metanorma-standoc 3.0.3 → 3.0.5

data/lib/metanorma/standoc/regex.rb

@@ -0,0 +1,78 @@
+module Metanorma
+  module Standoc
+    module Regex
+      # https://medium.com/@rickwang_wxc/in-ruby-given-a-string-detect-if-it-is-valid-numeric-c58275eace60
+      NUMERIC_REGEX = %r{^((\+|-)?\d*\.?\d+)([eE](\+|-){1}\d+)?$}
+
+      # extending localities to cover ISO referencing
+      CONN_REGEX_STR = "(?<conn>and|or|from|to)!".freeze
+
+      LOCALITIES = "section|clause|part|paragraph|chapter|page|line|" \
+        "table|annex|figure|example|note|formula|list|time|anchor|" \
+        "locality:[^ \\t\\n\\r:,;=]+".freeze
+
+      LOCALITY_REGEX_STR = <<~REGEXP.freeze
+        ^((#{CONN_REGEX_STR})?
+            (?<locality>#{LOCALITIES})(\\s+|=)
+               (?<ref>[^"][^ \\t\\n,:;-]*|"[^"]+")
+                 (-(?<to>[^"][^ \\t\\n,:;-]*|"[^"]"))?|
+          (?<locality2>whole|title|locality:[^ \\t\\n\\r:,;=]+))(?<punct>[,:;]?)\\s*
+         (?<text>.*)$
+      REGEXP
+
+      def to_regex(str)
+        Regexp.new(str.gsub(/\s/, ""), Regexp::IGNORECASE | Regexp::MULTILINE)
+      end
+
+      LOCALITY_REGEX_VALUE_ONLY_STR = <<~REGEXP.freeze
+        ^(?<conn0>(#{CONN_REGEX_STR}))
+          (?!whole|title|locality:)
+          (?<value>[^=,;:\\t\\n\\r]+)
+          (?<punct>[,;\\t\\n\\r]|$)
+      REGEXP
+
+      LOCALITY_REGEX_STR_TRIPLEDASH = <<~REGEXP.freeze
+        ^(?<locality>(#{CONN_REGEX_STR})?
+            (#{LOCALITIES})(\\s+|=))
+               (?<ref>[^"][^ \\t\\n,:;-]*
+                 -[^ \\t\\n,:;"-]+
+                 -[^ \\t\\n,:;"]+)
+          (?<text>[,:;]?\\s*
+         .*)$
+      REGEXP
+
+      TERM_REFERENCE_RE_STR = <<~REGEXP.freeze
+        ^(?<xref><(xref|concept)[^>]+>(.*?</(xref|concept)>)?)
+               (,\s(?<text>.*))?
+        $
+      REGEXP
+      TERM_REFERENCE_RE =
+        Regexp.new(TERM_REFERENCE_RE_STR.gsub(/\s/, "").gsub(/_/, "\\s"),
+                   Regexp::IGNORECASE | Regexp::MULTILINE)
+
+      ISO_REF =
+        %r{^<ref\sid="(?<anchor>[^"]+)">
+      \[(?<usrlbl>\([^)]+\))?(?<code>(?:ISO|IEC)[^0-9]*\s[0-9-]+|IEV)
+      (?::(?<year>[0-9][0-9-]+))?\]</ref>,?\s*(?<text>.*)$}xm
+
+      ISO_REF_NO_YEAR =
+        %r{^<ref\sid="(?<anchor>[^"]+)">
+      \[(?<usrlbl>\([^)]+\))?(?<code>(?:ISO|IEC)[^0-9]*\s[0-9-]+):
+      (?:--|–|—|&\#821[12];)\]</ref>,?\s*
+        (?:<fn[^>]*>\s*<p>(?<fn>[^\]]+)</p>\s*</fn>)?,?\s?(?<text>.*)$}xm
+
+      ISO_REF_ALL_PARTS =
+        %r{^<ref\sid="(?<anchor>[^"]+)">
+      \[(?<usrlbl>\([^)]+\))?(?<code>(?:ISO|IEC)[^0-9]*\s[0-9]+)
+      (?::(?<year>--|–|—|&\#821[12];|[0-9][0-9-]+))?\s
+      \(all\sparts\)\]</ref>,?\s*
+        (?:<fn[^>]*>\s*<p>(?<fn>[^\]]+)</p>\s*</fn>,?\s?)?(?<text>.*)$}xm
+
+      NON_ISO_REF = %r{^<ref\sid="(?<anchor>[^"]+)">
+      \[(?<usrlbl>\([^)]+\))?(?<code>.+?)\]</ref>,?\s*(?<text>.*)$}xm
+
+      NON_ISO_REF1 = %r{^<ref\sid="(?<anchor>[^"]+)">
+      (?<usrlbl>\([^)]+\))?(?<code>.+?)</ref>,?\s*(?<text>.*)$}xm
+    end
+  end
+end

