pubid 2.0.0.pre.alpha.6 → 2.0.0.pre.alpha.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 1b1fa5f38da07bc9a507c0c507ba9ddfc8f6541ab9f4b6793b09abe302ab8849
-  data.tar.gz: 63b7eba9c9942c177e95b9c78a09462679a11acde4fd6a2c2e486848fb43ca93
+  metadata.gz: 5987bc5bdc86f9fdf02244f7b5e1abbff308a0d7a5ae88c9ea6e26d20f316ec4
+  data.tar.gz: d768df87e4d1b06d0be0dbcc72dab8d5e5eda5476b4157bbf754c5863a2ad1f0
 SHA512:
-  metadata.gz: 915034be3882a43991ce3c0f934c4097f216f1fdac35dcb59ed6a0decb639dd4a19882de4f28940d4f29e6e9bb351de3383b07bd95cc0dc2a53bb4ad875f072d
-  data.tar.gz: 1cb7e935accc6e66bed8b599d25605b26cd415547c1ddec9ba2ecf1b7183883102c99078b830b2de9da436bc17613db042b5fc05e0d7d5515786fae71592aeb8
+  metadata.gz: 7214ebd2c2eac905def41dd1b4304ef299224378e703ab3d9883022b8fd2770adafd08c8e697d759a9ae6e2a984ff54927051090039c8b20cd56f98ae21bc3b8
+  data.tar.gz: 32227b91be942f7f59c758685ea7cbe105845eaa758c29984d43f81adeaf33321e68f6b4f7a8d9d71709f3395acf7011d008e1a2983ae43a61610b267c785d0e

data/README.adoc

@@ -1,54 +1,15 @@
 = PubID: Interoperable identifiers for information resources
 
-image:https://img.shields.io/gem/v/pubid-core.svg["pubid-core Gem Version", link="https://rubygems.org/gems/pubid-core"]
+image:https://img.shields.io/gem/v/pubid.svg["pubid Gem Version", link="https://rubygems.org/gems/pubid"]
 image:https://github.com/metanorma/pubid/actions/workflows/rake.yml/badge.svg["Build Status", link="https://github.com/metanorma/pubid/actions/workflows/rake.yml"]
 image:https://img.shields.io/github/issues-pr-raw/metanorma/pubid.svg["Pull Requests", link="https://github.com/metanorma/pubid/pulls"]
 image:https://img.shields.io/github/commits-since/metanorma/pubid/latest.svg["Commits since latest",link="https://github.com/metanorma/pubid/releases"]
 
-== Badges
-
-[cols="3*", options="header"]
-|===
-| Core & Meta | Standards Organizations | Regional Standards
-
-| pubid image:https://img.shields.io/gem/v/pubid.svg["pubid", link="https://rubygems.org/gems/pubid"]
-| pubid-iso image:https://img.shields.io/gem/v/pubid-iso.svg["pubid-iso", link="https://rubygems.org/gems/pubid-iso"]
-| pubid-cen image:https://img.shields.io/gem/v/pubid-cen.svg["pubid-cen", link="https://rubygems.org/gems/pubid-cen"]
-
-| pubid-core image:https://img.shields.io/gem/v/pubid-core.svg["pubid-core", link="https://rubygems.org/gems/pubid-core"]
-| pubid-iec image:https://img.shields.io/gem/v/pubid-iec.svg["pubid-iec", link="https://rubygems.org/gems/pubid-iec"]
-| pubid-bsi image:https://img.shields.io/gem/v/pubid-bsi.svg["pubid-bsi", link="https://rubygems.org/gems/pubid-bsi"]
-
-|
-| pubid-ieee image:https://img.shields.io/gem/v/pubid-ieee.svg["pubid-ieee", link="https://rubygems.org/gems/pubid-ieee"]
-| pubid-jis image:https://img.shields.io/gem/v/pubid-jis.svg["pubid-jis", link="https://rubygems.org/gems/pubid-jis"]
-
-|
-| pubid-nist image:https://img.shields.io/gem/v/pubid-nist.svg["pubid-nist", link="https://rubygems.org/gems/pubid-nist"]
-|
-
-|
-| pubid-etsi image:https://img.shields.io/gem/v/pubid-etsi.svg["pubid-etsi", link="https://rubygems.org/gems/pubid-etsi"]
-|
-
-|
-| pubid-itu image:https://img.shields.io/gem/v/pubid-itu.svg["pubid-itu", link="https://rubygems.org/gems/pubid-itu"]
-|
-
-|
-| pubid-ccsds image:https://img.shields.io/gem/v/pubid-ccsds.svg["pubid-ccsds", link="https://rubygems.org/gems/pubid-ccsds"]
-|
-
-|
-| pubid-plateau image:https://img.shields.io/gem/v/pubid-plateau.svg["pubid-plateau", link="https://rubygems.org/gems/pubid-plateau"]
-|
-|===
-
 == Quick Start
 
