pubid 2.0.0.pre.alpha.28 → 2.0.0.pre.alpha.29

data/lib/pubid/ieee/builder.rb

@@ -511,8 +511,15 @@ module Pubid
       # @param parsed_hash [Hash] parsed data with base and supplement info
       # @return [Identifiers::Corrigendum] corrigendum identifier
       def build_corrigendum_supplement(parsed_hash)
+        # The `corrigendum_identifier` rule's :base subtree binds as a
+        # top-level sequence (an Array of capture hashes — the same shape
+        # #build's dispatch merges at line 24) whenever the base carries a
+        # multi-capture narrative tail; fold it before reconstruction or
+        # reconstruct_base_string's Hash indexing crashes with TypeError.
+        base_data = parsed_hash[:base]
+        base_data = merge_parsed_array(base_data) if base_data.is_a?(Array)
         # Recursively parse the base identifier reconstructed from the subtree.
-        base = Identifier.parse(reconstruct_base_string(parsed_hash[:base]))
+        base = Identifier.parse(reconstruct_base_string(base_data))
 
         # Extract corrigendum attributes (parse-tree keys are cor_*; the
         # identifier stores them as the uniform `number`/`year`).
@@ -873,6 +880,10 @@ module Pubid
         end
         attributes[:parenthetical_content] ||= extract_value(parsed[:edition_marker]) if parsed[:edition_marker]
 
+        # A joint ISO-print redline is a distinct document of the same
+        # standard — same flag semantics as the generic path (pubid#215).
+        attributes[:redline] = true if parsed[:redline]
+
         # Extract edition, from relaton's "/E-<n>" suffix. The grammar's /E
         # alternative captures the bare ordinal (nil-residue hand-off item 1).
         if parsed[:edition]
@@ -1589,9 +1600,29 @@ module Pubid
         # Redline
         attributes[:redline] = true if parsed[:redline]
 
+        mark_glued_month(attributes)
         attributes
       end
 
+      # Glued identity year-month provenance (pubid#517): when the input's
+      # tail is the glued "-YYYY-MM" and the identity carries no draft or
+      # revision suffix, the month belongs to the code's date clause
+      # ("ANSI/IEEE 21-1976-11", "IEEE Std PC2-2016-01") and the renderer
+      # reproduces the glued face. The same month/year attribute shape is
+      # produced by trailing print dates (", 05 2014" — whose numeric form
+      # is its own documented law) and by dash-promoted draft tails
+      # ("/R-x-2011-04"), both of which keep the comma face: the input
+      # shape, checked here at the single choke point every assignment
+      # path converges to, is what distinguishes them (the AIEE
+      # date_separator precedent).
+      def mark_glued_month(attributes)
+        return unless attributes[:month] && attributes[:year] &&
+                      attributes[:draft].nil? && attributes[:revision].nil? &&
+                      original_input.to_s.match?(/-\d{4}-\d{2}\z/)
+
+        attributes[:month_glued] = true
+      end
+
       # Determine stage abbreviation for TYPED_STAGE lookup
       # @param type_value [String] the type value (e.g., "Std", "P")
       # @param draft_status_value [String] the draft status (e.g., "Unapproved")

data/lib/pubid/ieee/identifiers/base.rb

@@ -38,6 +38,12 @@ module Pubid
       attribute :revision, :string
       attribute :edition, :string                         # Edition 1.0
       attribute :month, :string
+      # Provenance flag (pubid#517): the month arrived glued to the identity
+      # year on the code ("ANSI/IEEE 21-1976-11", grammar
+      # date_with_month_numeric) rather than as a trailing ", 11 1976"
+      # print date. Serialization round-trips it so to_s reproduces the
+      # glued face.
+      attribute :month_glued, :boolean, default: -> { false }
       attribute :day, :string
       attribute :redline, :boolean, default: -> { false }
       attribute :amendments, Identifier, collection: true # Amendment identifiers
@@ -235,6 +241,7 @@ module Pubid
           Identifiers::Nesc::Draft,
           Identifiers::Nesc::Edition,
           Identifiers::Nesc::Handbook,
+          Identifiers::Nesc::Preprint,
           Identifiers::Nesc::Redline,
           Identifiers::Nesc::Standard,
         ]

data/lib/pubid/ieee/identifiers/joint_development.rb

@@ -177,6 +177,9 @@ module Pubid
           if parenthetical_content&.match?(%r{\A[A-Z](?:\s*[/&]\s*[A-Z])*\z})
             result += " (#{parenthetical_content})"
           end
+          # The redline marker is identity — the joint ISO print spells it
+          # with the space-dash form ("15289:2015(E) - Redline", pubid#215).
+          result += " - Redline" if redline
 
           result
         end
@@ -250,6 +253,7 @@ module Pubid
           result = parts.join(" ")
           result += "-#{year}" if year && !@designator_carries_year &&
                            !ieee_draft.to_s.end_with?("-#{year}")
+          result += " - Redline" if redline
 
           result
         end

data/lib/pubid/ieee/identifiers/nesc/base.rb

@@ -79,7 +79,7 @@ module Pubid
           #
           # @return [Boolean]
           def draft?
-            is_a?(Nesc::Draft)
+            is_a?(Nesc::Draft) || is_a?(Nesc::Preprint)
           end
 
