pubid 2.0.0.pre.alpha.8 → 2.0.0.pre.alpha.11

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 5987bc5bdc86f9fdf02244f7b5e1abbff308a0d7a5ae88c9ea6e26d20f316ec4
-  data.tar.gz: d768df87e4d1b06d0be0dbcc72dab8d5e5eda5476b4157bbf754c5863a2ad1f0
+  metadata.gz: 18402e85b2e19640920f3e216f62f704c6770a840f654d92f96d799b03fbda71
+  data.tar.gz: e9b1735e843fe5eb20574749d2581c19192799f86fe3f31a8e72b3d0a00ab047
 SHA512:
-  metadata.gz: 7214ebd2c2eac905def41dd1b4304ef299224378e703ab3d9883022b8fd2770adafd08c8e697d759a9ae6e2a984ff54927051090039c8b20cd56f98ae21bc3b8
-  data.tar.gz: 32227b91be942f7f59c758685ea7cbe105845eaa758c29984d43f81adeaf33321e68f6b4f7a8d9d71709f3395acf7011d008e1a2983ae43a61610b267c785d0e
+  metadata.gz: 1cbcaa48b3fb036aca2a62aabe9cdfc4733ad030dc833aa31b5be2dbba433ef6459f87dc22c0b811e707687ce18ee966a2882f928a44d2d0df89e159df5ea837
+  data.tar.gz: 0ea8984849aa9f56a400948fb5ab7b29f5de122f70ab036d35fe344655abe70fbe9e34d9cdc04b80673593190b1479d8ac47b31bd21e056ad9b64fa6743e18e6

data/README.adoc

@@ -147,18 +147,35 @@ id.to_json
 
 === Import from Hash
 
+`Pubid.from_hash` is the hash counterpart of `Pubid.parse`. It builds the
+correct identifier class from a hash that `#to_hash` emits. You do not name
+the flavor: the `_type` key (`pubid:<flavor>:<type>`) identifies the class.
+
 [source,ruby]
 ----
-require 'pubid/iso'
+require 'pubid'
 
 hash = {
-  flavor: "iso",
-  publisher: "ISO",
-  number: "9001",
-  year: "2015"
+  "_type" => "pubid:iso:international-standard",
+  "number" => "9001",
+  "year" => "2015"
 }
-id = Pubid::Serializable.from_h(hash)
-id.to_s # => "ISO 9001:2015"
+id = Pubid.from_hash(hash)
+id.class # => Pubid::Iso::Identifiers::InternationalStandard
+id.to_s  # => "ISO 9001:2015"
+----
+
+For a row of a relaton-data `index-v2.yaml` file, give the identifier hash in
+the `:id` key of the row: `Pubid.from_hash(row[:id])`.
+
+A hash that is not an identifier hash raises `Pubid::Errors::InvalidInputError`
+(an `ArgumentError`). This occurs when the input is not a Hash, has no `_type`,
+or has a `_type` that no flavor defines.
+
+[source,ruby]
+----
+Pubid.from_hash({ "number" => "9001" })
+# => Pubid::Errors::InvalidInputError: identifier hash has no _type key: ...
 ----
 
 === Import from JSON
@@ -181,11 +198,15 @@ All identifiers support full round-trip conversion preserving all attributes:
 require 'pubid/iso'
 
 original = Pubid::Iso.parse("ISO 9001:2015/Amd 1:2020/Cor 1:2021")
-hash = original.to_h
-restored = Pubid::Serializable.from_h(hash)
+hash = original.to_hash
+# => {"_type" => "pubid:iso:corrigendum", "number" => "1", "year" => "2021",
+#     "base" => {"_type" => "pubid:iso:amendment", "number" => "1", "year" => "2020",
+#       "base" => {"_type" => "pubid:iso:international-standard",
+#                  "number" => "9001", "year" => "2015"}}}
+restored = Pubid.from_hash(hash)
 
 restored.to_s # => "ISO 9001:2015/Amd 1:2020/Cor 1:2021"
-restored.to_urn # => "urn:iso:std:iso:9001:amd:2020:cor:2021:v1"
+restored.to_urn # => "urn:iso:std:iso:9001:amd:2020:v1:cor:2021:v1"
 ----
 
 == Utility Methods
@@ -1756,7 +1777,8 @@ require 'pubid'
 
