pubid 2.0.0.pre.alpha.38 → 2.0.0.pre.alpha.40

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e167bcdea8ccdc856546623449ce7ecb5662e48105794db9f36026b8d577f300
-  data.tar.gz: fdc189199738c34551faa29b7aa7abf0ccbdc006ef9aa5c1ae62b3f404b64cce
+  metadata.gz: c1e4c5a6889f61e535b4d05517aff5798eb5307c79bf7e318ab21822ebfe73ae
+  data.tar.gz: f129c70c4c9c646318c58e12d7a5e1097299c5698a083a0710e2866a2e7ad692
 SHA512:
-  metadata.gz: bd79b71b7fd2b457d78f225d3cc37eb2280a56b32aa24e604b14d8aa66f4f14812915d960769c4c69668f8c5b6b17d67f24399683c4e4dd6f0dd84b5ca882bd8
-  data.tar.gz: b0ce6c4e2d05416fc0ec1f77a29291bce32380efd13162ff62406bff78a60dc77c0b5404f3b61f036d868bc6b1f2c84003961cc6710b45ac0712facf6b972355
+  metadata.gz: d2e7e5cea5f0749a2aa5715521a2d6c192b43395bcb9f72470207629f6e7d1f0b9a7dbb089a1cc8e23c41a6ffc61a1e1424ce2546618249ed3eb074d611b3e4f
+  data.tar.gz: c4ee9f3166fa9caffb6f3b85d4684c11b6f08ac615ecf302af94831a9e7538a29264c2a862ffefa161fbeb4dbabc11b82405730c2435d5d666ab42c257fe2a61

data/data/ieee/update_codes.yaml

@@ -155,7 +155,9 @@ ISO/IEC/IEEE CD P26513/D1, August 2015: ISO/IEC/IEEE P26513/D1, August 2015
 "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"
+# The glued WD iteration re-spells to the face that canonicalises in one
+# round ("P42010/D-4-2019" renders "P42010.2019/D4", a second round).
+"ISO/IEC/IEEE P42010.WD4:2019(E)": "ISO/IEC/IEEE P42010.2019/D4"
 # the grammar parses the raw `_CDV/V3, Month YYYY` spelling directly
 # (joint_development_ieee_format slash-stage branch) — no shim needed
 "ISO/IEC/IEEE/P16085_DIS, March 2020": "ISO/IEC/IEEE DIS P16085-2020-03"

data/lib/pubid/ieee/builder.rb

@@ -16,7 +16,7 @@ module Pubid
         "Jun" => "06", "Jul" => "07", "Aug" => "08", "Sep" => "09",
         "Oct" => "10", "Nov" => "11", "Dec" => "12",
       }.freeze
-      attr_accessor :original_input
+      attr_accessor :original_input, :parser_input
       attr_reader :identifier_class
 
       def initialize(identifier_class = Identifier)
@@ -1535,10 +1535,12 @@ module Pubid
         type_value = "Std" if type_value == "Standard"
         draft_status_value = extract_value(parsed[:draft_status])
 
-        # Handle case where parser captured number without "P" prefix
-        # Check original input to see if there was a P before the number
-        # The ieee_p_identifier rule consumes P without capturing it
-        if !type_value && original_input&.match?(/IEEE\s+P/)
+        # Handle case where parser captured number without "P" prefix.
+        # The parser sees the NORMALIZED spelling: a rewrite that moves
+        # the project P behind a re-spelled type word ("IEEE
+        # PC37.30.2/D03 Rev 8" -> "IEEE Draft PC37.30.2Rev8/D03") must
+        # not be re-read as a project marker from the raw input.
+        if !type_value && parser_input&.match?(/IEEE\s+P/)
           type_value = "P"
         end
 
@@ -1760,6 +1762,7 @@ module Pubid
         end
 
         # Check for specific draft notation (D1, D2, etc.)
+        draft_version_present = false
         if parsed[:draft]
           draft_data = parsed[:draft]
           if draft_data.is_a?(Array)
@@ -1767,8 +1770,12 @@ module Pubid
               r.merge(e)
             end
           end
+          # A verbatim draft designator ("/D2.1.1") is a version in itself;
+          # only an attribute-less date hash ("{month:, year:}") lacks one.
+          draft_version_present = true unless draft_data.is_a?(Hash)
 
           if draft_data.is_a?(Hash) && draft_data[:draft_version]
+            draft_version_present = true
             dv = draft_data[:draft_version]
             version = if dv.is_a?(Array)
                         dv.map do |v|
@@ -1779,6 +1786,21 @@ module Pubid
                       end
 
