metanorma-standoc 3.4.8 → 3.4.10

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8cbc11c6b0e55ec1404d12dd309d0f711c1d4918d079375716d0b0406f6bce2f
-  data.tar.gz: 016b74cca05a97b06809a7e9af8d48b97b07c39e3d98fe0bd575da362f1bed57
+  metadata.gz: b5094a6f74d9378c42e68b279963b6ac50f9c94412e259175c5d689111ddf6e6
+  data.tar.gz: 193d1ecb96eaba4fe068290844230558579feb43a911ac55f11d0cac086bb073
 SHA512:
-  metadata.gz: 7510bd34addc7e4e2f6d123aa19c89dffe555f1a2f06ff572824136b2269e073721bc7ac8df2f4a94ed09180bd724dbcdb2d5f6d29dc076c71650fe6d5f3c597
-  data.tar.gz: d575b92f788e95d9754b9d0d2cf6af1f670d82d2f3710d82b165a6e453f8d8ec884fc407926bbf1f86594c61f2992ee3728e9fc3455e0eeb459a2dd4b1f5bcb1
+  metadata.gz: 793e4b2daecbde3ff0778cf36373bf9e6fc4ba63705bc7b42b545dc7fa6f52a40f6776325f15d55aaa18ea16e94fafbec88540896ae04f85989d97c67474a021
+  data.tar.gz: 90d033cbd1290db48e3a8fd9c33817970c8fb7488c13782b1fbd7ff7519105d4cf5608eeed51e014f9b85645c9d7036dc745a97d6c74a00adf5f92f6ccc173b5

data/.rubocop.yml

@@ -2,6 +2,15 @@
 # See https://github.com/metanorma/cimas
 inherit_from:
   - https://raw.githubusercontent.com/riboseinc/oss-guides/main/ci/rubocop.yml
+  # .rubocop_todo.yml MUST be the last entry. inherit_from is last-wins:
+  # if listed before oss-guides, the shared config's stricter Metrics/
+  # (MethodLength, BlockLength, etc.) rules override the todo's per-file
+  # grandfathering, and the todo becomes inert on those cops. Empirically
+  # verified on suma 2026-07-06: with todo listed first, 34 Metrics/* offenses
+  # remained; moved last, cleared. Every gem in the metanorma-org fleet
+  # already ships a `.rubocop_todo.yml` on its live tree (audited 2026-07-06,
+  # 54/54 have it), so this reference is safe to emit unconditionally.
+  - .rubocop_todo.yml
 
 # Rubocop plugins enabled centrally so every metanorma-org gem picks them up
 # on cimas sync — best practice belongs at the shared-template layer, not

data/Gemfile

@@ -6,5 +6,10 @@ git_source(:github) { |repo| "https://github.com/#{repo}" }
 
 gemspec
 
-eval_gemfile("Gemfile.devel") rescue nil
+# Stopgap: lutaml 0.11.x removed lib/lutaml/xmi.rb, which metanorma-plugin-lutaml
+# 0.7.x still `require`s. Hold lutaml at 0.10.x until plugin-lutaml follows the
+# file to its new home. Remove once resolved:
+# https://github.com/metanorma/metanorma-plugin-lutaml/issues/292
+gem "lutaml", "< 0.11"
 
+eval_gemfile("Gemfile.devel") rescue nil

data/lib/isodoc/html/htmlstyle.css

@@ -137,6 +137,10 @@ dl dd {
   grid-column-start: 2;
 }
 
+dl:has(dt .stem) {
+  grid-template-columns: fit-content(8em) auto;
+}
+
 b, strong {
   font-weight: bold;
 }
@@ -150,6 +154,11 @@ a.FootnoteRef, span.FootnoteRef {
   vertical-align: super;
 }
 
+a.TableFootnoteRef, a.FootnoteRef, a.footnote-number,
+sup a, a:has(> sup) {
+  vertical-align: baseline;
+}
+
 .addition {
   color: blue;
 }

data/lib/metanorma/cleanup/bibdata.rb

@@ -198,15 +198,23 @@ module Metanorma
         XML
       end
 
-      def published_base?(stage, _xmldoc)
+      def published_base?(stage, xmldoc)
         if @stage_published
           @stage_published == "true"
-        else published?(stage, _xmldoc)
+        else !draft_stage?(stage, xmldoc)
         end
       end
 
