) >>
-         str(")").maybe).as(:parameters)
-      end
-
-      # Parenthetical - try relationship_clause first, then fall back to additional_parameters
-      rule(:parenthetical) do
-        relationship_clause | additional_parameters
-      end
-
-      # IEC/IEEE copublished pattern - handle all variations comprehensively
-      # BUT exclude P prefix patterns (those are joint development)
-      rule(:iec_ieee_copublished) do
-        str("IEC/IEEE") >>
-          space >>
-          str("P").absent? >> # NOT a P prefix (would be joint development)
-          # The copublished number must contain at least one digit. This rejects
-          # an all-letter placeholder like "IEC/IEEE TR" (no real document
-          # number) — the same "require a digit" tightening the Standard number
-          # rule got, applied to the copublished number grammar.
-          (match("[^0-9\n]").repeat >> digit).present? >>
-          match("[^\n]").repeat(1).as(:content)
-      end
-
-      # Joint development patterns (ISO/IEC/IEEE in either IEEE or ISO format)
-      rule(:joint_development_ieee_format) do
-        # ISO/IEC/IEEE P26511/D8-2018 or ISO/IEEE P1003.1-2008 or IEC/IEEE P62582-1-2011
-        # ALSO handle: IEC/IEEE P60780-323, CDV1 2014 (comma before stage code)
-        # ALSO handle: IEEE/CSA P844.1/293.1/D2 (CSA dual numbering)
-        # ALSO handle: IEEE/ISO/IEC P42010/D=CD-2020 — the printed publisher
-        # order is the organization's perspective (pubid#469) and parses as
-        # printed; it is no longer rewritten to ISO-first.
-        (str("ISO/IEC/IEEE") | str("IEEE/ISO/IEC") | str("ISO/IEEE") | str("IEC/IEEE") | str("IEEE/CSA")).as(:joint_publishers) >>
-          space >>
-          # P = project (the document is a draft): identity-bearing, so it
-          # is captured and preserved, never silently consumed.
-          str("P").as(:project_marker) >>
-          digits.as(:number) >>
-          ((dot | dash) >> digits.as(:part)).maybe >> # Optional part like .1 or -1
-          # CSA dual numbering: /293.1 (second number)
-          (slash >> digits >> (dot >> digits).maybe >> (dash >> digits.as(:draft_version)).maybe).maybe >>
-          (
-            # Variant 1b: the ordinal-less stage draft "D=CDV-2020" -
-            # D (draft) = CDV (the IEC stage it drafts). The year rides in
-            # the draft clause (a distinct key, so the builder keeps the
-            # date inside the designator). The year follows IEEE format:
-            # dash-joined ("D=CD-2020", the ruled canonical); the colon
-            # spelling ("D=CDV:2020") stays accepted as the alias it was
-            # frozen with. A dash-year is guarded to 19xx/20xx so a
-            # YYMM monthcode tail can never be read as a year.
-            (slash >> str("D") >> str("=") >>
-              (str("CDV") | str("FDIS") | str("PWI") | str("WD") |
-               str("NP") | str("DIS") | str("CD")).as(:draft_iso_stage) >>
-              ((str(":") | str("-")) >>
-               ((str("19") | str("20")) >> digit >> digit).as(:draft_stage_year)).maybe) |
-            # Variant 1: /D8 notation (original), with the compound
-            # both-systems suffix "=DDIS.3" (docs/IEEE-DRAFT-STAGES.md §1.3)
-            (slash >> str("D") >> digits.as(:draft_version) >>
-              (str("=") >> (str("D") >> (str("FDIS") | str("CDV") | str("PWI") |
-                 str("DIS") | str("WD") | str("NP") | str("CD")) |
-               (str("FDIS") | str("CDV") | str("PWI") | str("DIS") | str("WD") |
-                str("NP") | str("CD"))).as(:draft_iso_stage) >>
-               (dot | dash).maybe >> digits.as(:draft_iso_iteration).maybe).maybe) |
-            # Variant 2: , CDV1 notation (comma before stage code) —
-            # the stage draft of the named ISO/IEC stage, its iteration
-            (comma >> (str("CDV") | str("FDIS") | str("CD") | str("DIS")).as(:iec_stage) >> digits.maybe.as(:stage_iteration))
-          ).maybe >>
-          # Optional edition, from relaton's "/E-<n>" suffix normalized to
-          # "Edition <n>.0[ YYYY]" (nil-residue hand-off item 1). The edition
-          # rule carries its own year, so the year clause below simply doesn't
-          # fire when an edition is present.
-          edition.maybe >>
-          ((dash >> year_digits.as(:year)) | # Either -YEAR
-           (comma.maybe >> space >> month_name.as(:month) >> space.maybe >> year_digits.as(:year))).maybe >> # Or Month YEAR (with optional comma)
-          revision_suffix.maybe
-      end
-
-      rule(:joint_development_iso_format) do
-        # ISO-led stage designations. Two spellings:
-        #   colon form  : "ISO/IEC/IEEE FDIS 26511:2018"           (already used)
-        #   corpus form : "ISO/IEC/IEEE FDIS P26515-2018-05"       (historical)
-        # The corpus form adds a leading "P" on the number, a trailing
-        # "-YYYY[-MM]" date (instead of ":YYYY"), multi-digit committee-draft
-        # stage codes (CD1..CD4) plus CDV, and a wider set of joint publishers.
