pubid 2.0.0.pre.alpha.27 → 2.0.0.pre.alpha.29

data/lib/pubid/ieee/renderer.rb

@@ -24,7 +24,29 @@ module Pubid
         # (Pubid::Identifier#render), so the flag is safe as instance state.
         @trademark = opts[:trademark]
 
-        render_id(@id)
+        result = render_id(@id)
+        # The long face spells draft months out ("Aug" → "August") over the
+        # bounded identifier — no title prose is ever in the string.
+        result = expand_months(result) if @long
+        result
+      end
+
+      # The long-form month names, longest key first ("Sept" before "Sep").
+      LONG_MONTHS = {
+        "Sept" => "September", "Jan" => "January", "Feb" => "February",
+        "Mar" => "March", "Apr" => "April", "Jun" => "June",
+        "Jul" => "July", "Aug" => "August", "Sep" => "September",
+        "Oct" => "October", "Nov" => "November", "Dec" => "December",
+      }.freeze
+      private_constant :LONG_MONTHS
+
+      # Expand abbreviated draft months to their long names (\b-bounded, so
+      # only whole month tokens move).
+      def expand_months(str)
+        LONG_MONTHS.each do |abbr, full|
+          str = str.gsub(/\b#{abbr}\b/, full)
+        end
+        str
       end
 
       private
@@ -125,6 +147,15 @@ module Pubid
             !id.draft_status.to_s.match?(/unapproved/i)
           type_str = id.type.dup
           type_str = type_str.sub(/^P/, "") if type_str.start_with?("P")
+          # A draft carries the word Draft alone in every face (the C2
+          # ruling): Std/Standard never renders on a draft, whatever
+          # spelling the source used — a status-less draft ("IEEE Draft
+          # Std P802.3/D2.0") renders "IEEE Draft P802.3/D2.0".
+          type_str = "Draft" if id.draft_obj && !id.draft_status
+          # The long face expands the type word ("IEEE Standard 495-2007");
+          # a draft never takes Std/Standard in any face, so the expansion
+          # only ever sees the published word.
+          type_str = "Standard" if @long && type_str == "Std"
           parts << type_str unless type_str.strip.empty?
         end
 
@@ -151,8 +182,22 @@ module Pubid
                          publishers: publishers_of(id))
 
 
-          # Only attach year to code if there's no edition, no month, and no draft
-          result += "-#{id.year}" if id.year && !id.draft_obj && !id.edition && !id.month
+          # A base year dash-attaches to the code when nothing else
+          # carries it: no edition, no identity-month date - and either
+          # 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).
+          # 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
@@ -226,12 +271,21 @@ 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
             result += " #{id.year}"
           end
+        elsif id.year && id.draft_obj && !id.edition && !year_on_code &&
+              !(id.publisher == "IEC" && id.copublisher == ["IEEE"])
+          # A year-only date trailing the draft ("ANSI N42.34/D9a,
+          # 2015") keeps its comma form - the grammar's comma-year
+          # draft tail - instead of being silently dropped. A year
+          # that dash-attached to the code above (the base year of a
+          # dated-draft family) does not repeat here.
+          result += ", #{id.year}"
         end
 
         # Bounded parenthetical markers only. The unbounded relationship/

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/ieee.rb

@@ -21,6 +21,7 @@ module Pubid
     autoload :PreParser, "#{__dir__}/ieee/pre_parser"
     autoload :Compaction, "#{__dir__}/ieee/compaction"
     autoload :ProjectRenderer, "#{__dir__}/ieee/project_renderer"
+    autoload :LongRenderer, "#{__dir__}/ieee/long_renderer"
     autoload :Renderer, "#{__dir__}/ieee/renderer"
     autoload :TypedStages, "#{__dir__}/ieee/typed_stages"
     autoload :UrnGenerator, "#{__dir__}/ieee/urn_generator"
@@ -164,3 +165,4 @@ Pubid::Registry.register(:ieee, Pubid::Ieee)
 Pubid::Ieee::Identifier.format_registry = Pubid::FormatRegistry.new(parent: Pubid::Identifier.format_registry)
 Pubid::Ieee::Identifier.format_registry.register(:human, renderer: Pubid::Ieee::Renderer)
 Pubid::Ieee::Identifier.format_registry.register(:project, renderer: Pubid::Ieee::ProjectRenderer)
+Pubid::Ieee::Identifier.format_registry.register(:long, renderer: Pubid::Ieee::LongRenderer)

data/lib/pubid/nist/components/edition.rb

@@ -1,5 +1,6 @@
 # frozen_string_literal: true
 
+require "date"
 require "lutaml/model"
 