+      # Is the stage a draft (unpublished) stage? Overridable seam for
+      # flavours whose stage repertoires are not the binary
+      # published/unpublished; the :docstage-published: document
+      # attribute (@stage_published, above) trumps it either way.
+      def draft_stage?(stage, xmldoc)
+        !published?(stage, xmldoc)
+      end
+
       def published?(stage, _xmldoc)
-        stage.casecmp("published").zero?
+        stage.to_s.casecmp("published").zero?
       end
 
       # allows us to deal with doc relation localities,

data/lib/metanorma/cleanup/bibitem.rb

@@ -34,6 +34,26 @@ module Metanorma
           bib.children = new.children.to_xml
         else bib << new.children.to_xml
         end
+        bibitem_reorder_elems(bib)
+      end
+
+      # canonical element order of BibliographicItem per biblio.rnc
+      BIBITEM_ELEM_ORDER =
+        %w(fetched formattedref title uri docidentifier docnumber date
+           contributor edition version note language locale script abstract
+           status copyright relation series medium place price extent size
+           accesslocation license classification keyword validity depiction)
+          .freeze
+
+      # the spans-derived fragment is appended after any elements already on
+      # the bibitem (e.g. formattedref, docnumber, date), which violates the
+      # ordered relaton content model; restore grammar order, stably
+      def bibitem_reorder_elems(bib)
+        bib.element_children
+          .sort_by
+          .with_index do |e, i|
+            [BIBITEM_ELEM_ORDER.index(e.name) || BIBITEM_ELEM_ORDER.size, i]
+          end.each { |e| bib << e }
       end
 
       def merge_bibitem_from_formattedref_span_attrs(bib, new)

data/lib/metanorma/cleanup/cleanup.rb

@@ -61,7 +61,7 @@ module Metanorma
           boilerplateauthority embed_hdr embed_id erefstyle originstyle
           xrefstyle blockunnumbered keepasciimath numberfmt_formula
           numberfmt_prof sort_biblio reqt_models default_requirement_model
-          document_scheme
+          document_scheme publisher_sort_config
         ]
       end
 

data/lib/metanorma/cleanup/index.rb

@@ -1,10 +1,13 @@
 module Metanorma
   module Standoc
     module Index
+      # normalize-space(.) rather than normalize-space(text()): index terms
+      # wrapped entirely in formatting (<em>a priori</em>, stem) have no
+      # direct text nodes but are not empty
       EMPTY_INDEX_XPATH =
-        "//index[not(.//primary[normalize-space(text())]) " \
-        "or .//secondary[. and not(normalize-space(text()))] " \
-        "or .//tertiary[. and not(normalize-space(text()))]]".freeze
+        "//index[not(.//primary[normalize-space(.)]) " \
+        "or .//secondary[. and not(normalize-space(.))] " \
+        "or .//tertiary[. and not(normalize-space(.))]]".freeze
 
       def index_cleanup(xmldoc)
         para_index_cleanup(xmldoc)