             if version
+              # A draft version that IS an ISO stage token ("/DCD1" and
+              # "/CD1" both capture version "CD1") takes that stage on a
+              # PLAIN IEEE document: the stage token outranks the ordinal
+              # ladder (docs/IEEE-DRAFT-STAGES.md §1.3, "DDIS" = DIS), so
+              # the render drops the Draft word. A joint ISO-led
+              # stage-word draft keeps its generic project stage — its
+              # publishers block composes the same face through the
+              # typed-stage project branch.
+              copubs = parsed[:publishers].is_a?(Hash) ? parsed[:publishers][:copublishers] : nil
+              joint = copubs && !(copubs.respond_to?(:empty?) ? copubs.empty? : copubs.nil?)
+              if !joint &&
+                  (m = version.to_s.match(/\A(FDIS|FCD|CDV|DIS|CD|WD|PWI|NP)\d*\z/))
+                return m[1]
+              end
+
               # relaton's hyphenated "/D-2" spelling captures the dash inside
               # draft_version; the stage registry keys on "D2". A date-embedded
               # version ("3-2017-07") keys on its leading ordinal.
@@ -1800,13 +1822,17 @@ module Pubid
             return "P"
           elsif type_value == "Std"
             return "Std"
-          elsif type_value == "Draft Std" && parsed[:draft].nil? && parsed[:digit_draft].nil?
+          elsif type_value == "Draft Std" && !draft_version_present && parsed[:digit_draft].nil?
             # A version-less "Draft Std" fabricates /D1 (issue #205);
             # deriving the stage from that fabricated version renders the
             # same P-prefixed form as an explicit "/D1" spelling, so the
             # render reaches its fixed point in one round (pubid#318).
-            # Gated on no-draft: a dotted version ("D2.0") fails the registry
-            # lookup above and must fall through, keeping "Draft Std".
+            # Gated on no captured VERSION: a dotted version ("D2.0")
+            # fails the registry lookup above and must fall through,
+            # keeping "Draft Std". A draft-version-less tree that carries
+            # only a date ("IEEE Draft Std P754/D1.2.6 Nov 2006") takes
+            # the same fabricated D1, the version the render already
+            # implies.
             return "D1"
           elsif type_value.match?(/^No\.?$/)
             return type_value

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

@@ -384,10 +384,12 @@ module Pubid
           result = PreParser.preprocess(normalized)
           return build_dual(result.parts) if result.dispatch == :dual_semicolon
         end
-        parsed = Pubid::Parg::Backend.parse(:ieee, Parser.normalize_input(normalized))
+        parser_input = Parser.normalize_input(normalized)
+        parsed = Pubid::Parg::Backend.parse(:ieee, parser_input)
         builder = Builder.new(Identifier)
         # Pass the original input string to builder for context
         builder.original_input = input
+        builder.parser_input = parser_input
         builder.build(parsed)
       end
 

data/lib/pubid/ieee/parser.rb

@@ -1295,6 +1295,39 @@ module Pubid
         # them to the generic standard face with a slashed copublisher.
         cleaned = cleaned.gsub(%r{((?:IEEE|ISO|IEC)(?:/(?:IEEE|ISO|IEC|CSA))+)\s+Draft\s+(P[A-Z0-9][A-Z0-9.-]*/D=[A-Z])}, '\1 \2')
         cleaned = cleaned.gsub(%r{(IEEE|ISO|IEC)//}, '\1/')
+        # The three-way ISO-led list glues the language marker to the
+        # year ("ISO/IEC/IEEE 8802-1AC-2018(E)"): the spaced spelling is
+        # read as the language marker and dropped, the glued one falls
+        # into parenthetical content and prints. Re-spell the boundary.
+        # Scoped to the three-way lists: the two-way IEC/IEEE copublished
+        # form keeps its glued (E) as identity ("…:2012(E)").
+        cleaned = cleaned.gsub(
+          %r{\A((?:ISO/IEC/IEEE|IEEE/ISO/IEC|IEEE/IEC/ISO)\s+[^()]+?[^()\s])\(E\)\z},
+          '\1 (E)',
+        )
+        # The stage-first joint draft's dash part carries a bare comma
+        # year ("IEC/IEEE FDIS P60780-323, 2015"): the joint route takes
+        # no bare year there and the copublished route claims the string,
+        # so drop the non-identity year and let the stage-first route
+        # read the dash part (the rendered face is the dot human).
+        cleaned = cleaned.gsub(
+          %r{\A((?:IEC/IEEE|IEEE/IEC)\s+(?:FDIS|FCD|CDV|DIS|CD|WD|PWI|NP)\s+P\d+(?:-\d+)+),\s+\d{4}\z},
+          '\1',
+        )
+        # A two-digit draft designator spelled without its D
+        # ("IEEE P1052/08, September 2018"): the corpus face is the D
+        # ordinal ("…/D08").
+        cleaned = cleaned.gsub(
+          %r{\AIEEE\s+(P[\d.-]+)/(\d{2}),\s+([A-Za-z]+)\s+(\d{4})\z},
+          'IEEE \1/D\2, \3 \4',
+        )
+        # The revision word rides after the draft ordinal
+        # ("IEEE PC37.30.2/D03 Rev 8, October 2013"): move it onto the
+        # number and mark the document as the draft it spells.
+        cleaned = cleaned.gsub(
+          %r{\AIEEE\s+((?:P)?C?[\d.]+)/D(\d+)\s+Rev\s+(\d+),\s+([A-Za-z]+)\s+(\d{4})\z},
+          'IEEE Draft \1Rev\3/D\2, \4 \5',
+        )
         # The IEEE-led joint list's dash/underscore stage tail carries a
         # month-name date ("IEEE/ISO/IEC P29119-2-DIS, December 2011",
         # "IEEE/ISO/IEC P29119-4_FDIS, December 2020"). The generic route

data/lib/pubid/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Pubid
-  VERSION = "2.0.0.pre.alpha.38"
+  VERSION = "2.0.0.pre.alpha.40"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pubid
 version: !ruby/object:Gem::Version
-  version: 2.0.0.pre.alpha.38
+  version: 2.0.0.pre.alpha.40
 platform: ruby
 authors:
 - Ribose Inc.