 [source,ruby]
 ----
-gem install pubid-core
+gem install pubid
 
 require 'pubid/iso'
 
@@ -292,6 +253,46 @@ base.root.equal?(base) # => true
 * ✅ **Works with supplements** - Traverses supplement chains correctly
 
 
+== Flavor Prefix Routing
+
+Every flavor exposes a static `prefixes` method returning the *leading
+identifier prefix tokens* it recognizes — the tokens a printed identifier can
+start with such that the string belongs to that SDO. This is the routing key a
+consumer (such as relaton) uses to dispatch a reference string to the right
+backend, without parsing.
+
+[source,ruby]
+----
+require 'pubid'
+
+Pubid::Iso.prefixes
+# => ["ISO", "ISO/IEC", "IEC/ISO", "ISO/IEC/IEEE"]
+
+Pubid::Bsi.prefixes
+# => ["BS", "BSI", "DD", "PD", "PAS", "NA", "TS", "HB", ... , "BS EN ISO"]
+
+# Look up a single flavor by name
+Pubid.prefixes(:iec)
+# => ["IEC", "CISPR", "IECEE", "IECEx", "IECQ", "ISO/IEC", "IEC/ISO", "ISO/IEC/IEEE"]
+
+# Reverse index: prefix token => owning flavor(s)
+Pubid.prefix_flavors["ISO/IEC"]       # => [:iec, :iso]
+Pubid.prefix_flavors["ISO/IEC/IEEE"]  # => [:iec, :ieee, :iso]
+Pubid.prefix_flavors["DD"]            # => [:bsi]
+----
+
+**Semantics** (see `Pubid::PrefixesSupport` for the full contract):
+
+* **Static** — no identifier string is parsed; callable at startup.
+* **Uniform** — every registered flavor returns a plain `Array<String>`.
+* **Symmetric joints** — a co-published prefix is listed by every co-publisher
+  (`"ISO/IEC"` is in both `Pubid::Iso.prefixes` and `Pubid::Iec.prefixes`).
+* **Inclusion** — publisher prefixes (`ISO`, `NIST`, `BS`…) and leading
+  series/type tokens that can *begin* a reference (BSI `DD`/`PD`, NIST `FIPS`).
+* **Exclusion** — pure sub-series that never lead a reference and
+  dangerously-ambiguous single letters (e.g. BSI aerospace `A`/`B`/`C`).
+
+
 == Advanced Rendering Styles
 
 ISO and IEC identifiers support multiple abbreviation forms for supplements to maintain round-trip fidelity with different official formats.
@@ -638,15 +639,13 @@ lib/tasks/export.rake                # Rake tasks: export:website_data, export:a
 
 == Repository
 
-This repository is a monorepo for the PubID Ruby gems, which implement the PubID
+This repository provides the single `pubid` Ruby gem, which implements the PubID
 identifier data model and its various components.
 
-The PubID Ruby gems implement identifiers from multiple standards organizations,
-making it easier for developers to work with them in their applications.
-
-This repository contains all the `pubid-*` gems consolidated into a single
-monorepo for easier development and maintenance while preserving individual gem
-releases.
+The `pubid` gem implements identifiers from multiple standards organizations,
+making it easier for developers to work with them in their applications. All 22+
+supported flavors (ISO, IEC, NIST, IEEE, …) live under `lib/pubid/{flavor}/` and
+ship together in the one gem.
 