data/lib/metanorma/cleanup/ref.rb

@@ -34,6 +34,77 @@ module Metanorma
         bib
       end
 
+      BIBLIO_PUBLISHER_ORG =
+        "./contributor[role/@type = 'publisher']/organization".freeze
+
+      # Configurable publisher sort ranking, shared by every flavour's
+      # bibliography sort. `table` is the flavour's default ordering, an array
+      # of { abbrev:, name:, rank: }. A per-document / metanorma-taste override
+      # (:sort-biblio-<abbrev>: <rank>: <name> attributes, captured into
+      # @publisher_sort_config) REPLACES the flavour default when present.
+      #
+      # A matched publisher gets its configured rank; an unmatched standards
+      # reference sorts immediately after the ranked publishers, and a
+      # non-standard reference last. With no override and the flavour's own
+      # default table, this reproduces the flavour's historical hardcoded ranks.
+      def publisher_sort_rank(bib, table)
+        table = publisher_sort_table(table)
+        if (entry = publisher_sort_match(bib, table))
+          return entry[:rank]
+        end
+
+        maxrank = table.map { |e| e[:rank] }.max || 0
+        biblio_standards_ref?(bib) ? maxrank + 1 : maxrank + 2
+      end
+
+      # Secondary sort token: the co-publisher(s) other than the matched
+      # primary. Deterministic when several co-publishers exist (returns the
+      # smallest token), which reproduces e.g. JIS's "JIS+IEC before
+      # JIS+ISO, JIS+IEC+ISO alongside JIS+IEC" ordering.
+      def publisher_sort_second(bib, table)
+        entry = publisher_sort_match(bib, publisher_sort_table(table))
+        entry or return ""
+        bib.xpath(BIBLIO_PUBLISHER_ORG)
+          .reject { |o| publisher_org_match?(o, entry) }
+          .map { |o| (o.at("./abbreviation") || o.at("./name"))&.text&.strip }
+          .compact.reject(&:empty?).min || ""
+      end
+
+      def publisher_sort_table(default_table)
+        cfg = @publisher_sort_config
+        cfg && !cfg.empty? ? cfg : default_table
+      end
+
+      # lowest-ranked table entry whose publisher appears on the bibitem
+      def publisher_sort_match(bib, table)
+        orgs = bib.xpath(BIBLIO_PUBLISHER_ORG)
+        table.sort_by { |e| e[:rank] }.find do |e|
+          orgs.any? { |o| publisher_org_match?(o, e) }
+        end
+      end
+
+      # A table entry matches a publisher organization on its abbreviation
+      # (case-insensitive, since AsciiDoc lowercases attribute-derived abbrevs)
+      # or on any of its names (`name` may be a single string or an array; the
+      # latter lets a flavour list, e.g., a Japanese and an English publisher
+      # name for the same rank).
+      def publisher_org_match?(org, entry)
+        abbr = org.at("./abbreviation")&.text&.strip
+        name = org.at("./name")&.text&.strip
+        names = Array(entry[:name]).map { |n| n.to_s.strip }.reject(&:empty?)
+        (!entry[:abbrev].to_s.empty? && abbr&.casecmp?(entry[:abbrev])) ||
+          (!name.to_s.empty? && names.include?(name))
+      end
+
+      # Whether a bibitem is a standards reference (sorts after the ranked
+      # publishers, before non-standard references). Base definition: it has a
+      # typed, non-DOI/ISSN/ISBN docidentifier. Flavours may widen this (ISO
+      # also treats an untyped docidentifier as a standards reference).
+      def biblio_standards_ref?(bib)
+        !!bib.at("./docidentifier[@type][not(#{@conv.skip_docid} or " \
+                 "@type = 'metanorma')]")
+      end
+
       # default presuppose that all citations in biblio numbered
       # consecutively, but that standards codes are preserved as is:
       # only numeric references are renumbered

data/lib/metanorma/cleanup/review.rb

@@ -35,6 +35,11 @@ module Metanorma
               when "ol", "ul" then node.at(".//p | .//name") || node.at("./li")
               when "dl" then node.at(".//p | .//name") || node.at("./dt | ./dd")
               when "table" then node.at(".//td[text()] | .//th[text()]")
+              # terms and their children have no legal place for a loose
+              # paragraph, so the bookmark must land inside existing content
+              when "term", "definition", "verbal-definition", "preferred",
+                "admitted", "deprecates", "related", "termnote", "termexample"
+                node.at(".//p | .//name")
               end or return nil
         first_non_stem_text(ret)
       end

data/lib/metanorma/cleanup/spans_to_bibitem.rb

@@ -58,7 +58,9 @@ module Metanorma
             ret += span_to_docid(s, "classification")
           end
           spans[:keyword]&.each do |s|
-            ret += span_to_docid(s, "keyword")
+            # keyword content is structured per the grammar: an uncontrolled
+            # keyword is a vocab entry, not bare text
+            ret += "<keyword><vocab>#{s[:val]}</vocab></keyword>"
           end
           spans[:image]&.each do |s|
             ret += "<depiction>#{s[:val]}</depiction>"