 module Pubid
@@ -81,6 +82,10 @@ module Pubid
               # For letter suffixes, append uppercase letter directly: " r5A" or "r5A"
               if additional_text.match?(/^[A-Za-z]$/)
                 result += additional_text.upcase
+              elsif additional_text.match?(/\Ar\d+\z/)
+                # A revision marker rides attached ("e2001r3"), the 1.x
+                # corpus form
+                result += additional_text
               else
                 # For other additional text (years, etc.), use dot separator
                 result += ".#{additional_text}"
@@ -106,6 +111,10 @@ module Pubid
             # Use UPPERCASE for consistency: "r5A" for SP patterns like 800-53r5a
             elsif additional_text.match?(/^[A-Za-z]$/)
               result += additional_text.upcase # Normalize to UPPERCASE: r5A
+            elsif additional_text.match?(/\Ar\d+\z/)
+              # A revision marker rides attached ("e2001r3"), the 1.x
+              # corpus form
+              result += additional_text
             else
               # For ALL years (2-digit or 4-digit) and month+year, use DOT separator
               # Examples: e2.50, e2.1915, e2.June1908
@@ -119,8 +128,20 @@ module Pubid
         # Build long format: "Edition 2021", "Revision 5", etc.
         def build_long_format
           case type
+          # An edition date renders as the parenthesised date
+          # ("1000e2021" long: "... (2021)", "ABCe20130101" long:
+          # "... (January 01, 2013)"), the 1.x corpus form; a numbered
+          # edition keeps its worded form ("Edition 2")
           when "e"
-            "Edition #{id}"
+            case id.to_s
+            when /\A\d{4}\z/ then "(#{id})"
+            when /\A(\d{4})(\d{2})(\d{2})\z/
+              "(#{::Date.new(Regexp.last_match[1].to_i,
+                             Regexp.last_match[2].to_i,
+                             Regexp.last_match[3].to_i)
+                   .strftime('%B %d, %Y')})"
+            else "Edition #{id}"
+            end
           when "r"
             "Revision #{id}"
           when "-"

data/lib/pubid/nist/components/translation.rb

@@ -31,11 +31,13 @@ module Pubid
 
           case format
           when :short, :long
-            " #{code}"
+            # 1.x corpus shape: the trailing parenthesised suffix
+            # "(en)" (pubid#497), not a leading segment.
+            "(#{code})"
           when :mr
             ".#{code}"
           else
-            " #{code}"
+            "(#{code})"
           end
         end
       end

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

@@ -438,32 +438,42 @@ module Pubid
       end
 
       def to_full_style
-        # "National Institute of Standards and Technology Special Publication 800-27, Revision A"
+        # "National Institute of Standards and Technology Special
+        # Publication 1000 Part 1 Supplement 3 Section 2 Index Update 5
+        # (2021)(en)" - the edition renders its own long form ("Revision
+        # 5" / "(2021)"), never the derived revision view alongside it
         result = publisher_full_name
         result += " #{series_full_name}" if series
         result += " #{number}" if number
         result += parts.map { |p| "-#{p}" }.join if parts&.any?
 
-        # Render volume and issue number in long form: "Vol. 6, No. 12"
+        # The 1.x corpus order: stage, volume, part, revision, version,
+        # supplement, section, update; an edition date trails as the
+        # parenthesised date before the translation
+        result += " #{stage.to_s(:long)}" if stage
+
+        # Volume and issue number in long form: "Volume 6, No. 12" - the
+        # component's own v-notation must not leak into the worded form
         if volume && issue_number
-          result += " Vol. #{volume}, #{issue_number.to_s(:long)}"
+          result += " Volume #{volume.value}, #{issue_number.to_s(:long)}"
         elsif volume
-          result += " Vol. #{volume}"
+          result += " Volume #{volume.value}"
         end
 
-        # NEW: Use edition component properly
-        result += " #{edition.to_s(:long)}" if edition
-
-        result += ", Revision #{revision.sub(/^r/, '')}" if revision
-
-        # V2: Use version_component
+        result += " Part #{part.value}" if part
+        result += " #{edition.to_s(:long)}" if edition && !edition_year?
+        # A revision riding on a date edition composes at the revision
+        # slot ("... Part 2, Revision 3 ... (2001)")
+        if edition&.additional_text.to_s =~ /\Ar(\d+)\z/
+          result += " Revision #{Regexp.last_match[1]}"
+        end
         result += " #{version_component.to_s(:long)}" if version_component
-
-        # V2: Use update_component
+        result += " #{supplement.to_s(:long)}" if supplement
+        result += " Section #{section}" if section
+        result += " Index" if index
         result += " #{update_component.to_s(:long)}" if update_component
 
-        # V2: Use stage
-        result += " #{stage.to_s(:long)}" if stage
+        result += " #{edition.to_s(:long)}" if edition && edition_year?
 