data/lib/metanorma/standoc/reqt.rng

@@ -1,5 +1,11 @@
 <?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">
+  <define name="requirement">
+    <element name="requirement">
+      <ref name="RequirementType"/>
+    </element>
+  </define>
+  <!-- ALERT: cannot have root comments, because of https://github.com/metanorma/metanorma/issues/437 -->
   <!--
     Presupposes isodoc.rnc, is included in it
     include "isodoc.rnc" { }
@@ -7,14 +13,9 @@
     This is the Metanorma default provisions model; it is overridden by other provisions models,
     such as Modspec
   -->
-  <define name="requirement">
+  <define name="recommendation">
     <a:documentation>Specification of an attribute of a subject that is required.
 NOTE: A requirement can contain other requirements</a:documentation>
-    <element name="requirement">
-      <ref name="RequirementType"/>
-    </element>
-  </define>
-  <define name="recommendation">
     <a:documentation>Specification of an attribute of a subject that is recommended</a:documentation>
     <element name="recommendation">
       <ref name="RequirementType"/>

data/lib/metanorma/standoc/section.rb

@@ -52,7 +52,7 @@ module Metanorma
 
       PREFACE_CLAUSE_NAMES =
         %w(abstract foreword introduction metanorma-extension termdocsource
-           misc-container acknowledgements).freeze
+           misc-container metanorma-extension acknowledgements).freeze
 
       MAIN_CLAUSE_NAMES =
         ["normative references", "terms and definitions", "scope",

data/lib/metanorma/standoc/terms.rb

@@ -140,15 +140,6 @@ module Metanorma
         end
       end
 
-      TERM_REFERENCE_RE_STR = <<~REGEXP.freeze
-        ^(?<xref><(xref|concept)[^>]+>(.*?</(xref|concept)>)?)
-               (,\s(?<text>.*))?
-        $
-      REGEXP
-      TERM_REFERENCE_RE =
-        Regexp.new(TERM_REFERENCE_RE_STR.gsub(/\s/, "").gsub(/_/, "\\s"),
-                   Regexp::IGNORECASE | Regexp::MULTILINE)
-
       def extract_termsource_refs(text, node)
         matched = TERM_REFERENCE_RE.match text
         matched.nil? and @log.add("AsciiDoc Input", node,
@@ -180,6 +171,8 @@ module Metanorma
           end
         end
       end
+
+      include ::Metanorma::Standoc::Regex
     end
   end
 end

data/lib/metanorma/standoc/utils.rb

@@ -100,6 +100,14 @@ module Metanorma
         para
       end
 
+      def insert_before(xmldoc, xpath)
+        unless ins = xmldoc.at(xpath).children.first
+          xmldoc.at(xpath) << " "
+          ins = xmldoc.at(xpath).children.first
+        end
+        ins
+      end
+
       def xml_encode(text)
         @c.encode(text, :basic, :hexadecimal)
           .gsub("&amp;gt;", ">").gsub("&amp;lt;", "<").gsub("&amp;amp;", "&")

data/lib/metanorma/standoc/validate.rb

@@ -2,8 +2,8 @@ require "metanorma/standoc/utils"
 require_relative "validate_section"
 require_relative "validate_table"
 require_relative "validate_term"
+require_relative "validate_schema"
 require "nokogiri"
-require "jing"
 require "iev"
 require "pngcheck"
 
@@ -98,49 +98,6 @@ module Metanorma
         @log.add("Style", i, err2)
       end
 