-        # (roadmap items 2/3, phase 1). longest publisher token first.
-        # The stage is OPTIONAL ONLY for ISO-led joint prefixes - the
-        # stage-less PUBLISHED form (pubid#317): "ISO/IEC/IEEE 26511:2018",
-        # "ISO/IEEE 11073-20101:2004(E)". Every other prefix keeps the
-        # stage REQUIREMENT so this rule cannot steal what belongs to
-        # iec_ieee_copublished ("IEC/IEEE 60076-2016"), the bare-IEEE
-        # project rules ("IEEE P802.16/D-3") or the ISO flavor
-        # ("ISO/IEC <n>:<year>").
-        iso_led = (str("ISO/IEC/IEEE") | str("IEEE/ISO/IEC") |
-                   str("IEEE/IEC/ISO") | str("ISO/IEEE"))
-        staged_only = (str("IEC/IEEE") | str("IEEE/IEC") |
-                       str("ISO/IEC") | str("IEEE"))
-        # ISO stage codes: FDIS, FCD, CDV; DIS/CD with an optional round
-        # digit (DIS2, CD1..CD4); WD/PWI/NP.
-        iso_stage = ((str("FDIS") >> digit.maybe) | (str("FCD") >> digit.maybe) |
-                     (str("CDV") >> digit.maybe) |
-                     (str("DIS") >> digit.maybe) |
-                     (str("CD") >> digit.maybe) |
-                     (str("WD") >> digit.maybe) | (str("PWI") >> digit.maybe) |
-                     (str("NP") >> digit.maybe)).as(:iso_stage)
-        std_noise = (space >> str("Std")).maybe
-        # The empty-string marker distinguishes the stage-less PUBLISHED
-        # tree in the builder (absence of :iso_stage alone means IEEE format).
-        ((iso_led.as(:joint_publishers) >> space >>
-           (iso_stage >> std_noise >> space |
-            str("").as(:iso_published) >> space.maybe)) |
-          (staged_only.as(:joint_publishers) >> space >>
-            iso_stage >> std_noise >> space)) >>
-          str("P").as(:project_marker).maybe >> # project marker (P =
-                                                # project/draft; its
-                                                # presence is identity)
-          digits.as(:number) >>
-          # part must not swallow the trailing year (year_digits.absent?). A
-          # DASH-joined part is tagged (:part_dash) - the catalogue-printed
-          # joint form spells the part with a dash ("21451-7") where the ISO
-          # form uses the dot.
-          ((dot >> year_digits.absent? >> digits.as(:part)) |
-           (dash >> str("").as(:part_dash) >> year_digits.absent? >>
-             digits.as(:part))).maybe >>
-          (
-            (str(":") >> year_digits.as(:year)) |
-            # The dash-year (and ", Month YYYY" text date below) spellings are
-            # the catalogue-PRINTED joint form; tag them (:printed_form) so the
-            # builder routes to a Standard instead of the ISO joint reference.
-            (dash >> year_digits.as(:year) >>
-             (dash >> month_numeric.as(:month)).maybe >>
-             str("").as(:printed_dash_year)) |
-            # Trailing text date ", April 2015" / " April 2015" (build_joint_development
-            # already reads :month/:year). No :year collision — the iso rule has no
-            # other :year capture (edition uses :edition_year).
-            ((comma | space) >> month_name.as(:month) >> space >>
-             year_digits.as(:year) >> str("").as(:printed_month_year))
-          ).maybe >>
-          # Optional /D<draft> tail; the historical hyphenated form
-          # "…/D-3-2017" carries its date inside the draft clause (bucket 5);
-          # a date-less "/D-4" keeps its hyphen (bucket 7), hence dash.maybe.
-          # A text date may trail the DRAFT itself (pubid#216:
-          # "CD P26515/D1, March 2017", "FDIS P15289/D3, 2017") — distinct
-          # keys (:draft_month/:draft_year), because a plain :month/:year
-          # here collides with the date clause above and parslet drops the
-          # subtree with a "Duplicate subtrees" warning.
-          # The ISO/IEC edition marker may sit between the year and an
-          # amendment/draft tail: "8802.11:2012 (E)/Amd 1-2014". A distinct
-          # key - a second :parameters capture here collides with the
-          # trailing parenthetical slot (parslet drops duplicate subtrees).
-          (space.maybe >> str("(") >>
-           match("[^)]").repeat(1).as(:edition_marker) >> str(")")).maybe >>
-          # The stage-draft clause of the ISO-led print: "/D=WD.5" -
-          # D (draft) = the ISO/IEC stage it drafts, with its iteration
-          # (docs/IEEE-DRAFT-STAGES.md §1.2).