         # V2: Use translation_component (already includes space)
         result += translation_component.to_s(:long) if translation_component
@@ -478,10 +488,9 @@ module Pubid
         result += " #{number}" if number
         result += " Part #{parts.first}" if parts&.any?
 
-        # NEW: Use edition component properly
-        result += " #{edition.to_s(:abbrev)}" if edition
-
-        result += ", Revision #{revision}" if revision
+        # The edition renders its own form ("Revision 5" / "(2021)");
+        # the derived revision view never renders alongside it
+        result += " #{edition.to_s(:long)}" if edition
 
         # V2: Use version_component
         result += " #{version_component.to_s(:abbrev)}" if version_component
@@ -565,7 +574,8 @@ module Pubid
 
         # Volume and Part components (v6n1 notation for CSM, pt1 for SP)
         if volume.is_a?(Components::Volume) && effective_part.is_a?(Components::Part)
-          # CSM series: v#n# notation
+          # the volume keeps its leading space before the part
+          # ("535 v1pt1"), the corpus form (testsuite 10491/10492)
           result += " #{volume}#{effective_part}"
         elsif effective_part.is_a?(Components::Part)
           # SP and other series: use Part.type to determine format
@@ -578,6 +588,11 @@ module Pubid
           # Render volume and issue number in short form: "v6n12"
           vol_str = volume.is_a?(Components::Volume) ? volume.to_s : "v#{volume}"
           result += "#{vol_str}n#{issue_number.number}"
+        elsif issue_number
+          # A standalone issue renders its own short form (pubid#497 -
+          # the metanorma-nist port passes the index here; n4 must not
+          # silently drop).
+          result += issue_number.to_s(:short)
         end
 
         # Use edition component properly (e2, e2021, r5, -3)
@@ -615,9 +630,12 @@ module Pubid
 
         # Add other attributes
         result += errata.to_s if errata
-        result += "index" if index
         result += "insert" if insert
+        # The 1.x order is section before index ("sup3sec2indx"); the index
+        # spelling is the abbreviated indx (pubid#497) — the full "index"
+        # input spelling canonicalizes onto it.
         result += "sec#{section}" if section
+        result += "indx" if index
         result += "app" if appendix
 
         # Add addendum - render as " Add." suffix
@@ -694,6 +712,9 @@ module Pubid
         elsif volume && issue_number
           vol_str = volume.is_a?(Components::Volume) ? volume.to_s : "v#{volume}"
           result += "#{vol_str}n#{issue_number.number}"
+        elsif issue_number
+          # A standalone issue renders its own short form (pubid#497)
+          result += issue_number.to_s(:mr)
         end
 
         # With a number, the edition glues to it per the NIST spec
@@ -727,19 +748,31 @@ module Pubid
           result += ".Add."
         end
 
+        # Section and index travel with the identifier in MR form too
+        # (pubid#497 - sec<N> / indx must not drop)
+        result += ".sec#{section}" if section
+        result += ".indx" if index
+
         # Use translation_component
         result += translation_component.to_s(:mr) if translation_component
 
         result
       end
 
+      # An e-type edition whose id is a date (year or full date)
+      # composes as the trailing parenthesised date
+      def edition_year?
+        edition&.type == "e" && edition.id.to_s =~ /\A\d{4}(\d{2}(\d{2})?)?\z/
+      end
+
       def series_full_name
         {
           "SP" => "Special Publication",
-          "FIPS" => "Federal Information Processing Standards",
+          "FIPS" => "Federal Information Processing Standards Publication",
           "IR" => "Interagency Report",
           "TN" => "Technical Note",
-        }[series] || series
+          "CSWP" => "Cybersecurity White Papers",
+        }[series.to_s] || series.to_s
       end
 
       def series_abbreviated_name

data/lib/pubid/nist/parser.rb

@@ -138,10 +138,15 @@ 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") |
+          str("ndx") |
           str("nsert") |
           str("rrata") |
           str("raft") | # NEW: Exclude "draft" from number suffix matching
@@ -264,7 +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")
@@ -282,7 +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
@@ -302,15 +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.27"
+  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.27
+  version: 2.0.0.pre.alpha.29
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-10-05 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
@@ -652,6 +653,7 @@ files:
 - lib/pubid/ieee/ire/builder.rb
 - lib/pubid/ieee/ire/identifier.rb
 - lib/pubid/ieee/ire/parser.rb
+- lib/pubid/ieee/long_renderer.rb
 - lib/pubid/ieee/nesc.rb
 - lib/pubid/ieee/nesc/builder.rb
 - lib/pubid/ieee/nesc/parser.rb