pubid 2.0.0.pre.alpha.25 → 2.0.0.pre.alpha.27

data/lib/pubid/format_registry.rb

@@ -1,6 +1,15 @@
 # frozen_string_literal: true
 
 module Pubid
+  # Per-flavor string formats (renderer/parser pairs), parent-chained
+  # along the identifier class hierarchy.
+  #
+  # When lutaml-model is loaded, every registration is mirrored into
+  # Lutaml::Model::FormatRegistry (the single format registry the
+  # grammar-backed format machinery consumes), so pubid's string
+  # formats and LML-declared formats live in one registry. pubid
+  # itself does not depend on lutaml-model — the mirror is inert
+  # without it.
   class FormatRegistry
     attr_reader :formats
 
@@ -11,6 +20,7 @@ module Pubid
 
     def register(format, renderer: nil, parser: nil)
       @formats[format.to_sym] = { renderer:, parser: }
+      mirror_to_model_registry(format.to_sym, renderer, parser)
     end
 
     def renderer_for(format)
@@ -38,5 +48,32 @@ module Pubid
     def has?(format)
       @formats.key?(format.to_sym) || (@parent&.has?(format) || false)
     end
+
+    private
+
+    def mirror_to_model_registry(format, renderer, parser)
+      return unless defined?(::Lutaml::Model::FormatRegistry)
+      return if ::Lutaml::Model::FormatRegistry.registered?(format)
+
+      adapter = Class.new
+      adapter.define_singleton_method(:pubid_renderer) { renderer }
+      if parser
+        adapter.define_singleton_method(:parse) { |data, _options = {}| parser.call(data) }
+      end
+      transformer = Class.new do
+        define_singleton_method(:name) { "Pubid#{format.to_s.capitalize}Transform" }
+
+        define_method(:data_to_model) do |data, _format, _options = {}|
+          data
+        end
+      end
+      ::Lutaml::Model::FormatRegistry.register(
+        format,
+        mapping_class: ::Lutaml::Model::Mapping,
+        adapter_class: adapter,
+        transformer: transformer,
+        error_types: [Pubid::Errors::ParseError],
+      )
+    end
   end
 end

data/lib/pubid/ieee/parser.rb

@@ -1255,12 +1255,6 @@ module Pubid
         cleaned = cleaned.gsub(/\s+/, " ")
 
 
-        # NEW Session 171: CONSERVATIVE data quality fixes for TODO.IEEE-MUST-DO.txt
-        # Only fix clear typos: space before dash + 4-digit year, OR dash + space + 4-digit year
-        # Do NOT touch " - " (space-dash-space) which is valid formatting
-        cleaned = cleaned.gsub(/(\d)\s+-(\d{4})\b/, '\1-\2')  # "C37.101 -2006" → "C37.101-2006"
-        cleaned = cleaned.gsub(/(\d)-\s+(\d{4})\b/, '\1-\2')  # "C62.35- 2010" → "C62.35-2010"
-
         # NEW Session 171: Remove wrong ! prefix
         cleaned = cleaned.gsub(/^!IEEE /, "IEEE ")
 
@@ -1270,10 +1264,6 @@ module Pubid
         cleaned = cleaned.gsub("&", "&")   # Double-encoded ampersand
         cleaned = cleaned.gsub("&", "&")       # Single-encoded ampersand
 
-        # NEW: Wrap P&V notation in parentheses (Paper & Video, etc.)
-        # Pattern: "IEEE Std 500-1984 P&V" → "IEEE Std 500-1984 (P&V)"
-        cleaned = cleaned.gsub(/\s+(P&V)\s*$/, ' (\1)')
-
         # NEW Phase 1: Fix number spacing issues (e.g., "C57.1 2.25" → "C57.12.25")
         # This handles cases where a space appears in the middle of a number
         cleaned = cleaned.gsub(/(\d+\.\d+)\s+(\d+\.)/, '\1\2')
@@ -1282,16 +1272,6 @@ module Pubid
         # Remove spaces within 4-digit years
         cleaned = cleaned.gsub(/\b(1|2)\s+(\d{3})\b/, '\1\2')
 
-        # NEW: Fix month+year spacing (e.g., "March2016" → "March 2016")
-        # Add space between month name and 4-digit year when they're concatenated
-        cleaned = cleaned.gsub(
-          /\b(January|February|March|April|May|June|July|August|September|October|November|December)(\d{4})\b/, '\1 \2'
-        )
-        # Also handle abbreviated months
-        cleaned = cleaned.gsub(
-          /\b(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Sept|Oct|Nov|Dec)(\d{4})\b/, '\1 \2'
-        )
-
         # NEW: Convert IEC/IEEE space-separated to semicolon format
         # Pattern: "IEC 61523-3 First edition 2004-09; IEEE 1497" → already semicolon
         # Pattern: "IEC 62539 First Edition 2007-07 IEEE 930" → needs semicolon

data/lib/pubid/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Pubid
-  VERSION = "2.0.0.pre.alpha.25"
+  VERSION = "2.0.0.pre.alpha.27"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: pubid
 version: !ruby/object:Gem::Version
-  version: 2.0.0.pre.alpha.25
+  version: 2.0.0.pre.alpha.27
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-10-03 00:00:00.000000000 Z
+date: 2026-10-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: lutaml-model