-          ((slash >> str("D") >> str("=") >>
-           ((str("FDIS") >> digit.maybe) | (str("FCD") >> digit.maybe) |
-            (str("CDV") >> digit.maybe) |
-            (str("DIS") >> digit.maybe) |
-            (str("CD") >> digit.maybe) |
-            (str("WD") >> digit.maybe) | (str("PWI") >> digit.maybe) |
-            (str("NP") >> digit.maybe)).as(:draft_iso_stage) >>
-           (dot >> (digits >> match("[A-Za-z]").repeat(0, 1)).as(:draft_iso_iteration)).maybe) |
-          (slash >> str("D") >> dash.maybe >>
-           match('[0-9.]').repeat(1).as(:draft_version) >>
-           (((comma | space) >> month_name.as(:draft_month) >> space >>
-               year_digits.as(:draft_year)) |
-             (comma >> year_digits.as(:draft_year))).maybe).maybe).maybe >>
-          # Optional amendment tail (pubid#317:
-          # "8802-11:2012/Amd.1:2014(E)") - the flat tree keys reuse
-          # build_flat_amendment.
-          (slash >> str("Amd") >> (dot | space).maybe >>
-           digits.as(:amd_number) >>
-           ((str(":") | dash) >> year_digits.as(:amd_year)).maybe).maybe >>
-          # Optional corrigendum tail, the ISO print's dotted spelling
-          # (pubid#215): "8802-1AC:2018/Cor.1:2020(E)". The shared
-          # corrigendum rule accepts the slash/dot/colon separators; the
-          # flat tree routes through build_flat_corrigendum.
-          corrigendum.maybe >>
-          # Optional edition, from relaton's "/E-<n>" suffix normalized to
-          # "Edition <n>.0[ YYYY]" (nil-residue hand-off item 1).
-          edition.maybe >>
-          revision_suffix.maybe >>
-          # A redline is a distinct document of the joint ISO print
-          # (pubid#215): "15289:2015(E) - Redline". Same flag semantics as
-          # the generic rules - the builder sets redline: true.
-          redline.maybe >>
-          # Published joint docs print a language marker, after either the
-          # year or the amendment tail: "9945:2009(E)", "...Amd.1:2014(E)".
-          # The crawl sometimes spaces it ("…(E), January 2017" family,
-          # pubid#216), and a relationship parenthetical may follow
-          # ("…CD2 P15288-2013-09 (Revision of …)").
-          parenthetical.maybe
-      end
-
-      # Embedded (stage-LAST) ISO-led designations: the corpus writes the ISO
-      # stage AFTER the (dotted or dashed) part —
-      #   "ISO/IEC/IEEE 29119.4.FDIS, April 2015"   (dot part .4, stage .FDIS)
-      #   "ISO/IEC/IEEE 24748-5.CD3, February 2015" (dash part -5, stage .CD3)
-      #   "IEEE P24748.5.CD3, July 2015"            (bare IEEE + P)
-      # — rather than before the number (joint_development_iso_format). Same
-      # meaning as the stage-first form: `29119.4`/`29119-4` = "29119 part 4" and
-      # the trailing `.FDIS` is the ISO stage, NOT a second part. Routes through
-      # the same build_joint_development (via :joint_publishers/:iso_stage), so the
-      # stage is modeled correctly and the numeric part stays separate.
-      # The stage-less ISO/IEC label with a colon year and the ISO/IEC
-      # edition marker: "ISO/IEC 13210:1994 (E)" - the ISO/IEC portion of a
-      # double-labeled standard. The (E) parenthetical is REQUIRED: without
-      # it this spelling is the ISO flavor's own form and must not be
-      # stolen. Routes through build_joint_development like every ISO-led
-      # joint reference.
-      rule(:joint_development_iso_iec_edition) do
-        str("ISO/IEC").as(:joint_publishers) >> space >>
-          str("").as(:iso_published) >>
-          str("P").as(:project_marker).maybe >>
-          digits.as(:number) >>
-          # The part may be dot- or dash-joined ("8802-9"); a dash-year
-          # ("1234-2012") is not a part.
-          # A distinct key: the iso-format route's :part_dash prints as a
-          # dot in the joint code (the "P24748.2:2018" convention); only
-          # the ISO/IEC LABEL keeps its printed dash.
-          ((dot | (dash >> str("").as(:iec_label_dash) >> year_digits.absent?)) >>
-            digits.as(:part)).maybe >>
-          (str(":") >> space.maybe >> year_digits.as(:year)) >>
-          space.maybe >> parenthetical
-      end
-
-      rule(:joint_development_embedded_stage) do
-        # publisher set mirrors joint_development_iso_format (incl. bare IEEE);
-        # longest token first.
-        (str("ISO/IEC/IEEE") | str("IEEE/ISO/IEC") | str("IEEE/IEC/ISO") |
-         str("ISO/IEEE") | str("IEC/IEEE") | str("IEEE/IEC") | str("ISO/IEC") |
-         str("IEEE")).as(:joint_publishers) >>
-          space >>
-          str("P").as(:project_marker).maybe >> # project marker (P =
-                                                # project/draft; its
-                                                # presence is identity)
-          digits.as(:number) >>
-          # optional numeric part (dot or dash); must not swallow a year
-          ((dot | dash) >> year_digits.absent? >> digits.as(:part)).maybe >>
-          # the embedded stage, dot-separated, closed vocab (same as iso_stage).