 == Structure
 
@@ -671,11 +670,11 @@ The repository is organized as follows:
 ├── spec/                   # Tests
 │   ├── pubid/              # Per-flavor tests
 │   ├── fixtures/           # Bulk test fixtures
-│   └── integration/        # Cross-gem tests
+│   └── integration/        # Cross-flavor integration tests
 ├── docs/                   # Documentation
 ├── .github/workflows/      # CI/CD workflows
 ├── Gemfile                 # Root Gemfile for development
-├── Rakefile               # Monorepo management tasks
+├── Rakefile               # Test, lint, build, and release tasks
 ├── .rubocop.yml           # Shared RuboCop configuration
 └── LICENSE.txt            # Shared license
 ----

data/lib/pubid/amca.rb

@@ -2,6 +2,12 @@
 
 module Pubid
   module Amca
+    extend Pubid::PrefixesSupport
+
+    # AMCA publisher token; the ANSI/AMCA joint form comes from
+    # Pubid::JOINT_PREFIXES.
+    PREFIXES = ["AMCA"].freeze
+
     autoload :Identifier, "#{__dir__}/amca/identifier"
     autoload :Identifiers, "#{__dir__}/amca/identifiers"
     autoload :Builder, "#{__dir__}/amca/builder"

data/lib/pubid/ansi.rb

@@ -2,6 +2,13 @@
 
 module Pubid
   module Ansi
+    extend Pubid::PrefixesSupport
+
+    # ANSI publisher token; ANSI/ASHRAE and ANSI/AMCA joint forms come from
+    # Pubid::JOINT_PREFIXES. Parser::ORGANIZATIONS are co-publishers, not
+    # leading ANSI tokens, so they are excluded.
+    PREFIXES = ["ANSI"].freeze
+
     autoload :Builder, "#{__dir__}/ansi/builder"
     autoload :Identifier, "#{__dir__}/ansi/identifier"
     autoload :Identifiers, "#{__dir__}/ansi/identifiers"

data/lib/pubid/api.rb

@@ -2,6 +2,11 @@
 
 module Pubid
   module Api
+    extend Pubid::PrefixesSupport
+
+    # Sole API publisher token (see the parser's `publisher` rule).
+    PREFIXES = ["API"].freeze
+
     autoload :Builder, "#{__dir__}/api/builder"
     autoload :Identifier, "#{__dir__}/api/identifier"
     autoload :Identifiers, "#{__dir__}/api/identifiers"

data/lib/pubid/ashrae.rb

@@ -2,6 +2,12 @@
 
 module Pubid
   module Ashrae
+    extend Pubid::PrefixesSupport
+
+    # ASHRAE publisher token; the ANSI/ASHRAE joint form comes from
+    # Pubid::JOINT_PREFIXES.
+    PREFIXES = ["ASHRAE"].freeze
+
     autoload :Builder, "#{__dir__}/ashrae/builder"
     autoload :Identifier, "#{__dir__}/ashrae/identifier"
     autoload :Identifiers, "#{__dir__}/ashrae/identifiers"

data/lib/pubid/asme.rb

@@ -2,6 +2,11 @@
 
 module Pubid
   module Asme
+    extend Pubid::PrefixesSupport
+
+    # Sole ASME publisher token (see the parser's `asme_publisher` rule).
+    PREFIXES = ["ASME"].freeze
+
     autoload :Builder, "#{__dir__}/asme/builder"
     autoload :Components, "#{__dir__}/asme/components/code"
     autoload :Identifier, "#{__dir__}/asme/identifier"

data/lib/pubid/astm.rb

@@ -2,6 +2,11 @@
 
 module Pubid
   module Astm
+    extend Pubid::PrefixesSupport
+
+    # Sole ASTM publisher token (see the parser's `publisher` rule).
+    PREFIXES = ["ASTM"].freeze
+
     autoload :Components, "#{__dir__}/astm/components"
     autoload :Identifier, "#{__dir__}/astm/identifier"
     autoload :Identifiers, "#{__dir__}/astm/identifiers"

data/lib/pubid/bsi.rb

@@ -2,6 +2,21 @@
 
 module Pubid
   module Bsi
+    extend Pubid::PrefixesSupport
+
+    # Leading tokens a printed BSI reference can start with, curated from the
+    # parser grammar (lib/pubid/bsi/parser.rb): the publisher (`bs` => BS/BSI),
+    # the native series/type rules that begin references (dd, pd, pas, na, ts,
+    # handbook => HB, pp, bip, ep_type), and the "BS <adopted-org>" combos the
+    # grammar builds from `bs` + `adopted_org_prefix`. Deliberately EXCLUDES the
+    # aerospace single/multi-letter prefixes (see
+    # identifiers/aerospace_standard.rb SINGLE_LETTER_PREFIXES / the parser's
+    # `single_letter_prefix` rule — A/B/C/S/X ...) because they are dangerously
+    # ambiguous and never lead a routable reference on their own; and the bare
+    # adopted forms (ISO/IEC/EN alone), which route to their own SDOs.
+    PREFIXES = (%w[BS BSI DD PD PAS NA TS HB PP BIP EP] +
+      ["BS EN", "BS ISO", "BS IEC", "BS EN ISO", "BS EN IEC"]).freeze
+
     module Components
       autoload :Code, "#{__dir__}/bsi/components/code"
       autoload :Date, "#{__dir__}/bsi/components/date"