 # update_codes.yaml: "NBS HB: NIST HB"
 id = Pubid::Nist.parse("NBS HB 105-1")
-id.to_s  # => "NIST HB 105-1" (NBS→NIST correction applied)
+id.to_s  # => "NBS HB 105-1" (NBS identity preserved - documents published
+         # by NBS, i.e. before/equal 1988, keep their NBS identifiers forever)
 
 # update_codes.yaml: "FIPS.140-2: FIPS 140-2" (normalizes dotted format)
 id = Pubid::Nist.parse("FIPS.140-2")

data/conformance/pending.yaml

@@ -0,0 +1,4 @@
+# Cases this implementation explicitly cannot handle yet.
+# Rules: entries need reason + ref; a pending case that passes raises
+# the pending-satisfied alarm and MUST be unmarked (TODO.correct-test-suite/04).
+---

data/data/bipm/update_codes.yaml

@@ -0,0 +1,9 @@
+# Legacy / docnumber-style BIPM identifier spellings normalized to a canonical,
+# parseable Pubid::Bipm form before parsing. Keys are full-line-anchored plain
+# strings (Regexp.escaped) or /regex/ literals — see lib/pubid/core/update_codes.rb.
+#
+# CIPM MRA:2005 interpretation document (relaton-data-bipm docnumber
+# "CIPM/2005-06(REV)"): strip the slash and the "(REV)" revision marker so both
+# the docnumber and docidentifier spellings map onto the bare "CIPM 2005-06" form.
+"CIPM/2005-06(REV)": CIPM 2005-06
+"CIPM 2005-06(REV)": CIPM 2005-06

data/data/iec/update_codes.yaml

@@ -1,4 +1,7 @@
 IECEx TRF 60079-11v3-4:2002: IECEx TRF 60079-11v4C:2002
+# IECEE TRF data-source anomaly: the OD-010-1 ampersand separator (&) was
+# mistyped as a semicolon (;) in IEC's source data. See pubid/pubid#239.
+IECEE TRF 60335-2-4;7F:2012: IECEE TRF 60335-2-4&7F:2012
 IECEE TRF cispr 15N:2022: IECEE TRF CISPR 15N:2022
 IECEE TRF cispr 15M:2021: IECEE TRF CISPR 15M:2021
 IECEE TRF cispr 15L:2020: IECEE TRF CISPR 15L:2020