-          # The (digit|letter) look-ahead keeps it at a token boundary so "CD"
-          # can't match inside a longer token and a glued no-space date
-          # (".DISMay2013") is left to fall through.
-          dot >>
-          ((str("FDIS") >> digit.maybe) | (str("FCD") >> digit.maybe) |
-           (str("CDV") >> digit.maybe) |
-           (str("DIS") >> digit.maybe) |
-           (str("CD") >> digit.maybe) |
-           (str("WD") >> digit.maybe) | (str("PWI") >> digit.maybe) |
-           (str("NP") >> digit.maybe)).as(:iso_stage) >>
-          (digit | match("[A-Za-z]")).absent? >>
-          # optional trailing date: :YYYY, -YYYY[-MM], or text "Month YYYY"
-          (
-            (str(":") >> year_digits.as(:year)) |
-            (dash >> year_digits.as(:year) >>
-             (dash >> month_numeric.as(:month)).maybe) |
-            ((comma | space) >> month_name.as(:month) >> space >> year_digits.as(:year))
-          ).maybe
-      end
-
-      # Number-first pattern: "1873-2015 IEEE Standard..."
-      rule(:number_first_identifier) do
-        number >>
-          (dash >> year_digits.as(:year)).maybe >>
-          space >>
-          (publisher >> copublisher.repeat.as(:copublishers)).as(:publishers) >>
-          space >>
-          (type_word.as(:type) >> space?).maybe >>
-          match("[^\n]").repeat(0).as(:title)
-      end
-
-      # IEEE P pattern (without Std): "IEEE P1003.1..." OR just "P1003.1..." (prefix optional)
-      # `P` is not consumed here: the `number` rule carries its own optional
-      # leading P and captures it into :number ("P1003.1"), so Code.parse
-      # peels it as code.prefix and the project marker survives the parse —
-      # a bare `str("P")` swallowed it and "P1201/D0.3" came back without
-      # its P at all (pubid#18's data-loss complaint).
-      rule(:ieee_p_identifier) do
-        (str("IEEE").as(:publisher) >> space).maybe >> # Make IEEE prefix optional
-          number >>
-          (part_subpart_year | edition).maybe >>
-          # Pattern for /08 style drafts (digits without D prefix) - MUST come before corrigendum
-          (slash >> digits.as(:draft_version)).as(:digit_draft).maybe >>
-          # FDIS and other ISO stage codes without D prefix (Pattern 3)
-          fdraft.maybe >>
-          # Trailing "Month YYYY" print/reaffirm date. Captured under distinct
-          # keys so it never collides with the base -YYYY identity year (a
-          # collision made Parslet drop the base year and warn "Duplicate
-          # subtrees … keys: [:year]"). The builder promotes it to the identity
-          # only when there is no base year.
-          trailing_month_year.maybe >>
-          corrigendum.maybe >>
-          draft.maybe >>
-          # Revision trails the draft (before any date), matching normalization's
-          # ".../D<n>/R-<x>" repositioning of "P802.16Rev2/D3 Feb 2008".
-          revision_suffix.maybe >>
-          # ALSO accept month/year after draft (some patterns like /DX, Month YEAR)
-          trailing_month_year.maybe >>
-          # Trailing corrigendum AFTER the draft+date ("…/D1, Jan 2007/Cor. 1").
-          # The draft's own draft_date has already consumed the date, leaving
-          # "/Cor. 1" here; the resulting flat corrigendum+draft tree routes to
-          # build_flat_corrigendum (disjoint from the pre-draft corrigendum slot,
-          # which only fires when no draft precedes it).
-          corrigendum.maybe >>
-          parenthetical.maybe
-      end
-
-      # ANSI P pattern: "ANSI PN42.34-D9a, 2015" OR "ANSI P1234/D5"
-      rule(:ansi_p_identifier) do
-        str("ANSI").as(:publisher) >> space >>
-          str("P") >> space.maybe >> # Make space after P optional
-          number >>
-          (part_subpart_year | edition).maybe >>
-          # Trailing "Month YYYY"/bare-year date under distinct keys so they
-          # never collide with the base -YYYY identity year (see ieee_p_identifier).
-          trailing_month_year.maybe >>
-          corrigendum.maybe >>
-          draft.maybe >>
-          revision_suffix.maybe >>
-          # ALSO accept month/year after draft
-          trailing_month_year.maybe >>
-          # Accept bare year after draft: ", 2015"
-          ((comma | space) >> year_digits.as(:trailing_year)).maybe >>
-          # Trailing corrigendum after the draft+date (parity with ieee_p_identifier).
-          corrigendum.maybe >>
-          parenthetical.maybe
-      end
-
-      # IEEE Draft P pattern: "IEEE Draft P802.11..." OR "Draft P802.11..." (IEEE prefix optional)
-      # An optional status phrase may precede "Draft" — "IEEE Unapproved Draft P…",
-      # "IEEE Active Approved Draft P…" — the largest index-v2 draft bucket. The
-      # status is captured (draft_status ends in a space, so "Approved Draft"
-      # splits cleanly) and round-trips as the draft_status attribute; the literal
-      # "Draft" stays a bare marker (dropped on render, like the plain form).