-      def schema_validate(doc, schema)
-        Tempfile.open(["tmp", ".xml"], encoding: "UTF-8") do |f|
-          schema_validate1(f, doc, schema)
-        rescue Jing::Error => e
-          clean_abort("Jing failed with error: #{e}", doc)
-        ensure
-          f.close!
-        end
-      end
-
-      def schema_validate1(file, doc, schema)
-        file.write(to_xml(doc))
-        file.close
-        errors = Jing.new(schema, encoding: "UTF-8").validate(file.path)
-        warn "Syntax Valid!" if errors.none?
-        errors.each do |e|
-          @log.add("Metanorma XML Syntax",
-                   "XML Line #{'%06d' % e[:line]}:#{e[:column]}", e[:message])
-        end
-      end
-
-      SVG_NS = "http://www.w3.org/2000/svg".freeze
-
-      WILDCARD_ATTRS =
-        "//*[@format] | //stem | //bibdata//description | " \
-        "//formattedref | //bibdata//note | //bibdata/abstract | " \
-        "//bibitem/abstract | //bibitem/note | //metanorma-extension".freeze
-
-      # RelaxNG cannot cope well with wildcard attributes. So we strip
-      # any attributes from FormattedString instances (which can contain
-      # xs:any markup, and are signalled with @format) before validation.
-      def formattedstr_strip(doc)
-        doc.xpath(WILDCARD_ATTRS, "m" => SVG_NS).each do |n|
-          n.elements.each do |e|
-            e.traverse do |e1|
-              e1.element? and e1.each { |k, _v| e1.delete(k) } # rubocop:disable Style/HashEachMethods
-            end
-          end
-        end
-        doc.xpath("//m:svg", "m" => SVG_NS).each { |n| n.replace("<svg/>") }
-        doc
-      end
-
       def image_validate(doc)
         image_exists(doc)
         image_toobig(doc)
@@ -198,8 +155,7 @@ module Metanorma
 
       def validate(doc)
         content_validate(doc)
-        schema_validate(formattedstr_strip(doc.dup),
-                        File.join(File.dirname(__FILE__), "isodoc-compile.rng"))
+        schema_validate(formattedstr_strip(doc.dup), schema_location)
       end
 
       def repeat_id_validate1(elem)

data/lib/metanorma/standoc/validate_schema.rb

@@ -0,0 +1,104 @@
+require "jing"
+
+module Metanorma
+  module Standoc
+    module Validate
+      def schema_location
+        self.class.respond_to?(:_file) and ret = self.class::_file
+        ret ||= caller_locations(1..1).first.absolute_path
+        ret ||= __FILE__
+        File.join(File.dirname(ret), schema_file)
+      end
+
+      def schema_file
+        "isodoc-compile.rng"
+      end
+
+      def schema_validate(doc, schema)
+        Tempfile.open(["tmp", ".xml"], encoding: "UTF-8") do |f|
+          schema_validate1(f, doc, schema)
+        rescue Jing::Error => e
+          clean_abort("Jing failed with error: #{e}", doc)
+        ensure
+          f.close!
+        end
+      end
+
+      def schema_validate1(file, doc, schema)
+        file.write(to_xml(doc))
+        file.close
+        errors = Jing.new(schema, encoding: "UTF-8").validate(file.path)
+        warn "Syntax Valid!" if errors.none?
+        errors.each do |e|
+          @log.add("Metanorma XML Syntax",
+                   "XML Line #{'%06d' % e[:line]}:#{e[:column]}", e[:message])
+        end
+      end
+
+      def validate_document_fragment(xml_fragment)
+        f = add_ns_to_fragment(xml_fragment) or
+          return [true,
+                  "Fragment is not well-formed XML, not validating"]
+        begin
+          temp_schema, schema = fragment_schema(f.root.name)
+          schema or return [false, "Did not expect element #{f.root.name}"]
+          validation_errors = schema.validate(f)
+          [validation_errors.none? do |x|
+            x.to_s.include?("Did not expect element")
+          end, validation_errors]
+        ensure
+          temp_schema.unlink
+        end
+      end
+
+      def add_ns_to_fragment(xml_fragment)
+        f = Nokogiri::XML(xml_fragment, &:strict)
+        f.errors.any? || f.root.nil? and return nil
+        root_tag = f.root.name
+        f.root.namespace or
+          f = Nokogiri::XML(xml_fragment
+          .sub(/<#{root_tag}([^>]*)>/,
+               "<#{root_tag}\\1 xmlns='#{xml_namespace}'>"))
+        f
+      rescue StandardError
+        nil
+      end
+
+      def fragment_schema(root_element)
+        temp_schema = Tempfile.new(["dynamic_schema", ".rng"])
+        temp_schema.write(<<~SCHEMA)
+                  <grammar xmlns="http://relaxng.org/ns/structure/1.0">
+            <include href="#{schema_location}">
+              <start combine="choice">
+                  <ref name="#{root_element}"/>
+              </start>
+          </include>
+                  </grammar>
+        SCHEMA
+        temp_schema.close
+        [temp_schema, Nokogiri::XML::RelaxNG(File.open(temp_schema.path))]
+      rescue StandardError # error because root_element is not in schema
+        [temp_schema, nil]
+      end
+
+      SVG_NS = "http://www.w3.org/2000/svg".freeze
+
+      WILDCARD_ATTRS = "//stem | //metanorma-extension".freeze
+
+      # RelaxNG cannot cope well with wildcard attributes. So we strip
+      # any attributes from FormattedString instances (which can contain
+      # xs:any markup, and are signalled with @format) before validation.
+      def formattedstr_strip(doc)
+        doc.xpath(WILDCARD_ATTRS, "m" => SVG_NS).each do |n|
+          n.elements.each do |e|
+            e.traverse do |e1|
+              e1.element? and e1.each { |k, _v| e1.delete(k) } # rubocop:disable Style/HashEachMethods
+            end
+          end
+        end
+        doc.xpath("//m:svg", "m" => SVG_NS).each { |n| n.replace("<svg/>") }
+        doc
+      end
+    end
+  end
+end