data/data/ieee/update_codes.yaml

@@ -1,4 +1,9 @@
 IEEE Unapproved Draft Std PC37.110/D911Oct 07: IEEE Unapproved Draft Std PC37.110/D711, Oct 07
+# Issue #206: "IEEE Unapproved Draft Std P15289, 06" — the ", 06" suffix
+# has no clear meaning in IEEE's own grammar. Treating it as a draft
+# iteration number (D6) gives a parseable form. Conservative targeted
+# normalization rather than a general rule; revisit if more cases surface.
+IEEE Unapproved Draft Std P15289, 06: IEEE Unapproved Draft Std P15289/D6
 IEEE C57.139/D14June 2010: IEEE C57.139/D14, June 2010
 IEEE P1653.5/D7d1 November, 2019: IEEE P1653.5/D7.1 November, 2019
 IEEE P1017/D062012: P1017/D062012, Jun 2012
@@ -32,7 +37,13 @@ ISO/IEC/IEEE P26515 FDIS, May 2018: ISO/IEC/IEEE FDIS P26515, May 2018
 IEEE Std Std 1139-2008: IEEE Std 1139-2008
 IEC/IEEE P60214-2, FDIS 2018: IEC/IEEE FDIS P60214-2-2018(E)
 IEC/IEEE P62582-6, FDIS 2018: IEC/IEEE FDIS P62582-6-2018(E)
-ISO/IEC DIS P26511.2, May 2017: ISO/IEC DIS P26511-2, May 2017
+# ISO/IEC DIS P26511.2 (May 2017) is an IEEE-style citation of an ISO/IEC
+# DIS draft — ".2" is a draft version, not a part, so the prior mapping
+# to "ISO/IEC DIS P26511-2" produced a phantom document that doesn't exist
+# (ISO/IEC 26511 has no part 2). The ISO parser doesn't yet support the
+# IEEE-style "P" prefix or the ", Month Year" date suffix on ISO/IEC DIS
+# citations, so normalize to the bare DIS base + year. See pubid/pubid#212.
+ISO/IEC DIS P26511.2, May 2017: ISO/IEC DIS 26511.2:2017
 ISO/IEC/IEEE P90003_FDIS, August 2018: ISO/IEC/IEEE FDIS P90003, August 2018
 ISO/IEC/IEEE P29119-1/FDIS, September 2021: ISO/IEC/IEEE FDIS P29119-1, September 2021
 ISO/IEC/IEEE/ P12207-2/D3_FDIS, July 2020: ISO/IEC/IEEE FDIS P12207-2/D3, July 2020
@@ -102,3 +113,98 @@ IEEE Unapproved Draft Std P1635/ASHRAE 21_D4, Jan 2010: IEEE Unapproved Draft St
 IEC/IEEE 62582-1:2011 Edition 1.0 2011-08: IEC/IEEE 62582-1-2011 Edition 1.0 2011-08
 ISO/IEC/IEEE CD P26513/D1, August 2015: ISO/IEC/IEEE P26513/D1, August 2015
 /C37.0I0d-1984/: C37.010d-1984
+
+# --- month spellings pubid rejects -> canonical (relaton-ieee normtitles) ---
+# Surgical, case-sensitive: only forms the grammar genuinely REJECTS. Do NOT
+# add full-name->abbr rules (pubid accepts "June"/"September") or a blanket
+# underscore rule — a broader ruleset regressed 12 already-parsing normtitles.
+# Validated over relaton's 8,767 normtitles: +33 gained, 0 regressions.
+# 4-letter abbreviation
+/\bSept\b/: Sep
+# OCR typos seen in the IEEE rawbib feed
+/\bFeburary\b/: Feb
+/\bNovemer\b/: Nov
+/\bOctobor\b/: Oct
+/\bJanurary\b/: Jan
+# lowercase-only month words -> proper case (case-sensitive: leaves correct case alone)
+/\bjanuary\b/: January
+/\bfebruary\b/: February
+/\bmarch\b/: March
+/\bapril\b/: April
+/\bjune\b/: June
+/\bjuly\b/: July
+/\baugust\b/: August
+/\bseptember\b/: September
+/\boctober\b/: October
+/\bnovember\b/: November
+/\bdecember\b/: December
+
+# --- one-off IEEE normtitles migrated from relaton-ieee RawbibIdParser ---
+# Exact-match, anchored (plain keys compile to /^…$/), zero-regression: each
+# maps a specific malformed catalog string to relaton's canonical id, which
+# pubid parses losslessly. (hand-off: ieee-update-codes-oneoffs.) Group 1
+# parses on its own; the "dotted-draft" group relies on the verbatim-draft fix.
+"IEEE Std P1073.1.3.4/D3.0": "IEEE Std P11073.00101"
+"IEEE P802.1Qbu/03.0, July 2015": "IEEE P802.1Qbu/D-3.0-2015"
+"IEEE P802.11aqTM/013.0 October 2017": "IEEE P802.11aqTM/D-13.0-2017"
+"IEC/IEEE P62271-37-013:2015 D13.4": "IEC/IEEE P62271.37-013/D-13.4-2015"
+"ISO/IEC/IEEE DIS P42020:201x(E), June 2017": "ISO/IEC/IEEE DIS P42020-2017-06"
+"ISO/IEC/IEEE DIS P16326/201x, December 2018": "ISO/IEC/IEEE DIS P16326-2018-12"
+"ISO/IEC/IEEE/P21839, 2019(E)": "ISO/IEC/IEEE P21839-2019"
+"ISO/IEC/IEEE CD2 P12207-2: 201x(E), February 2019": "ISO/IEC/IEEE CD2 P12207.2-2019-02"