-      # `P` is not consumed here: the `number` rule carries its own optional
-      # leading P and captures it into :number ("P802.11"), so Code.parse peels
-      # it as code.prefix and the project marker round-trips (pubid#318) — a
-      # bare `str("P").maybe` here swallowed it and the render lost the P.
-      rule(:ieee_draft_p_identifier) do
-        (str("IEEE").as(:publisher) >> space).maybe >> # Make IEEE prefix optional
-          draft_status.as(:draft_status).maybe >>
-          str("Draft") >> space >>
-          # A status-word draft may carry a bare number
-          # ("IEEE Unapproved Draft 802.1ah/D4.2"). The `number` rule
-          # already accepts the bare forms (802.1ah, C57.15, 11073-10471).
-          number >>
-          (part_subpart_year | edition).maybe >>
-          # Trailing "Month YYYY" date under distinct keys so it never collides
-          # with the base -YYYY identity year (see ieee_p_identifier).
-          ((space >> month_name.as(:trailing_month) >> space >> year_digits.as(:trailing_year)) |
-           (space >> month_numeric.as(:trailing_month) >> (space | dash) >> year_digits.as(:trailing_year))).maybe >>
-          draft.maybe >>
-          # A print date may trail a draft that carries its own date
-          # ("IEEE Unapproved Draft 24765/D1 2009, Oct 2009" - the draft's
-          # year is "2009", the print date is ", Oct 2009").
-          (comma >> space? >> month_name.as(:trailing_month) >> space >>
-            year_digits.as(:trailing_year)).maybe >>
-          revision_suffix.maybe >>
-          # …and the print date may also trail the repositioned revision
-          # ("IEEE Unapproved Draft P802.16Rev2/D9a, March 2009" reaches the
-          # grammar with its inline revision parsed natively).
-          (comma >> space? >> month_name.as(:trailing_month) >> space >>
-            year_digits.as(:trailing_year)).maybe >>
-          # Trailing corrigendum after the draft ("…/D2.0/Cor. 1", or
-          # "…/D1.0, Dec 2007/Cor. 1" where the draft's own draft_date consumes
-          # the date, leaving "/Cor. N"). The flat corrigendum+draft tree (no
-          # :base) routes to build_flat_corrigendum, which rebuilds the base
-          # standard (carrying the draft + draft_status) and wraps it.
-          corrigendum.maybe >>
-          parenthetical.maybe
-      end
-
-      # IEEE Approved Draft pattern: "IEEE Approved Draft Std P..."
-      rule(:ieee_approved_draft_identifier) do
-        str("IEEE").as(:publisher) >>
-          space >>
-          str("Approved") >> space >>
-          (str("Draft Std") | str("Std")).as(:type) >> space >>
-          str("P").maybe >>
-          number >>
-          (part_subpart_year | edition).maybe >>
-          draft.maybe >>
-          revision_suffix.maybe >>
-          parenthetical.maybe
-      end
-
-      # Combined AIEE identifier pattern: "AIEE No 72-1932 and AIEE No 73-1932"
-      # Handles "and"-separated AIEE identifiers (from "Nos X and Y" preprocessing)
-      rule(:combined_aiee_identifier) do
-        # First AIEE identifier
-        Aiee::Parser.new.aiee_identifier.as(:first_aiee) >>
-          # "and" separator
-          space >> str("and") >> space >>
-          # Second AIEE identifier
-          Aiee::Parser.new.aiee_identifier.as(:second_aiee)
-      end
-
-      # AIEE (American Institute of Electrical Engineers) patterns
-      # Detect AIEE patterns and delegate to AIEE parser
-      rule(:aiee_identifier) do
-        # Lookahead for AIEE patterns - do not consume input
-        (
-          # IEEE-AIEE transitional pattern
-          (str("IEEE-AIEE") >> space >> (str("No.") | str("Nos") | str("No") | str("Standard") | str("Trans."))) |
-          # A.I.E.E. pattern (with dots, no spaces)
-          (str("A.I.E.E.") >> space >> (str("No.") | str("Nos") | str("No"))) |
-          # A. I. E. E. pattern (with dots and spaces)
-          (str("A. I. E. E.") >> space >> (str("No.") | str("Nos") | str("No") | str("Standard"))) |
-          # AIEE pattern - extended to include more type words
-          (str("AIEE") >> space >> (str("No.") | str("Nos") | str("No") | str("Standard") | str("Trans.") | str("Std")))
-        ).present? >>
-          # Delegate to AIEE parser if pattern detected
-          Aiee::Parser.new.aiee_identifier.as(:aiee)
-      end
-
-      # IRE (Institute of Radio Engineers) patterns
-      # Detect IRE patterns and delegate to IRE parser
-      rule(:ire_identifier) do
-        # Lookahead for IRE patterns - do not consume input
-        (
-          # Year-first pattern: "52 IRE 7.S2" or "60 IRE 28 PS7"
-          ((match("[1-6]") >> digit >> space >> str("IRE")) | # 2-digit year format
-           (str("19") >> digit.repeat(2, 2) >> space >> str("IRE"))) |
-          # Prefix-first pattern: "IRE 7.S2-1952" (pubid#316 family 2) —
-          # IRE then a designation (digits or a type word). Loose on
-          # purpose: the sub-parser is strict, and a failed delegation just
-          # falls through to the remaining alternatives.