data/lib/pubid/ccsds.rb

@@ -2,6 +2,11 @@
 
 module Pubid
   module Ccsds
+    extend Pubid::PrefixesSupport
+
+    # Sole CCSDS publisher token (see the parser's `ccsds_prefix` rule).
+    PREFIXES = ["CCSDS"].freeze
+
     autoload :Builder, "#{__dir__}/ccsds/builder"
     autoload :Identifier, "#{__dir__}/ccsds/identifier"
     autoload :Identifiers, "#{__dir__}/ccsds/identifiers"

data/lib/pubid/cen_cenelec.rb

@@ -2,6 +2,13 @@
 
 module Pubid
   module CenCenelec
+    extend Pubid::PrefixesSupport
+
+    # CEN/CENELEC publisher tokens, mirroring the parser's publisher rules
+    # (lib/pubid/cen_cenelec/parser.rb). Adopted-org prefixes (ISO/IEC/CISPR)
+    # are excluded — they route to their own SDOs, not to CEN.
+    PREFIXES = %w[EN CEN CLC CWA HD ES CR ENV].freeze
+
     autoload :Builder, "#{__dir__}/cen_cenelec/builder"
     autoload :Identifier, "#{__dir__}/cen_cenelec/identifier"
     autoload :Identifiers, "#{__dir__}/cen_cenelec/identifiers"

data/lib/pubid/cie.rb

@@ -2,6 +2,11 @@
 
 module Pubid
   module Cie
+    extend Pubid::PrefixesSupport
+
+    # Sole CIE publisher token (see the parser grammar).
+    PREFIXES = ["CIE"].freeze
+
     module Components
       autoload :Code, "#{__dir__}/cie/components/code"
       autoload :Language, "#{__dir__}/cie/components/language"

data/lib/pubid/csa.rb

@@ -2,6 +2,11 @@
 
 module Pubid
   module Csa
+    extend Pubid::PrefixesSupport
+
+    # Sole CSA publisher token (see the parser's `publisher` rule).
+    PREFIXES = ["CSA"].freeze
+
     autoload :Builder, "#{__dir__}/csa/builder"
     autoload :Identifier, "#{__dir__}/csa/identifier"
     autoload :Identifiers, "#{__dir__}/csa/identifiers"

data/lib/pubid/etsi.rb

@@ -4,6 +4,13 @@ require "lutaml/model"
 
 module Pubid
   module Etsi
+    extend Pubid::PrefixesSupport
+
+    # Sole ETSI publisher token (see the parser's `etsi_prefix` rule). The
+    # document-type tokens (EN, TS, TR, ...) are sub-types that follow the ETSI
+    # prefix, not standalone routing tokens, so they are excluded.
+    PREFIXES = ["ETSI"].freeze
+
     autoload :Builder, "#{__dir__}/etsi/builder"
     autoload :Components, "#{__dir__}/etsi/components"
     autoload :Identifier, "#{__dir__}/etsi/identifier"

data/lib/pubid/iala/builder.rb

@@ -6,10 +6,13 @@ module Pubid
     class Builder
       def build(parsed)
         hash = parsed
+
+        return build_annex(hash) if hash[:annex_marker]
+
         type_letter = stringify(hash[:type_letter])&.upcase
 
         attrs = {
-          number:   build_number(hash),
+          number:   build_number(hash, type_letter),
           edition:  stringify(hash.dig(:edition, :edition_value)),
           language: stringify(hash.dig(:language_group, :language))&.upcase,
         }.compact
@@ -23,13 +26,66 @@ module Pubid
 