+"ISO/IEC/IEEE P42010.WD4:2019(E)": "ISO/IEC/IEEE P42010/D-4-2019"
+"IEC/IEEE P63195_CDV/V3, February 2020": "IEC/IEEE CDV P63195-2020-02"
+"ISO/IEC/IEEE/P16085_DIS, March 2020": "ISO/IEC/IEEE DIS P16085-2020-03"
+"ANSI/IEEE Std: Outdoor Apparatus Bushings": "ANSI/IEEE Std 21-1976-11"
+"Draft National Electrical Safety Code, January 2016": "IEEE PC2-2016-01"
+"International Standard ISO/IEC 8802-9: 1996(E) ANSI/IEEE Std 802.9, 1996 Edition": "ISO/IEC/IEEE Std 802.9-1996"
+"ISO/IEC13210: 1994 (E) ANSI/IEEE Std 1003.3-1991": "ISO/IEC/IEEE 13210-1994"
+"J-STD-016-1995": "IEEE 016-1995"
+"Nuclear EQ Sourcebook and Supplement": "IEEE 7438946"
+"PC37.30.2/D043 Rev 18, May 2015": "IEEE PC37.30-2/D-043/R-18-2015"
+"IEEE/IEC P62582 CD2 proposal, May 2017": "IEEE/IEC CD2 P62582-2017-05"
+"ISO/IEC/IEEE P16326:201x WD.4a, July 2017": "ISO/IEC/IEEE P16326/D-4a-2017-07"
+"ISO/IEC/IEEE CD.1 P21839, October 2017": "ISO/IEC/IEEE CD1 P21839-2017-10"
+"ISO/IEC/IEEE P16326:201x WD5, December 2017": "ISO/IEC/IEEE P16326/D-5-2017-12"
+"IEEE/ISO/IEC P42010.CD1-V1.0, April 2020": "IEEE/ISO/IEC CD1 P42010-2020-04"
+"Unapproved Draft Std ISO/IEC FDIS 15288:2007(E) IEEE P15288/D3,": "ISO/IEC/IEEE FDIS Std P15288/D-3-2007"
+"IS0/IEC/IEEE 8802-11:2012/Amd.5:2015(E) (Adoption of IEEE Std 802.11af-2014)": "ISO/IEC/IEEE 802.11/Amd5-2012"
+# `/D-` corrupted IEEE draft docids — malformed draft designations emitted by
+# relaton-ieee's RawbibIdParser fallback rendering (the `/D-`, `_CDV_`, `.pdf`,
+# `201x`, stray dots/spaces). Each maps to a canonical parseable form that
+# retains number/part/draft/stage/date in the hash (the relaton index-v2 gate
+# is the hash round-trip + non-empty root.number, not `to_s == input`).
+# Interim measure; the durable fix is upstream in RawbibIdParser. See
+# HANDOFFS/metanorma__pubid__ieee-update-codes-dashD-drafts.md.
+# --- 8 pre-verified ---
+"IEEE Approved Std P1609.3/D-23.pdf-2007-02": "IEEE Approved Std P1609.3/D23, 02 2007"
+"IEEE P3004.5/D-6.2.": "IEEE P3004.5/D6.2"
+"IEEE P730/D-0.16.-2025": "IEEE P730/D0.16-2025"
+"IEEE PC37.107/D-4.-2023-11": "IEEE PC37.107/D4, 11 2023"
+"IEEE PC37.113/D-8.0.-2015-08": "IEEE PC37.113/D8.0, 08 2015"
+"IEEE PC37.98/D-5.1D-2013": "IEEE PC37.98/D5.1D-2013"
+"IEEE PC57.637/D-6.2.-2014-03": "IEEE PC57.637/D6.2, 03 2014"
+"IEEE Std P650/D-9.-2006": "IEEE Std P650/D9-2006"
+# --- 21 derived (revision markers dropped, corrigendum reordered, joint
+#     stage+draft repositions year onto number with stage moved front) ---
+"IEC/IEEE P63198.2775/D-6.1_CDV_March-2022": "IEC/IEEE CDV P63198.2775/D6.1, March 2022"
+"IEEE 3.Rev.B/D-4.0-2023-01": "IEEE P3/D4.0, 01 2023"
+"IEEE 3.Rev.B/D-5.0-2023-03": "IEEE P3/D5.0, 03 2023"
+"IEEE P1003.1/D-5/Cor1-201x-2012-04": "IEEE P1003.1/Cor 1/D5, 04 2012"
+"IEEE P1003.1/D-5/Cor1-201x-2012-08": "IEEE P1003.1/Cor 1/D5, 08 2012"
+"IEEE P1127/D-4.1_July-2023": "IEEE P1127/D4.1, July 2023"
+"IEEE P2660_1/D-2-2020-06": "IEEE P2660.1/D2, 06 2020"
+"IEEE P802.1AE/D-1.0/R-/Cor1": "IEEE P802.1AE/Cor 1/D1.0"
+"IEEE P802.1AE_Rev/D-1.2-2018": "IEEE P802.1AE/D1.2-2018"
+"IEEE P802.1AE_Rev/D-1.3-2018": "IEEE P802.1AE/D1.3-2018"
+"IEEE Unapproved Std P802.16Rev2/D-9-2009-01": "IEEE Unapproved Std P802.16Rev2/D9-2009"
+"IEEE Unapproved Std P802.1X_REV/D-4.5-2009": "IEEE Unapproved Std P802.1X/D4.5-2009"
+"IEEE Unapproved Std PC37.91/D-8-2007-20": "IEEE Unapproved Std PC37.91/D8-2007"
+"IEEEE P1243/D-3-2023-02": "IEEE P1243/D3, 02 2023"
+"ISO/IEC/IEEE FDIS P15289./E-3/D-2-2016": "ISO/IEC/IEEE FDIS P15289-2016/D2"
+"ISO/IEC/IEEE P24641/D-2_CD-2020": "ISO/IEC/IEEE CD P24641-2020/D2"
+"ISO/IEC/IEEE P24641/D-3_CD2": "ISO/IEC/IEEE CD2 P24641/D3"
+"ISO/IEC/IEEE P24748.3/D-3_FDIS-2020": "ISO/IEC/IEEE FDIS P24748.3-2020/D3"
+"ISO/IEC/IEEE/ P12207.2/D-3_FDIS-2020": "ISO/IEC/IEEE FDIS P12207.2-2020/D3"
+"ISO/IEC/IEEE/ P24748.3/D-IS-2019": "ISO/IEC/IEEE DIS P24748.3-2019"
+"P1635/D10/ASHARE 21/D-10-2016": "IEEE P1635/D10-2016"

