metanorma-iso 3.5.2 → 3.5.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 41f2468c77bd949ee94e2f8cbfa5f60c01b5a5633db5d1ee05b248a977966f6f
-  data.tar.gz: 2413d3b2af2054253b248a3661929282e2eb43d26f4660304805bd86abc7b19d
+  metadata.gz: 4f1dc7166ec0bc4f7cfef9674b44ab2536dc8fcca3945005c6414049de029543
+  data.tar.gz: 88a444de3e410a1bfc61427b8c563528c726ac97982d583900fa7ea6c01a5fb5
 SHA512:
-  metadata.gz: 0536b71c11b64054b8f8efc51e4eabe7430759d5c3f70189a53770b6425fdd585a141bee630c2adc63a07fa8d4da9b87cce4112b1a0e8651824191823b3903a9
-  data.tar.gz: ae3d8b25bf269f9efe7b71ff210599f30f3af7f2a222b10663714394dda19402467102774f849ee2f0c7f702d471f21426038f6c477438c98d052afa813a5f5f
+  metadata.gz: c07733849883b99aaa801630ff27fcda045fd6c7a512e7faed3f37b73af0a4925c6d92ddc6edef85adad50c34c65ddb32c1c265a96cde3c789e51c5b7782c0a6
+  data.tar.gz: 0e4f6bd8b003198fab58d3b3a30c0786b05011f4bb56ba6fb28995091b50f82baabf6f1adb9158ba9de0e6c80c8fd54c8f20c9997a30e230a097759c8a62a8f5

data/lib/metanorma/iso/validate.rb

@@ -332,6 +332,12 @@ module Metanorma
           vocab: @vocab, amd: @amd, i18n: @i18n,
           novalid: @novalid, document: @localdir
         )
+        # ModelValidator deserializes the full document into a typed model
+        # tree — for large documents that tree is gigabytes on top of the
+        # converter-phase garbage (asciidoctor tree, macro temporaries).
+        # Collect the garbage first so the spike stacks on the live set,
+        # not on the high-water mark of earlier phases.
+        GC.start
         Metanorma::Iso::Validation::ModelValidator.run(
           doc.to_xml, log: @log, state: state,
           profile: validation_profile

data/lib/metanorma/iso/validation/rules/section_sequence_rule.rb

@@ -68,7 +68,9 @@ module Metanorma
                 items << SectionItem.new(kind: clause_kind(clause), node: clause) if clause
                 clause_idx += 1
               when "terms"
-                terms = sections.terms
+                # Multiple terms clauses are an Array; the order walk is
+                # per-occurrence, so index like the clause branch does.
+                terms = Array(sections.terms)[terms_idx]
                 items << SectionItem.new(kind: :terms, node: terms) if terms
                 terms_idx += 1
               when "definitions"

data/lib/metanorma/iso/validation/rules/tree_traversal.rb

@@ -36,7 +36,11 @@ module Metanorma
             return enum_for(__method__, root) unless block_given?
 
             sections = root&.sections or return
-            visit_terms_section(sections.terms) { |term| yield(term) }
+            # sections.terms is an Array when the document carries multiple
+            # terms clauses (vocabulary corpora carry one per concept).
+            Array(sections.terms).each do |terms_section|
+              visit_terms_section(terms_section) { |term| yield(term) }
+            end
             visit_clause_subtree_for_terms(sections.clause) { |t| yield(t) }
           end
 
@@ -68,7 +72,7 @@ module Metanorma
             return enum_for(__method__, root) unless block_given?
 
             sections = root&.sections or return
-            yield(sections.terms) if sections.terms
+            Array(sections.terms).each { |terms_section| yield(terms_section) }
             visit_clauses_for_terms(sections.clause) { |t| yield(t) }
           end
 
@@ -462,7 +466,9 @@ module Metanorma
 
           def visit_clauses_for_terms(clauses)
             Array(clauses).each do |clause|
-              yield(clause.terms) if clause.class.method_defined?(:terms) && clause.terms
+              if clause.class.method_defined?(:terms) && clause.terms
+                Array(clause.terms).each { |terms_section| yield(terms_section) }
+              end
               visit_clauses_for_terms(clause.clause) { |t| yield(t) } if clause.class.method_defined?(:clause)
             end
           end

data/lib/metanorma/iso/validation/rules/unique_id_rule.rb

@@ -57,12 +57,12 @@ module Metanorma
 
           def add_id(shared, id, node)
             shared.doc_ids[id] ||= { node: node }
-            shared.id_seq << id unless shared.id_seq.include?(id)
+            shared.id_seq << id
           end
 
           def add_anchor(shared, anchor, node)
             shared.doc_anchors[anchor] ||= { node: node }
-            shared.anchor_seq << anchor unless shared.anchor_seq.include?(anchor)
+            shared.anchor_seq << anchor
           end
 
           def collect_duplicate_issues(context)

data/lib/metanorma/iso/validation/shared_state.rb

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+require "set"
+
 module Metanorma
   module Iso
     module Validation
@@ -7,6 +9,10 @@ module Metanorma
       # collection) and read by later rules (e.g. xref integrity).
       #
       # Lives on {Context} so every rule sees the same shared map within a run.
+      #
+      # The seq members are Sets: they collect one entry per identifier
+      # across the whole document, and duplicate checks run per identifier,
+      # so Array#include? made collection quadratic in document size.
       SharedState = Struct.new(
         :doc_ids, :doc_anchors, :doc_xrefs, :id_seq, :anchor_seq,
         keyword_init: true
@@ -16,8 +22,8 @@ module Metanorma
           self.doc_ids     ||= {}
           self.doc_anchors ||= {}
           self.doc_xrefs   ||= {}
-          self.id_seq      ||= []
-          self.anchor_seq  ||= []
+          self.id_seq      ||= Set.new
+          self.anchor_seq  ||= Set.new
         end
       end
     end

data/lib/metanorma/iso/version.rb

@@ -1,5 +1,5 @@
 module Metanorma
   module Iso
-    VERSION = "3.5.2".freeze
+    VERSION = "3.5.4".freeze
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: metanorma-iso
 version: !ruby/object:Gem::Version
-  version: 3.5.2
+  version: 3.5.4
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-10-30 00:00:00.000000000 Z
+date: 2026-10-31 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: metanorma-document