       private
 
-      # Compose "1070" or "0126-1" from doc_number + optional subpart.
-      def build_number(hash)
+      # Build the Annex wrapper from the parse tree. The base sub-tree
+      # (captured under :base) is recursively built via the same logic as
+      # a flat identifier, then wrapped.
+      def build_annex(hash)
+        base_hash = hash[:base]
+        base_type = stringify(base_hash[:type_letter])&.upcase
+        base_number = build_number(base_hash, base_type)
+        base = Iala.identifier_klass_for_type_letter(base_type).new(number: base_number)
+
+        attrs = {
+          base_identifier: base,
+          annex_form:      stringify(hash[:annex_marker]),
+          letter:          stringify(hash[:annex_letter]),
+          edition:         stringify(hash.dig(:edition, :edition_value)),
+          language:        stringify(hash.dig(:language_group, :language))&.upcase,
+        }.compact
+
+        Identifiers::Annex.new(**attrs)
+      end
+
+      # Compose the full number string from the leading doc_number plus any
+      # dotted continuation (GA01.13 → "01.13", L2.1.11 → "2.1.11") plus
+      # the optional dash-separated subpart (C0103-1 → "0103-1", L2.7.1-2
+      # → "2.7.1-2"). Padding is type-aware per IALA's canonical form:
+      # typed S/R/G/M/C zero-pad to 4 digits; GeneralAssembly (GA) zero-
+      # pads each segment to 2; Advices (A) and Letters (L) preserve input
+      # (their canonical form is already what the IALA site prints).
+      def build_number(hash, type_letter = nil)
         base = stringify(hash[:doc_number])
-        return base unless hash[:subpart]
+        return base unless base
+
+        base = canonical_base(base, type_letter)
+
+        dots = hash[:doc_number_dots]
+        dots_str = dots.is_a?(Array) ? "" : canonical_dots(stringify(dots).to_s, type_letter)
+
+        subpart_str = hash[:subpart] ? "-#{subpart_to_s(hash[:subpart])}" : ""
+        subpart_str = "" if subpart_str == "-"
+        "#{base}#{dots_str}#{subpart_str}"
+      end
+
+      # Per-type canonical width for the leading numeric run.
+      def canonical_base(base, type_letter)
+        case type_letter
+        when "S", "R", "G", "M", "C", "X", "P" then base.rjust(4, "0")
+        when "GA" then base.rjust(2, "0")
+        else base
+        end
+      end
+
+      # Per-type canonical width for each dotted continuation segment.
+      # `str` looks like ".13" or ".1.11" (leading dot, then digits).
+      def canonical_dots(str, type_letter)
+        return "" if str.empty?
 
-        subpart_str = subpart_to_s(hash[:subpart])
-        subpart_str.empty? ? base : "#{base}-#{subpart_str}"
+        if type_letter == "GA"
+          str.split(".", -1).map { |seg| seg.empty? ? seg : seg.rjust(2, "0") }.join(".")
+        else
+          str
+        end
       end
 
       # The subpart capture may be an array of {subpart_number: "1"},

data/lib/pubid/iala/identifier.rb

@@ -33,13 +33,17 @@ module Pubid
       attribute :language,  :string    # "E", "F", "S", "C", "A", "R", nil
 
       IALA_TYPE_MAP = {
-        "pubid:iala:standard"       => "Pubid::Iala::Identifiers::Standard",
-        "pubid:iala:recommendation" => "Pubid::Iala::Identifiers::Recommendation",
-        "pubid:iala:guideline"      => "Pubid::Iala::Identifiers::Guideline",
-        "pubid:iala:manual"         => "Pubid::Iala::Identifiers::Manual",
-        "pubid:iala:model-course"   => "Pubid::Iala::Identifiers::ModelCourse",
-        "pubid:iala:report"         => "Pubid::Iala::Identifiers::Report",
-        "pubid:iala:resolution"     => "Pubid::Iala::Identifiers::Resolution",
+        "pubid:iala:standard"          => "Pubid::Iala::Identifiers::Standard",
+        "pubid:iala:recommendation"    => "Pubid::Iala::Identifiers::Recommendation",
+        "pubid:iala:guideline"         => "Pubid::Iala::Identifiers::Guideline",
+        "pubid:iala:manual"            => "Pubid::Iala::Identifiers::Manual",
+        "pubid:iala:model-course"      => "Pubid::Iala::Identifiers::ModelCourse",
+        "pubid:iala:advice"            => "Pubid::Iala::Identifiers::Advice",
+        "pubid:iala:general-assembly"  => "Pubid::Iala::Identifiers::GeneralAssembly",
+        "pubid:iala:letter"            => "Pubid::Iala::Identifiers::Letter",
+        "pubid:iala:annex"             => "Pubid::Iala::Identifiers::Annex",
+        "pubid:iala:report"            => "Pubid::Iala::Identifiers::Report",
+        "pubid:iala:resolution"        => "Pubid::Iala::Identifiers::Resolution",
       }.freeze
 