-          (str("IRE") >> space >> (digits | str("Standard") | str("Std") | str("Trans"))) |
-          # IEEE-IRE transitional pattern
-          (str("IEEE-IRE") >> space)
-        ).present? >>
-          # Delegate to IRE parser if pattern detected
-          Ire::Parser.new.ire_identifier.as(:ire)
-      end
-
-      # NESC (National Electrical Safety Code) patterns
-      # Detect NESC patterns and delegate to NESC parser
-      rule(:nesc_identifier) do
-        # Lookahead for NESC patterns - do not consume input
-        (
-          # C2-YYYY pattern
-          (str("C2-") >> year_digits) |
-          # YYYY NESC pattern
-          (year_digits >> space >> (str("NESC") | str("National Electrical Safety Code"))) |
-          # Draft NESC pattern
-          (str("Draft") >> space >> (str("NESC") | str("National Electrical Safety Code"))) |
-          # Name-first pattern (NEW)
-          (str("National Electrical Safety Code") >> str(",") >> space >> str("C2-")) |
-          # Catalogue form: "IEEE Std YYYY NESC ..." / "IEEE Std YYYY National
-          # Electrical Safety Code" - the prefix is consumed here (the NESC
-          # renderer prepends "IEEE Std" itself) but must not fall through to
-          # the generic standard grammar, which would read YYYY as a number.
-          (str("IEEE") >> space >> str("Std") >> space >> year_digits >> space >>
-            (str("NESC") | str("National Electrical Safety Code")))
-        ).present? >>
-          # Delegate to NESC parser if pattern detected
-          ((str("IEEE") >> space >> str("Std") >> space).maybe >>
-            Nesc::Parser.new.nesc_identifier).as(:nesc)
-      end
-
-      # Draft notation for PSI (e.g., /D2, /D3, rawbib's /D-2)
-      rule(:psi_draft) do
-        slash >> str("D") >> dash.maybe >> digits.as(:draft_version)
-      end
-
-      # Date on an SI/PSI identifier: ", Month Year" or "-YEAR[-MM]". The
-      # comma branch also accepts a NUMERIC month and a DASH before the
-      # year — the renderer emits ", 05 2010" for a numeric-month draft,
-      # and the preprocessing gsub rewrites a trailing " <digits> <year>"
-      # to "<digits>-<year>" (the same load-bearing coupling draft_date
-      # documents), so the re-parse must read its own rendering back.
-      rule(:psi_date) do
-        (comma >> (month_name | month_numeric).as(:month) >> (space | dash) >>
-           year_digits.as(:year)) |
-          (dash >> year_digits.as(:year) >>
-            (dash >> digit.repeat(2, 2).as(:month)).maybe)
-      end
-
-      # IEEE/ASTM SI/PSI (Système International) patterns
-      # SI = Published metric system standard
-      # PSI = Proposed SI (draft)
-      # Rawbib spellings (pubid#316 family 3): a bare "IEEE" publisher, a dot
-      # separator after the type ("PSI.10", "SI 10.1997"), relaton's
-      # hyphenated "/D-<n>" draft, and a dash year-month ("-2010-05").
-      rule(:ieee_astm_si_psi) do
-        (str("IEEE/ASTM") | str("IEEE")).as(:publishers) >>
-          space >>
-          (str("PSI") | str("SI")).as(:si_type) >>
-          (space | str(".")) >>
-          digits.as(:number) >>
-          # Glued ".YEAR" edition date ("IEEE/ASTM SI 10.1997")
-          (str(".") >> year_digits.as(:year)).maybe >>
-          # The date and the draft appear in EITHER order: the legacy
-          # spelling puts the draft first ("PSI 10/D2, October 2015"),
-          # the rawbib hyphenated form ("PSI 10/D-3-2010") parses in place
-          # with its date on the draft clause. Draft-first is tried first so the legacy
-          # comma-date keeps its original match; each side is optional so
-          # a date-only or draft-less form still parses.