data/{archived-gems/pubid-nist → data/nist}/series.yaml

@@ -69,6 +69,40 @@ abbrev:
   CSRC Building Block: CSRC Building Block
   JPCRD: J. Phys. & Chem. Ref. Data
   JRES: J. Res. Natl. Inst. Stan.
+  # Issue #150: abbreviated descriptions for series that previously had
+  # none. Convention follows existing entries (shortened, period-suffixed).
+  # Ronald Tse should review and refine as needed.
+  AI: AI
+  BH: Bldg. & Hous.
+  BMS: Bldg. Mat. Struct.
+  CIRC: Circ.
+  CRPL: CRPL
+  CS: Commer. Stand.
+  CSM: Commer. Stand. Mo.
+  EAB: Econ. Anal. Br.
+  GCR: Grant/Contract Rpt.
+  IR: Interag. Rep.
+  ITL Bulletin: ITL Bull.
+  MP: Misc. Publ.
+  NBS BRPD-CRPL-D: Bas. Radio Prop. Pred.
+  NBS CIS: Consum. Inf. Ser.
+  NBS CRPL-F-A: CRPL Ionos. Data
+  NBS CRPL-F-B: CRPL Solar-Geophys. Data
+  NBS CS-E: Commer. Stand. (Emerg.)
+  NBS HR: Hydraul. Res.
+  NBS IP: CRPL Ionos. Pred.
+  NBS IRPL: Interserv. Radio Prop. Lab.
+  NBS LC: Lett. Circ.
+  NIST LC: Lett. Circ.
+  OWMWP: Off. Weights Meas. WP
+  PC: Photo. Lab. Circ.
+  RPT: Rpt.
+  SIBS: Spec. Inter. Ball. Stud.
+  TIBM: Tech. Inf. Bldg. Mat.
+  TTB: Tech. Transf. Br.
+  VTS: Voting Tech. Ser.
+  JPCRD: J. Phys. & Chem. Ref. Data
+  JRES: J. Res. Natl. Inst. Stan.
 mr:
   AMS: AMS
   BSS: BSS