data/lib/metanorma/cleanup/term_lookup.rb

@@ -48,6 +48,7 @@ module Metanorma
       def concept_cleanup
         xmldoc.xpath("//concept").each do |n|
           refterm = n.at("./refterm") or next
+          refterm.xpath(".//index").each(&:remove)
           lookup = norm_ref_id_text(refterm.text.strip)
           p = @lookup[:sec2prim][lookup] and refterm.children = @c.encode(p)
         end

data/lib/metanorma/converter/base.rb

@@ -58,7 +58,7 @@ module Metanorma
             pronunciation|grammar|term|terms|termnote|termexample|source|
             origin|termref|modification)>)}x, "\\1\n")
           .gsub(%r{(<(title|name))}, "\n\\1")
-          .gsub(%r{(<sourcecode[^<>]*>)\s+(<name[^<>]*>[^<]+</name>)\s+},
+          .gsub(%r{(<sourcecode[^<>]*+>)\s++(<name[^<>]*+>(?>.*?</name>))\s++}m,
                 "\\1\\2")
       end
 

data/lib/metanorma/converter/converter.rb

@@ -54,6 +54,12 @@ module Metanorma
         preprocessor Metanorma::Standoc::NamedEscapePreprocessor
         inline_macro Metanorma::Standoc::PreferredTermInlineMacro
         inline_macro Metanorma::Standoc::DateInlineMacro
+        # PassFormat must run before Span (and the other bracketed inline macros
+        # below) so that a nested pass/pass-format passthrough is resolved to a
+        # bracket-free placeholder before Span delimits its own [...]; otherwise
+        # Span stops at the inner passthrough's ] and over-captures. See
+        # metanorma-standoc#1221.
+        inline_macro Metanorma::Standoc::PassFormatInlineMacro
         inline_macro Metanorma::Standoc::SpanInlineMacro
         inline_macro Metanorma::Standoc::AltTermInlineMacro
         inline_macro Metanorma::Standoc::AdmittedTermInlineMacro
@@ -79,7 +85,6 @@ module Metanorma
         inline_macro Metanorma::Standoc::FormSelectMacro
         inline_macro Metanorma::Standoc::FormOptionMacro
         inline_macro Metanorma::Standoc::ToCInlineMacro
-        inline_macro Metanorma::Standoc::PassFormatInlineMacro
         inline_macro Metanorma::Standoc::StdLinkInlineMacro
         inline_macro Metanorma::Standoc::NumberInlineMacro
         inline_macro Metanorma::Standoc::TrStyleInlineMacro
@@ -111,7 +116,8 @@ module Metanorma
       $xreftext = {}
 
       attr_reader :bibdb, :lang, :script, :locale, :relaton_log, :libdir, :backend,
-                  :novalid, :local_log, :output_dir, :filename, :files_to_delete
+                  :novalid, :local_log, :output_dir, :filename, :files_to_delete,
+                  :docstage_valid
 
       def initialize(backend, opts)
         doc = opts&.dig(:document)

data/lib/metanorma/converter/front.rb

@@ -174,7 +174,12 @@ module Metanorma
       def metadata_keywords(node, xml)
         node.attr("keywords") or return
         node.attr("keywords").split(/,\s*/).each do |kw|
-          add_noko_elem(xml, "keyword", kw)
+          kw.empty? and next
+          # keyword content is structured per the grammar: an uncontrolled
+          # keyword is a vocab entry, not bare text
+          xml.keyword do |k|
+            add_noko_elem(k, "vocab", kw)
+          end
         end
       end
 

data/lib/metanorma/converter/init.rb

@@ -27,6 +27,7 @@ module Metanorma
         @document_scheme = document_scheme(node)
         @default_doctype = "standard"
         @stage_published = node.attr("docstage-published")
+        @docstage_valid = node.attr("docstage-valid")&.split(/,\s*/)
         @metadata_attrs = metadata_attrs(node)
       end
 