           # Rendering for year-first NESC identifiers (the C2-code standard form

data/lib/pubid/ieee/identifiers/nesc/preprint.rb

@@ -0,0 +1,41 @@
+# frozen_string_literal: true
+
+module Pubid
+  module Ieee
+    module Identifiers
+      module Nesc
+        # Preprint NESC identifier — the draft stage preceding a C2 edition.
+        # "Preprint" is a draft stage (the NESC's own terminology for its
+        # pre-edition drafts), not a variant marker: the document IS a draft
+        # of the C2 code, distinct from the published edition.
+        #
+        # @example
+        #   nesc = Pubid::Ieee.parse("IEEE Std C2.2012.Preprint")
+        #   nesc.to_s  # => "IEEE Std C2.2012.Preprint"
+        class Preprint < Base
+          include Pubid::Ieee::Identifiers::CodeNumber
+
+          def self.polymorphic_name
+            "pubid:ieee:nesc-preprint"
+          end
+
+          def draft?
+            true
+          end
+
+          # Render preprint identifier. The canonical spelling is dotted
+          # (C2.2012.Preprint); the dash spelling is a non-normalized alias.
+          #
+          # @param trademark [Boolean] append the IEEE trademark symbol (™/®)
+          # @return [String] IEEE Std C2.YYYY.Preprint
+          def to_s(trademark: false, **opts)
+            body = ["C2", year, "Preprint"].compact.join(".")
+            result = ["IEEE Std", body].join(" ")
+            result += trademark_symbol if trademark
+            annotate_plain_render(result, **opts)
+          end
+        end
+      end
+    end
+  end
+end

data/lib/pubid/ieee/identifiers/nesc.rb

@@ -8,6 +8,7 @@ module Pubid
         autoload :Draft, "#{__dir__}/nesc/draft"
         autoload :Edition, "#{__dir__}/nesc/edition"
         autoload :Handbook, "#{__dir__}/nesc/handbook"
+        autoload :Preprint, "#{__dir__}/nesc/preprint"
         autoload :Redline, "#{__dir__}/nesc/redline"
         autoload :Standard, "#{__dir__}/nesc/standard"
       end

data/lib/pubid/ieee/nesc/builder.rb

@@ -88,6 +88,11 @@ module Pubid
         # @param parsed_hash [Hash] Parsed attributes
         # @return [Class] Identifier class to instantiate
         def determine_identifier_class(parsed_hash)
+          # Preprint — a draft STAGE of the C2 code (checked before the
+          # Standard route below: the c2-preprint tree carries :code "C2"
+          # too, and the verbose proposal forms carry none).
+          return Identifiers::Nesc::Preprint if parsed_hash[:preprint]
+
           # Draft identifiers
           return Identifiers::Nesc::Draft if parsed_hash[:draft]
 

data/lib/pubid/ieee/parser.rb

@@ -735,10 +735,19 @@ module Pubid
           (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,
@@ -1465,10 +1474,6 @@ module Pubid
           cleaned = "IEEE P#{cleaned}"
         end
 
-        # Part D: Suffix Normalization
-        # 5. "/Preprint" -> remove (data quality - not standard suffix)
-        cleaned = cleaned.gsub(/\/Preprint\b/, "")
-
         # Part E: Relationship Text Normalization
         # 6. "Proposed Revision of" -> "Revision of"
         cleaned = cleaned.gsub("Proposed Revision of", "Revision of")

data/lib/pubid/ieee/renderer.rb

@@ -187,9 +187,17 @@ module Pubid
           # no draft at all, or a draft that carries its OWN date (the
           # draft+corrigendum family: "PC37.09-1999/D7.2, Nov 2006" -
           # the 1999 is the base's year, the Nov 2006 the draft's).
-          year_on_code = id.year && !id.edition && !id.month &&
+          # A numeric identity month is the SHORT format: it keeps the
+          # year on the code and appends itself dash-joined, whatever the
+          # input spelling ("ANSI/IEEE 21-1976-11" and "IEEE P15026.1,
+          # 05 2014" both render glued - pubid#517). The comma face is
+          # the LONG format, a month name ("..., May 2014").
+          glued_month = id.month.to_s.match?(/\A\d{1,2}\z/)
+          year_on_code = id.year && !id.edition &&
+                         (!id.month || glued_month) &&
                          (!id.draft_obj || id.draft_obj&.year)
           result += "-#{id.year}" if year_on_code
+          result += "-#{id.month}" if glued_month
 