data/data/nist/update_codes.yaml

@@ -25,8 +25,16 @@
 # by earlier ones. Run `bundle exec rspec spec/pubid/nist/` after edits.
 NBS MN: NBS MONO
 NIST MN: NIST MONO
+# Long-form series names — normalize to the abbreviated form the parser expects.
+/\bMonograph\b/: MONO
+# Citation-form comma before long parameter prefixes — strip so the parser's
+# " Rev. " / " Part. " / " Vol. " prefix rules match (issue #197).
+'/,\s+(Rev\.|Part\.|Vol\.|Revision|Volume|Part)(?=\s|\d)/': ' \1'
 NIST MP: NBS MP
 NIST SP 260-162 2006ed.: NIST SP 260-162e2006
+# Doubled publisher+series prefix — an error in the NIST Tech Pubs source data
+# (pubid/pubid#152). Collapse the repeated "NIST AI" so the remainder parses.
+NIST AI NIST AI 100-2e2023 ipd: NIST AI 100-2e2023 ipd
 NBS CIRC 154suprev: NBS CIRC 154r1sup
 NBS CIRC supJun1925-Jun1926: NBS CIRC 24e7sup2
 NBS CIRC supJun1925-Jun1927: NBS CIRC 24e7sup3

data/lib/pubid/adobe/builder.rb

@@ -0,0 +1,52 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Adobe
+    # Builds an Adobe identifier object from the Parslet parse tree.
+    class Builder
+      def build(parsed)
+        return build_tech_note(parsed[:tech_note]) if parsed[:tech_note]
+        return build_tech_note_bare(parsed[:tech_note_bare]) if parsed[:tech_note_bare]
+        return build_publication_version(parsed[:publication_version]) if parsed[:publication_version]
+        return build_publication(parsed[:slug]) if parsed[:slug]
+
+        raise ArgumentError, "Unrecognized Adobe parse tree: #{parsed.inspect}"
+      end
+
+      def self.build(parsed)
+        new.build(parsed)
+      end
+
+      private
+
+      def build_tech_note(hash)
+        Identifiers::TechNote.new(
+          number: stringify(hash[:number]),
+          slug:   stringify(hash[:slug]),
+        )
+      end
+
+      def build_tech_note_bare(number_hash)
+        Identifiers::TechNote.new(number: stringify(number_hash))
+      end
+
+      def build_publication(slug_value)
+        Identifiers::Publication.new(slug: stringify(slug_value))
+      end
+
+      def build_publication_version(hash)
+        Identifiers::Publication.new(
+          slug:    stringify(hash[:slug]),
+          version: stringify(hash[:version]),
+        )
+      end
+
+      def stringify(value)
+        return nil if value.nil?
+
+        str = value.to_s
+        str.empty? ? nil : str
+      end
+    end
+  end
+end

data/lib/pubid/adobe/identifier.rb

@@ -0,0 +1,47 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Adobe
+    # Base class for all Adobe identifiers. Canonical name
+    # Pubid::Adobe::Identifier; every concrete Adobe identifier
+    # (Identifiers::*) descends from it.
+    #
+    # Adobe identifiers come in two shapes:
+    #
+    #   * TechNote    — `Adobe Technical Note #<number>` / `ATN<number>`
+    #                   e.g. `Adobe Technical Note #5014`, `ATN5014`.
+    #   * Publication — slug-keyed named specs that have no number
+    #                   e.g. `adobe-glyph-list`, `adobe-japan1-7`.
+    class Identifier < ::Pubid::Identifier
+      # Parse an Adobe identifier string into an identifier object.
+      # @param identifier [String]
+      # @return [Pubid::Adobe::Identifier]
+      # @raise [Pubid::Errors::ParseError] If parsing fails
+      def self.parse(identifier)
+        parsed = Parser.parse(identifier)
+        Builder.build(parsed)
+      end
+
+      attribute :publisher, :string, default: "Adobe"
+
+      ADOBE_TYPE_MAP = {
+        "pubid:adobe:tech-note"   => "Pubid::Adobe::Identifiers::TechNote",
+        "pubid:adobe:publication" => "Pubid::Adobe::Identifiers::Publication",
+      }.freeze
+
+      key_value do
+        map "_type", to: :_type, polymorphic_map: ADOBE_TYPE_MAP
+      end
+
+      def to_urn
+        UrnGenerator.new(self).generate
+      end
+
+      # Short type code (e.g. "ATN", nil for Publication) — convenience
+      # accessor used by Renderer and UrnGenerator.
+      def type_short
+        self.class.type[:short]
+      end
+    end
+  end
+end