@@ -116,6 +117,11 @@ module Metanorma
       end
 
       def init_toc(node)
+        init_toc_levels(node)
+        init_toc_types(node)
+      end
+
+      def init_toc_levels(node)
         @htmltoclevels = node.attr("toclevels-html") ||
           node.attr("htmltoclevels") ||
           node.attr("toclevels") || toc_default[:html_levels]
@@ -125,10 +131,13 @@ module Metanorma
         @pdftoclevels = node.attr("toclevels-pdf") ||
           node.attr("toclevels") || toc_default[:pdf_levels]
         @toclevels = node.attr("toclevels") || toc_default[:html_levels]
-        @tocfigures = node.attr("toc-figures")
-        @toctables = node.attr("toc-tables")
-        @tocrecommendations = node.attr("toc-recommendations")
-        @tocexamples = node.attr("toc-examples")
+      end
+
+      def init_toc_types(node)
+        @tocfigures = node.attr("toc-figures") != "false"
+        @toctables = node.attr("toc-tables") != "false"
+        @tocrecommendations = node.attr("toc-recommendations") != "false"
+        @tocexamples = node.attr("toc-examples") != "false"
       end
 
       def toc_default
@@ -175,6 +184,7 @@ module Metanorma
         @no_isobib = node.attr("no-isobib")
         @flush_caches = node.attr("flush-caches")
         @sort_biblio = node.attr("sort-biblio") != "false"
+        @publisher_sort_config = init_publisher_sort_config(node)
         init_bib_db(node)
         init_bib_caches(node)
         init_iev_caches(node)
@@ -182,6 +192,21 @@ module Metanorma
         @biblio_cutoff = biblio_cutoff(node)
       end
 
+      # Per-publisher bibliography sort ranks, from
+      # :sort-biblio-<abbrev>: <rank>: <name> document attributes (set directly
+      # in a document or injected by a metanorma-taste config). AsciiDoc
+      # lowercases attribute names, so <abbrev> is matched case-insensitively
+      # downstream. Returns [] when none are set (flavours keep their default
+      # publisher ordering).
+      def init_publisher_sort_config(node)
+        node.attributes.each_with_object([]) do |(k, v), acc|
+          m = /\Asort-biblio-(?<abbrev>.+)\z/.match(k.to_s) or next
+          rank_s, name = v.to_s.split(":", 2)
+          rank = Integer(rank_s.to_s.strip, exception: false) or next
+          acc << { abbrev: m[:abbrev].strip, name: name&.strip, rank: rank }
+        end
+      end
+
       def init_bib_db(node)
         @bibdb = nil
         @local_bibdb =

data/lib/metanorma/converter/inline.rb

@@ -160,12 +160,19 @@ module Metanorma
         uri
       end
 
+      # empty-string dimensions (image::x.png["","",""]) must fall back to
+      # "auto": ImageSize admits only numbers, percentages, or "auto"
+      def image_dim_attr(node, key)
+        ret = node.attr(key)
+        ret.nil? || ret.empty? ? "auto" : ret
+      end
+
       def image_attributes1(node, uri, sourceuri, type, altmedia)
         /^data:/.match?(sourceuri) and sourceuri = nil
         attr_code(altmedia_id_attr(node, altmedia)
           .merge(src: uri, mimetype: type,
-                 height: node.attr("height") || "auto",
-                 width: node.attr("width") || "auto",
+                 height: image_dim_attr(node, "height"),
+                 width: image_dim_attr(node, "width"),
                  filename: node.attr("filename") || sourceuri,
                  title: node.attr("titleattr"),
                  media: node.attr("media"), altmedia: altmedia,

data/lib/metanorma/converter/lists.rb

@@ -103,6 +103,7 @@ module Metanorma
           .merge(
             metadata: node.option?("metadata") ? "true" : nil,
             key: node.option?("key") ? "true" : nil,
+            class: node.attr("class"),
           )))
       end
 