-          (
-            (psi_draft >> psi_date.maybe) | (psi_date >> psi_draft.maybe)
-          ).maybe >>
-          # Optional parenthetical (revision relationships)
-          parenthetical.maybe
-      end
-
-      # No-prefix IEEE identifier (characteristic patterns without "IEEE Std")
-      # These are patterns that are distinctly IEEE even without explicit publisher
-      rule(:no_prefix_ieee) do
-        characteristic_ieee_number.as(:number) >>
-          # Optional suffix (like -a, -b)
-          (dash >> match("[A-Za-z]")).maybe.as(:suffix) >>
-          # Optional year
-          (dash >> year_digits).maybe.as(:year) >>
-          # Optional draft notation
-          draft.maybe >>
-          # Optional language portion
-          (str("(E)") | str("(F)")).maybe >>
-          # Optional parenthetical content
-          parenthetical.maybe
-      end
-
-      # Corrigendum identifier with recursive base parsing
-      # Captures base identifier for recursive parsing, then corrigendum supplement
-      # Example: IEEE Std 535-2013/Cor. 1-2017
-      rule(:corrigendum_identifier) do
-        # Match a complete base identifier (reuse existing patterns)
-        # Try standard patterns that would match "IEEE Std 535-2013"
-        (
-          ((publisher >> copublisher.repeat.as(:copublishers)).as(:publishers) >> space).maybe >>
-          (type_word.as(:type) >> space?).maybe >>
-          number >>
-          part_subpart_year.maybe # This captures the full identifier before /Cor
-        ).as(:base) >>
-          # Now match the corrigendum portion
-          (slash | dash | space) >>
-          str("Cor") >>
-          (dash | dot | space).maybe >> # More flexible separator after "Cor"
-          space? >>
-          digits.as(:cor_number) >>
-          ((dash | str(":") | space) >> year_digits.as(:cor_year)).maybe >> # Optional cor year suffix
-          parenthetical.maybe
-      end
-
-      # Interpretation identifier with recursive base parsing
-      # Captures base identifier for recursive parsing, then interpretation supplement
-      # Example: IEEE Std 1076/INT-1991, IEEE Std 1003.1-1988/INT
-      rule(:interpretation_identifier) do
-        # Match a complete base identifier
-        (
-          ((publisher >> copublisher.repeat.as(:copublishers)).as(:publishers) >> space).maybe >>
-          (type_word.as(:type) >> space?).maybe >>
-          number >>
-          part_subpart_year.maybe
-        ).as(:base) >>
-          # Now match the interpretation portion
-          (slash | dash | space) >>
-          str("INT") >>
-          ((dash | str(":") | space) >> year_digits.as(:int_year)).maybe >> # Optional year suffix
-          parenthetical.maybe
-      end
-
-      # Conformance identifier with recursive base parsing
-      # Captures base identifier for recursive parsing, then conformance supplement
-      # Example: IEEE Std 802.16/Conformance01-2003
-      rule(:conformance_identifier) do
-        # Match a complete base identifier
-        (
-          ((publisher >> copublisher.repeat.as(:copublishers)).as(:publishers) >> space).maybe >>
-          (type_word.as(:type) >> space?).maybe >>
-          number >>
-          part_subpart_year.maybe
-        ).as(:base) >>
-          # Now match the conformance portion
-          (slash | dash | space) >>
-          str("Conformance") >>
-          match("[0-9]").repeat(1).as(:conf_number) >>
-          dash >>
-          year_digits.as(:conf_year) >>
-          parenthetical.maybe
-      end
-
-      # Multi-numbered identifier: same document with multiple numbers
-      # Examples: IEEE Std 1299/C62.22.1-1996, IEEE Std 960-1989, Std 1177-1989
-      rule(:multi_numbered_identifier) do
-        # Primary identifier (full IEEE identifier)
-        ((
-          (publisher >> space).maybe >>
-          (type_word.as(:type) >> space?).maybe >>
-          number >>
-          (part_subpart_year | edition).maybe
-        ).as(:primary_identifier) >>
-          # Separator: slash for cross-ref format, comma for joint standard
-          (slash >> str("C") >> digits >> dot >> digits >> dot >> digits >> dash >> year_digits).as(:secondary_crossref)) |
-          (comma >> space >> (type_word.as(:type) >> space?).maybe >> number >> dash >> year_digits).as(:secondary_joint)
-      end
-
-      # CSA dual published pattern: IEEE Std 844.1-2017/CSA C22.2 No. 293.1-17
-      rule(:csa_dual_published) do
-        # IEEE portion (full identifier)
-        (
-          publisher >> space >>
-          (type_word.as(:type) >> space?).maybe >>
-          number >>
-          (part_subpart_year | edition).maybe
-        ).as(:ieee_portion) >>
-          # CSA portion with slash separator
-          slash >>
-          str("CSA") >> space >>
-          # CSA number formats (various patterns observed)
-          (
-            # Format 1: C22.2 No. 293.1-17 (with NO., either case)
-            (str("C") >> digit.repeat(2) >> dot >> digit >> space >> (str("No") | str("NO")) >> dot >> space >>
-             match("[0-9.]").repeat(1) >> (dash | str(":")) >> digit.repeat(2)) |
-            # Format 2: C293.2-17 (without NO., dash year)