data/lib/pubid/adobe/identifiers/publication.rb

@@ -0,0 +1,31 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Adobe
+    module Identifiers
+      # Adobe named publication. Slug-keyed specs that have no tech-note
+      # number — Adobe Glyph List, PostScript Language, character
+      # collections, etc.
+      #
+      # Surface forms:
+      #   * slug `adobe-glyph-list`, `adobe-postscript-language`
+      #   * versioned character collections: `adobe-japan1` + version `7`
+      #   * human render is the slug title-cased (`Adobe Glyph List`,
+      #     `Adobe-Japan1-7`); the full publication title lives in the
+      #     data repo's title attribute, not here.
+      class Publication < Identifier
+        attribute :slug,    :string   # "adobe-glyph-list", "adobe-japan1"
+        attribute :version, :string   # "7", "3", nil
+
+        key_value do
+          map "slug",    to: :slug
+          map "version", to: :version
+        end
+
+        def self.type
+          { key: :publication, title: "Publication", short: nil }
+        end
+      end
+    end
+  end
+end

data/lib/pubid/adobe/identifiers/tech_note.rb

@@ -0,0 +1,32 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Adobe
+    module Identifiers
+      # Adobe Technical Note. Numbered 4-digit (5xxx series observed).
+      #
+      # Surface forms:
+      #   * `Adobe Technical Note #5014` — formal citation form
+      #   * `ATN5014`                    — short alias used in Metanorma
+      #                                    anchor ids ([[atn5014,…]])
+      #   * filename `5014.CIDFont_Spec.pdf` (data repo splits filename
+      #     into number + slug before calling the parser)
+      #
+      # The optional `slug` carries the filename's title segment
+      # (`CIDFont_Spec`) so round-trip through the data repo is lossless.
+      class TechNote < Identifier
+        attribute :number, :string   # "5014", "5902"
+        attribute :slug,   :string   # "CIDFont_Spec", "AFM_Spec", nil
+
+        key_value do
+          map "number", to: :number
+          map "slug",   to: :slug
+        end
+
+        def self.type
+          { key: :"tech-note", title: "Technical Note", short: "ATN" }
+        end
+      end
+    end
+  end
+end

data/lib/pubid/adobe/identifiers.rb

@@ -0,0 +1,10 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Adobe
+    module Identifiers
+      autoload :TechNote,    "#{__dir__}/identifiers/tech_note"
+      autoload :Publication, "#{__dir__}/identifiers/publication"
+    end
+  end
+end