       key_value do

data/lib/pubid/iala/identifiers/advice.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Iala
+    module Identifiers
+      # IALA Advices (A prefix). Numbering is dashed: A12-01, A13-04.
+      # Carries no edition in the corpus.
+      class Advice < Base
+        def self.type
+          { key: :advice, title: "Advice", short: "A" }
+        end
+      end
+    end
+  end
+end

data/lib/pubid/iala/identifiers/annex.rb

@@ -0,0 +1,57 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Iala
+    module Identifiers
+      # IALA Annex to a base publication. Two surface forms appear in the
+      # corpus, both wrapping a Guideline:
+      #
+      #   "G1045 Annex Ed 1"           — bare Annex, no letter
+      #   "G1128 ANNEX A Ed 1.6"        — uppercase + letter (A/B/C/D)
+      #
+      # The case of the marker is preserved (Annex vs ANNEX) for round-trip
+      # fidelity. Edition and language apply to the Annex itself, not the base.
+      class Annex < Base
+        # The publication this annex supplements. Always present; carries the
+        # type letter and number (e.g. G1045).
+        attribute :base_identifier, ::Pubid::Iala::Identifier, polymorphic: true
+        # "Annex" or "ANNEX" — preserved verbatim for exact round-trip.
+        attribute :annex_form, :string
+        # "A", "B", "C", "D", or nil for the bare form.
+        attribute :letter, :string
+
+        key_value do
+          map "base_identifier",
+              with: { to: :base_identifier_to_kv, 
+                      from: :base_identifier_from_kv }
+          map "annex_form", to: :annex_form
+          map "letter",     to: :letter
+        end
+
+        def self.type
+          { key: :annex, title: "Annex", short: nil }
+        end
+
+        # Public because lutaml-model invokes the `to:` / `from:` mapping
+        # methods via public_send during (de)serialization — a `private`
+        # section here raises NoMethodError on to_hash (matches the public
+        # visibility of Oiml::SupplementIdentifier's identical helpers).
+        def base_identifier_to_kv(model, doc)
+          base = model.base_identifier
+          return unless base
+
+          doc.add_child(
+            Lutaml::KeyValue::DataModel::Element.new("base_identifier",
+                                                     base.to_hash),
+          )
+        end
+
+        def base_identifier_from_kv(model, value)
+          return unless value
+
+          model.base_identifier = ::Pubid::Iala::Identifier.from_hash(value)
+        end
+      end
+    end
+  end
+end

data/lib/pubid/iala/identifiers/general_assembly.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Iala
+    module Identifiers
+      # IALA General Assembly resolutions (GA prefix). Numbering is dotted
+      # with a series and an index: GA01.01, GA01.13. No edition.
+      class GeneralAssembly < Base
+        def self.type
+          { key: :"general-assembly", title: "General Assembly Resolution", 
+            short: "GA" }
+        end
+      end
+    end
+  end
+end

data/lib/pubid/iala/identifiers/letter.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Iala
+    module Identifiers
+      # IALA Letters (L prefix). Numbering is dotted with a series, sub-series,
+      # and item: L2.1.11, L2.7.1-2 (the trailing -2 is a sub-part on top of
+      # the dotted number).
+      class Letter < Base
+        def self.type
+          { key: :letter, title: "Letter", short: "L" }
+        end
+      end
+    end
+  end
+end