-            (str("C") >> match("[0-9.]").repeat(1) >> dash >> digit.repeat(2)) |
-            # Format 3: C22.2 No. 293.3:19 (with NO., either case, colon year)
-            (str("C") >> digit.repeat(2) >> dot >> digit >> space >> (str("No") | str("NO")) >> dot >> space >>
-             match("[0-9.]").repeat(1) >> str(":") >> digit.repeat(2)) |
-            # Format 4: C293.4:19 (without NO., colon year)
-            (str("C") >> match("[0-9.]").repeat(1) >> str(":") >> digit.repeat(2))
-          ).as(:csa_portion)
-      end
-
-      # Basic IEEE identifier (no dual PubIDs or complex revisions yet)
-      rule(:identifier) do
-        combined_aiee_identifier |
-          aiee_identifier |
-          combined_aiee_identifier |
-          ire_identifier |
-          nesc_identifier |
-          ieee_astm_si_psi | # NEW Session 171: Add IEEE/ASTM SI/PSI support
-          multi_numbered_identifier | # NEW: Try multi-numbered identifiers before generic patterns
-          csa_dual_published | # NEW: Try CSA dual published before generic patterns
-          s_designation | # Historical IEEE/IPCEA cable designation (S-135)
-          corrigendum_identifier | # NEW: Try corrigendum before generic patterns
-          interpretation_identifier | # NEW: Try interpretation identifier before generic patterns
-          conformance_identifier | # NEW: Try conformance identifier before generic patterns
-          joint_development_ieee_format |
-          joint_development_iso_format |
-          joint_development_iso_iec_edition | # stage-less ISO/IEC label with (E) - double-label portion
-
-          joint_development_embedded_stage | # stage-LAST embedded form (before generic)
-          iec_ieee_copublished |
-          number_first_identifier |
-          ieee_approved_draft_identifier |
-          ieee_draft_p_identifier |
-          ieee_p_identifier |
-          ansi_p_identifier | # NEW: ANSI P prefix support
-          (((publisher >> copublisher.repeat.as(:copublishers)).as(:publishers) >> space).maybe >> # Make publisher optional
-          draft_status.as(:draft_status).maybe >>
-          (str("Draft Std").as(:type) >> space?).maybe >>
-          (type_word.as(:type) >> (space >> str("No") >> space).maybe >> space?).maybe >>
-          number >>
-          (part_subpart_year | edition).maybe >>
-          corrigendum.maybe >>
-          amendment.maybe >>
-          interpretation.maybe >> # NEW: Add /INT support
-          conformance.maybe >> # NEW: Add /Conformance support
-          ashrae_copub.maybe >> # NEW: Add /ASHRAE Guideline support
-          ieee_crossref.maybe >> # NEW: Add /C62.22.1-1996 cross-reference support
-          draft.maybe >>
-          revision_suffix.maybe >>
-          # Trailing "Month YYYY" print/reaffirm date under distinct keys so it
-          # never collides with the base -YYYY identity year (see
-          # ieee_p_identifier). The builder promotes it only when no base year.
-          trailing_month_year.maybe >>
-          # The ASHRAE joint suffix may also trail a draft+date
-          # ("IEEE P1635/D13, December, 2017/ASHRAE Guideline 21"), so a
-          # second slot catches it after the date.
-          ashrae_copub.maybe >>
-          # A bracketed narrative may sit between the base and a trailing
-          # corrigendum: "IEEE Std 671-1985 [Corrigendum to IEEE Std 671-1985
-          # (Reaff 2008)]/Cor. 1-2010" captures the bracket as the base's
-          # nickname before the wrapper is parsed.
-          book_nickname.maybe >>
-          # Trailing corrigendum after the draft+date ("IEEE Approved P1015/D1,
-          # Jan 2007/Cor. 1"): the generic bucket is the only path a status-word
-          # form reaches. Routes via build_flat_corrigendum (disjoint from the
-          # pre-draft corrigendum slot at line 954).
-          corrigendum.maybe >>
-          edition.maybe >>
-          parenthetical.maybe >>  # REVERT: Back to single parenthetical
-          book_nickname.maybe >>  # NEW: Add book nickname support
-          redline.maybe >>
-          title_portion.maybe >>
-          approved_draft_suffix.maybe) |
-          no_prefix_ieee # NEW: Try no-prefix patterns last (lowest priority)
-      end
-
-      root(:identifier)
+    class Parser
       def self.normalize_input(string)
         # Strip .pdf extension if present (Pattern 3: File Extensions)
         cleaned = string.sub(/\.pdf$/i, "")
@@ -1269,7 +19,6 @@ module Pubid
         # No valid IEEE identifier pattern needs more than 1 space
         cleaned = cleaned.gsub(/\s+/, " ")
 
-
         # NEW Session 171: Remove wrong ! prefix
         cleaned = cleaned.gsub(/^!IEEE /, "IEEE ")
 
@@ -1647,7 +396,6 @@ module Pubid
         # Remove period after "Std": "IEEE Std." -> "IEEE Std"
         cleaned = cleaned.gsub(/\bStd\.\s+/, "Std ")
 
-
         # Title portion removal after year: "YYYY - IEEE Standard for..."
         cleaned = cleaned.gsub(
           /(\d{4})(\s+\([^)]+\))?\s+-\s+IEEE\s+Standard\s+for.*$/, '\1\2'
@@ -1787,10 +535,6 @@ module Pubid
 
         cleaned
       end
-
-      def self.parse(string)
-        new.parse(normalize_input(string))
-      end
     end
   end
 end