data/lib/pubid/adobe/parser.rb

@@ -0,0 +1,136 @@
+# frozen_string_literal: true
+
+require "parslet"
+
+module Pubid
+  module Adobe
+    # Parslet grammar for Adobe identifiers.
+    #
+    # Canonical citation forms (what the renderer produces):
+    #
+    #   Adobe TN 5014                  (tech note; replaces the older
+    #                                    "Adobe Technical Note #5014")
+    #   Adobe Publication adobe-glyph-list
+    #   Adobe Publication adobe-japan1-7
+    #
+    # The parser also accepts these legacy forms for back-compat:
+    #
+    #   Adobe Technical Note #5014
+    #   Adobe Technical Note 5014
+    #   Adobe TN5014                   (no space)
+    #   ATN5014                        (short alias)
+    #   5014                           (bare number; data-repo context)
+    #   adobe-glyph-list               (bare slug)
+    class Parser < ::Pubid::Parser::Grammar
+      include ::Pubid::Parser::CommonParseRules
+
+      root :identifier
+
+      rule(:space)  { str(" ") }
+      rule(:space?) { space.maybe }
+      rule(:hash)   { str("#") }
+      rule(:dot)    { str(".") }
+      rule(:dash)   { str("-") }
+
+      # Canonical short prefix: "Adobe TN" with optional trailing space.
+      # Matches both "Adobe TN 5014" and "Adobe TN5014".
+      rule(:tech_note_tn_prefix) do
+        str("Adobe") >> space >> str("TN") >> space?.maybe
+      end
+
+      # Legacy long prefix: "Adobe Technical Note" with optional "#".
+      rule(:tech_note_long_prefix) do
+        str("Adobe") >> space >> str("Technical") >> space >> str("Note") >>
+          (space? >> hash.maybe >> space?).maybe
+      end
+
+      # Legacy short alias: "ATN" with optional trailing space.
+      rule(:tech_note_atn_prefix) do
+        str("ATN") >> space?.maybe
+      end
+
+      rule(:tech_note_prefix) do
+        tech_note_tn_prefix | tech_note_long_prefix | tech_note_atn_prefix
+      end
+
+      rule(:digits) do
+        match("[0-9]").repeat(1)
+      end
+
+      # 4-digit tech note number, optionally followed by ".<Slug>" from
+      # the filename form. The slug is captured but not required.
+      rule(:slug_chars) do
+        match("[A-Za-z0-9_]").repeat(1)
+      end
+
+      rule(:tech_note) do
+        (tech_note_prefix >>
+          digits.as(:number) >>
+          (dot >> slug_chars.as(:slug)).maybe).as(:tech_note)
+      end
+
+      # Bare number form (e.g. when the data repo passes just "5014"
+      # extracted from a filename). Not advertised in user docs; used
+      # internally by AdobeFetcher::Docid.
+      rule(:bare_number) do
+        digits.as(:number).as(:tech_note_bare)
+      end
+
+      # "Adobe Publication " prefix. Matched literally then discarded —
+      # the publication slug/version is parsed by the inner rule.
+      rule(:publication_prefix) do
+        str("Adobe") >> space >> str("Publication") >> space
+      end
+
+      # Publication slug. Lowercase ASCII letters and digits joined by `-`.
+      # Each segment starts with a letter (so a trailing `-<digits>` is
+      # unambiguously a version suffix, not part of the slug).
+      rule(:slug_segment) do
+        match("[a-z]") >> match("[a-z0-9]").repeat
+      end
+
+      rule(:slug_inner) do
+        slug_segment >> (dash >> slug_segment).repeat
+      end
+
+      rule(:publication_with_version) do
+        (slug_inner.as(:slug) >> dash >> digits.as(:version)).as(:publication_version)
+      end
+
+      rule(:publication_slug_or_version) do
+        # Try with-version first so `adobe-japan1-7` doesn't get captured
+        # as slug `adobe-japan1-7` with no version. The flat `:slug` form
+        # is the fallback for unversioned slugs.
+        publication_with_version | slug_inner.as(:slug)
+      end
+
+      rule(:publication) do
+        # Optional "Adobe Publication " prefix. The prefix is consumed
+        # but not captured — the inner rule produces :slug or
+        # :publication_version as before.
+        publication_prefix.maybe >> publication_slug_or_version
+      end
+
+      rule(:identifier) do
+        tech_note | bare_number | publication
+      end
+
+      # Parse a string and return the raw parslet tree.
+      # @param string [String]
+      # @return [Hash]
+      def self.parse(string)
+        unless string.is_a?(String)
+          raise ::Pubid::Errors::InvalidInputError,
+                ::Pubid::INPUT_NOT_A_STRING_MESSAGE
+        end
+
+        if string.length > ::Pubid::MAX_INPUT_LENGTH
+          raise ::Pubid::Errors::InvalidInputError,
+                ::Pubid::INPUT_TOO_LONG_MESSAGE
+        end
+
+        new.parse(string.strip)
+      end
+    end
+  end
+end

data/lib/pubid/adobe/renderer.rb

@@ -0,0 +1,39 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Adobe
+    # Human-readable renderer for Adobe identifiers.
+    #
+    # Produces the canonical citation forms:
+    #   "Adobe TN 5014"                       (tech note)
+    #   "Adobe Publication adobe-glyph-list"  (publication, slug)
+    #   "Adobe Publication adobe-japan1-7"    (publication, slug + version)
+    #
+    # The data repo (AdobeFetcher::Docid) overrides #to_s for publications
+    # to use the title from metadata instead of the slug, producing e.g.
+    # "Adobe Publication Adobe Glyph List". Pubid::Adobe itself is purely
+    # structural and renders slug-form citations.
+    class Renderer < ::Pubid::Renderers::Base
+      def render(context: nil, **_opts)
+        id = @id
+
+        return render_tech_note(id) if id.is_a?(Identifiers::TechNote)
+        return render_publication(id) if id.is_a?(Identifiers::Publication)
+
+        raise ArgumentError, "Unknown Adobe identifier class: #{id.class}"
+      end
+
+      private
+
+      def render_tech_note(id)
+        "Adobe TN #{id.number}"
+      end
+
+      def render_publication(id)
+        rendered = "Adobe Publication #{id.slug}"
+        rendered << "-#{id.version}" if id.version
+        rendered
+      end
+    end
+  end
+end