           # Append the numbered/lettered revision inline, in IEEE's native
           # position — right after the code number and before the draft
@@ -263,7 +271,8 @@ module Pubid
         # An exactly-IEC/IEEE co-published reference drops the comma date
         # entirely (same catalogue-metadata rule as the draft date above):
         # "IEC/IEEE P60076-16, May 2016" prints as "IEC/IEEE 60076-16".
-        if id.month && !(id.publisher == "IEC" && id.copublisher == ["IEEE"])
+        if id.month && !glued_month &&
+           !(id.publisher == "IEC" && id.copublisher == ["IEEE"])
           result += ", #{id.month}"
           result += " #{id.day}" if id.day
           if id.year && !id.edition

data/lib/pubid/ieee/urn_generator.rb

@@ -40,6 +40,13 @@ module Pubid
           parts << "int"
         end
 
+        # A NESC preprint is a draft-stage document distinct from the
+        # published edition it precedes — without this component it would
+        # share the edition's URN (urn:ieee:ieee:C2:2012).
+        if identifier.is_a?(Identifiers::Nesc::Preprint)
+          parts << "preprint"
+        end
+
         # Conformance is always a ConformanceIdentifier wrapper (special_type
         # "conformance"), never inline on a base standard — the base no longer
         # carries a conf_number attribute, so there is no inline conf part.

data/lib/pubid/nist/parser.rb

@@ -138,7 +138,11 @@ module Pubid
       # Two-letter suffixes: Ur (Unclassified Revised), Ua (Unclassified Amended), Ub-Uj (series variants)
       # Single letter: any letter not followed by excluded keywords
       rule(:number_suffix) do
-        (str("U") >> lower_letter) | (match("[a-zA-Z]") >> (
+        (str("U") >> lower_letter) |
+          # An "r" directly followed by a digit is the revision marker (pubid#170),
+          # not a suffix — leave it for the second_number/edition rules.
+          # The guard sits before the letter so it tests the suffix start itself.
+          ((str("r") >> digit).absent? >> match("[a-zA-Z]") >> (
           # Match suffixes
           str("ec") |
           str("ndex") |
@@ -265,7 +269,9 @@ module Pubid
             (digits >> str("r") >> digits >> match("[a-zA-Z]")) |
             # NEW: Revision pattern with year (e.g., "126r2013")
             # This handles SP revision format where revision is attached to second_number
-            (digits.as(:number_only) >> str("r") >> digits.as(:edition_id)) |
+            # Dotted minor accepted atomically (pubid#170): "53r4.1" binds id "4.1"
+            (digits.as(:number_only) >> str("r") >>
+             (digits >> (str(".") >> digits).maybe).as(:edition_id)) |
             # CRPL range with underscore (e.g., "2_3-1A")
             (digits >> str("_") >> digits >> dash >> digits >> upper_letter.maybe) |
             # Letter followed by dash and digits (e.g., "m-5")
@@ -283,7 +289,9 @@ module Pubid
             # For patterns like NIST SP 800-27ra where revision 'ra' is attached directly to number
             (str("r") >> match("[a-zA-Z]")).as(:revision_letter_suffix) |
             # NEW: Simple revision pattern r followed by digits (e.g., "r1", "r2") for trailing revision
-            (str("r") >> digits.as(:edition_id)).as(:revision_simple) |
+            # Dotted minor accepted atomically (pubid#170): "r4.1" binds id "4.1"
+            (str("r") >>
+             (digits >> (str(".") >> digits).maybe).as(:edition_id)).as(:revision_simple) |
             # Special patterns like "NCNR", "PERMIS", "BFRL"
             str("NCNR") | str("PERMIS") | str("BFRL") |
             # Just capital letters (e.g., "A", "B", "C") - standalone
@@ -303,15 +311,22 @@ module Pubid
       # Enhanced: Support space-separated format from preprocessing (r1 separated from number)
       rule(:edition) do
         # Edition with "e" prefix: e2, e3, e2021 (1-4 digits for ID)
-        (space.maybe >> str("e") >> digits.as(:edition_id)).as(:edition_e) |
+        # Dotted minor accepted atomically (pubid#170): "e2006.1" binds id "2006.1"
+        (space.maybe >> str("e") >>
+         (digits >> (str(".") >> digits).maybe).as(:edition_id)).as(:edition_e) |
           # Revision with "r" prefix and SPACE, with letter: r 5A (preserve format)
-          (space >> str("r") >> digits.as(:edition_id) >> match("[a-zA-Z]").as(:edition_letter)).as(:edition_r_with_space_letter) |
+          # Dotted minor accepted atomically (pubid#170)
+          (space >> str("r") >> (digits >> (str(".") >> digits).maybe).as(:edition_id) >>
+           match("[a-zA-Z]").as(:edition_letter)).as(:edition_r_with_space_letter) |
           # Revision with "r" prefix and SPACE: r 5 (preserve format)
-          (space >> str("r") >> digits.as(:edition_id)).as(:edition_r_with_space) |
+          (space >> str("r") >>
+           (digits >> (str(".") >> digits).maybe).as(:edition_id)).as(:edition_r_with_space) |
           # Revision with "r" prefix NO space, with letter: r5A (compact format)
-          (str("r") >> digits.as(:edition_id) >> match("[a-zA-Z]").as(:edition_letter)).as(:edition_r_no_space_letter) |
+          (str("r") >> (digits >> (str(".") >> digits).maybe).as(:edition_id) >>
+           match("[a-zA-Z]").as(:edition_letter)).as(:edition_r_no_space_letter) |
           # Revision with "r" prefix NO space: r5 (compact format)
-          (str("r") >> digits.as(:edition_id)).as(:edition_r_no_space) |
+          (str("r") >>
+           (digits >> (str(".") >> digits).maybe).as(:edition_id)).as(:edition_r_no_space) |
           # Revision with "rev" prefix (verbose): rev2013, rev 2013
           (space.maybe >> str("rev") >> space.maybe >> digits.as(:edition_id)).as(:edition_rev) |
           # Historical with "-" prefix: -2, -3 (ONLY if followed by non-digit or end)