data/lib/metanorma/converter/macros.rb

@@ -67,16 +67,23 @@ module Metanorma
         end.join
       end
 
-      # debugging output of results of all preprocessing,
-      # including include files concatenation and Lutaml/Liquid processing
+      # debugging output of results of all preprocessing, including include
+      # files concatenation and Lutaml/Liquid processing.
+      #
+      # Skipped for docfile-less (string/stdin) input: the target would
+      # degenerate to a single CWD-relative "./metanorma.asciidoc.log.txt" that
+      # the thousands of collection sub-compiles all truncate-rewrite (so it
+      # carries no useful diagnostics anyway). And a diagnostic write must never
+      # abort the build, so a write failure degrades to a warning rather than
+      # propagating out of the preprocessor. metanorma/metanorma-standoc#1209
       def log(doc, text)
-        source = doc.attr("docfile") || "metanorma"
+        source = doc.attr("docfile") or return
         dirname  = File.dirname(source)
         basename = File.basename(source, ".*")
         fname = File.join(dirname, "#{basename}.asciidoc.log.txt")
-        File.open(fname, "w:UTF-8") do |f|
-          f.write(text.join("\n"))
-        end
+        File.open(fname, "w:UTF-8") { |f| f.write(text.join("\n")) }
+      rescue SystemCallError => e
+        warn "metanorma: could not write preprocessing log #{fname}: #{e.message}"
       end
     end
 

data/lib/metanorma/standoc/version.rb

@@ -19,6 +19,6 @@ module Metanorma
   end
 
   module Standoc
-    VERSION = "3.4.8".freeze
+    VERSION = "3.4.10".freeze
   end
 end

data/lib/metanorma/validate/basicdoc.rng

@@ -1527,12 +1527,16 @@ to indicate that the index range covers a range of locations between the current
       <ref name="index-primary">
         <a:documentation>Primary index term to be recorded at the current location</a:documentation>
       </ref>
-      <ref name="index-secondary">
-        <a:documentation>Secondary index term to be recorded at the current location</a:documentation>
-      </ref>
-      <ref name="index-tertiary">
-        <a:documentation>Tertiary index term to be recorded at the current </a:documentation>
-      </ref>
+      <optional>
+        <ref name="index-secondary">
+          <a:documentation>Secondary index term to be recorded at the current location</a:documentation>
+        </ref>
+      </optional>
+      <optional>
+        <ref name="index-tertiary">
+          <a:documentation>Tertiary index term to be recorded at the current </a:documentation>
+        </ref>
+      </optional>
     </element>
   </define>
   <define name="index-xref">
@@ -1547,12 +1551,16 @@ The text in the inline element is the primary index term to be be cross-referenc
       <ref name="index-primary">
         <a:documentation>The primary index term to be cross-referenced</a:documentation>
       </ref>
-      <ref name="index-secondary">
-        <a:documentation>The secondary index term to be cross-referenced</a:documentation>
-      </ref>
-      <ref name="index-tertiary">
-        <a:documentation>The tertiary index term to be cross-referenced</a:documentation>
-      </ref>
+      <optional>
+        <ref name="index-secondary">
+          <a:documentation>The secondary index term to be cross-referenced</a:documentation>
+        </ref>
+      </optional>
+      <optional>
+        <ref name="index-tertiary">
+          <a:documentation>The tertiary index term to be cross-referenced</a:documentation>
+        </ref>
+      </optional>
       <element name="target">
         <a:documentation>The index term to be cross-referenced to</a:documentation>
         <oneOrMore>
@@ -1931,9 +1939,10 @@ or as the string "auto"</a:documentation>
     </element>
   </define>
   <define name="mathml">
-    <a:documentation>Encoding of MathML: the official W3C MathML 3.0 grammar (namespace
+    <a:documentation>Encoding of MathML: the official W3C MathML 4 grammar (namespace
 http://www.w3.org/1998/Math/MathML), via the metanorma wrapper in
-grammars/mathml/. See grammars/mathml/README.adoc and basicdoc-models#35.</a:documentation>
+grammars/mathml/. See grammars/mathml/README.adoc, basicdoc-models#35,
+and the MathML 3 -&gt; 4 migration: basicdoc-models#39 / metanorma#585.</a:documentation>
     <externalRef href="metanorma-mathml.rng"/>
   </define>
   <define name="StemAttributes">

data/lib/metanorma/validate/isodoc.rng

@@ -1,6 +1,6 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <grammar xmlns:a="http://relaxng.org/ns/compatibility/annotations/1.0" xmlns="http://relaxng.org/ns/structure/1.0" datatypeLibrary="http://www.w3.org/2001/XMLSchema-datatypes">
-<!-- VERSION v2.1.6 -->
+<!-- VERSION v2.2.0 -->
 
   <!--
     ALERT: cannot have root comments, because of https://github.com/metanorma/metanorma/issues/437