data/lib/pubid/iala/identifiers.rb

@@ -3,14 +3,18 @@
 module Pubid
   module Iala
     module Identifiers
-      autoload :Base,           "#{__dir__}/identifiers/base"
-      autoload :Standard,       "#{__dir__}/identifiers/standard"
-      autoload :Recommendation, "#{__dir__}/identifiers/recommendation"
-      autoload :Guideline,      "#{__dir__}/identifiers/guideline"
-      autoload :Manual,         "#{__dir__}/identifiers/manual"
-      autoload :ModelCourse,    "#{__dir__}/identifiers/model_course"
-      autoload :Report,         "#{__dir__}/identifiers/report"
-      autoload :Resolution,     "#{__dir__}/identifiers/resolution"
+      autoload :Base,             "#{__dir__}/identifiers/base"
+      autoload :Advice,           "#{__dir__}/identifiers/advice"
+      autoload :Annex,            "#{__dir__}/identifiers/annex"
+      autoload :Standard,         "#{__dir__}/identifiers/standard"
+      autoload :Recommendation,   "#{__dir__}/identifiers/recommendation"
+      autoload :Guideline,        "#{__dir__}/identifiers/guideline"
+      autoload :Manual,           "#{__dir__}/identifiers/manual"
+      autoload :ModelCourse,      "#{__dir__}/identifiers/model_course"
+      autoload :GeneralAssembly,  "#{__dir__}/identifiers/general_assembly"
+      autoload :Letter,           "#{__dir__}/identifiers/letter"
+      autoload :Report,           "#{__dir__}/identifiers/report"
+      autoload :Resolution,       "#{__dir__}/identifiers/resolution"
     end
   end
 end

data/lib/pubid/iala/parser.rb

@@ -28,16 +28,32 @@ module Pubid
       rule(:lparen) { str("(") }
       rule(:rparen) { str(")") }
 
-      # Type letter — one of IALA's seven document classes. Captured to
-      # route to the right Identifiers::* subclass.
+      # Type prefix — IALA document classes. GA must be tried before G so it
+      # isn't left with a dangling "A". Captured to route to the right
+      # Identifiers::* subclass.
       rule(:type_letter) do
-        (str("S") | str("R") | str("G") | str("M") |
-         str("C") | str("X") | str("P")).as(:type_letter)
+        (str("GA") | str("S") | str("R") | str("G") | str("M") |
+         str("C") | str("A") | str("L") | str("X") | str("P")).as(:type_letter)
       end
 
-      # 4-digit document number, e.g. "1070", "0126".
+      # Document number — one or more decimal digits. The dataset historically
+      # zero-pads to 4 (S1070, R0126, M0001) but the canonical form going
+      # forward is unpadded (M1, GA1.1). Accept any length so both shapes
+      # round-trip; the number attribute preserves whatever was parsed.
       rule(:doc_number) do
-        match("[0-9]").repeat(4, 4).as(:doc_number)
+        match("[0-9]").repeat(1).as(:doc_number)
+      end
+
+      # Dotted continuation that some prefixes use as part of the number:
+      # GA01.01 (series.index), L2.1.11 (series.sub.item). Captured into the
+      # same :doc_number atom so the number attribute stays a single string
+      # (e.g. "01.01", "2.1.11") and the existing builder needs no changes.
+      rule(:dotted_number) do
+        dot >> match("[0-9]").repeat(1)
+      end
+
+      rule(:number_with_dots) do
+        doc_number >> dotted_number.repeat.as(:doc_number_dots)
       end
 
       # Numeric sub-part suffix(es): "-1", "-9-10", "-11". The catalogue
@@ -48,7 +64,7 @@ module Pubid
       end
 
       rule(:code) do
-        type_letter >> doc_number >> subpart.maybe
+        type_letter >> number_with_dots >> subpart.maybe
       end
 
       # Edition forms observed on covers and listing pages:
@@ -83,7 +99,22 @@ module Pubid
       end
 
       rule(:identifier) do
-        publisher >> code >> edition.maybe >> language.maybe
+        annex_identifier | (publisher >> code >> edition.maybe >> language.maybe)
+      end
+
+      # Annex to a base publication. Wraps the base code and adds an
+      # "Annex" / "ANNEX" marker (case preserved verbatim), an optional
+      # single-letter suffix (A/B/C/D), then the usual edition + language.
+      # Tried before the flat `identifier` so the annex marker isn't left
+      # dangling after a successful base-code match.
+      #
+      # The letter suffix uses a negative lookahead for a following lowercase
+      # letter so the "E" in "Ed 1" isn't mis-captured as annex letter "E".
+      rule(:annex_identifier) do
+        (publisher >> type_letter >> number_with_dots >> subpart.maybe).as(:base) >>
+          space >> (str("Annex") | str("ANNEX")).as(:annex_marker) >>
+          (space >> match("[A-Z]").as(:annex_letter) >> match("[a-z]").absent?).maybe >>
+          edition.maybe >> language.maybe
       end
 
       # Parse a string and return the raw parslet tree.