data/lib/pubid/nist/preprocessor.rb

@@ -76,6 +76,7 @@ module Pubid
         normalize_supplement_and_part!
         normalize_version_notation!
         normalize_edition_year_suffix!
+        normalize_glued_dotted_edition!
         normalize_revision_with_letter!
         normalize_version_dotted_spaces!
         normalize_update_markers!
@@ -264,10 +265,24 @@ module Pubid
       end
 
       # Edition year suffix shorthand: "2006ed." → "e2006".
+      # A dotted minor after the dot stays attached (pubid#170): "2006ed.1" → "e2006.1",
+      # so the edition id binds atomically instead of corrupting into "e20061".
       def normalize_edition_year_suffix!
+        @cleaned = @cleaned.gsub(/(\d{4})ed\.(\d+)/, 'e\1.\2')
         @cleaned = @cleaned.gsub(/(\d{4})ed\./, 'e\1')
       end
 
+      # A glued dotted e-edition needs the space the spaced form has (pubid#170):
+      # "800-53e2.1" → "800-53 e2.1", so the edition rule binds "2.1" atomically.
+      # The minor is 1-2 digits — a 4-digit tail is the additional-text year of
+      # forms like "NBS CIRC 11e2.1915", which must keep its existing tree.
+      # Only dotted e-editions are unglued; the dot-less glued form "800-53e2"
+      # keeps its number-bound tree. Mirrors the "v1.1" unglue in
+      # normalize_draft_and_volume!.
+      def normalize_glued_dotted_edition!
+        @cleaned = @cleaned.gsub(/(\d)e(\d+\.\d{1,2})(?!\d)/, '\1 e\2')
+      end
+
       # Revision attached to a number with optional letter suffix. When
       # a letter suffix is present, keep them together for the
       # second_number grammar rule; otherwise insert a space before

data/lib/pubid/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Pubid
-  VERSION = "2.0.0.pre.alpha.28"
+  VERSION = "2.0.0.pre.alpha.29"
 end

metadata

@@ -1,29 +1,29 @@
 --- !ruby/object:Gem::Specification
 name: pubid
 version: !ruby/object:Gem::Version
-  version: 2.0.0.pre.alpha.28
+  version: 2.0.0.pre.alpha.29
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-10-08 00:00:00.000000000 Z
+date: 2026-10-12 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: lutaml-model
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.8'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.8'
 - !ruby/object:Gem::Dependency
   name: parslet
   requirement: !ruby/object:Gem::Requirement
@@ -640,6 +640,7 @@ files:
 - lib/pubid/ieee/identifiers/nesc/draft.rb
 - lib/pubid/ieee/identifiers/nesc/edition.rb
 - lib/pubid/ieee/identifiers/nesc/handbook.rb
+- lib/pubid/ieee/identifiers/nesc/preprint.rb
 - lib/pubid/ieee/identifiers/nesc/redline.rb
 - lib/pubid/ieee/identifiers/nesc/standard.rb
 - lib/pubid/ieee/identifiers/parenthetical_identifier.rb