data/lib/metanorma/standoc/version.rb

@@ -19,6 +19,6 @@ module Metanorma
   end
 
   module Standoc
-    VERSION = "3.0.3".freeze
+    VERSION = "3.0.5".freeze
   end
 end

data/metanorma-standoc.gemspec

@@ -32,10 +32,10 @@ Gem::Specification.new do |spec|
   spec.add_dependency "asciidoctor", "~> 2.0.0"
   spec.add_dependency "crass", "~> 1.0.0"
   spec.add_dependency "iev", "~> 0.3.5"
-  spec.add_dependency "isodoc", "~> 3.1.0"
+  spec.add_dependency "isodoc", "~> 3.1.4"
   spec.add_dependency "metanorma", ">= 1.6.0"
   spec.add_dependency "metanorma-plugin-datastruct", "~> 0.3.0"
-  spec.add_dependency "metanorma-plugin-glossarist", "~> 0.2.0"
+  spec.add_dependency "metanorma-plugin-glossarist", "~> 0.2.3"
   spec.add_dependency "metanorma-plugin-lutaml", "~> 0.7.0"
   spec.add_dependency "metanorma-utils", "~> 1.10.0"
   spec.add_dependency "ruby-jing"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: metanorma-standoc
 version: !ruby/object:Gem::Version
-  version: 3.0.3
+  version: 3.0.5
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-03-17 00:00:00.000000000 Z
+date: 2025-04-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: addressable
@@ -72,14 +72,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 3.1.0
+        version: 3.1.4
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 3.1.0
+        version: 3.1.4
 - !ruby/object:Gem::Dependency
   name: metanorma
   requirement: !ruby/object:Gem::Requirement
@@ -114,14 +114,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.2.0
+        version: 0.2.3
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.2.0
+        version: 0.2.3
 - !ruby/object:Gem::Dependency
   name: metanorma-plugin-lutaml
   requirement: !ruby/object:Gem::Requirement
@@ -434,7 +434,6 @@ files:
 - ".tex"
 - CODE_OF_CONDUCT.md
 - Gemfile
-- Gemfile.devel
 - LICENSE
 - Makefile
 - README.adoc
@@ -518,6 +517,7 @@ files:
 - lib/metanorma/standoc/cleanup_mathvariant.rb
 - lib/metanorma/standoc/cleanup_ref.rb
 - lib/metanorma/standoc/cleanup_reqt.rb
+- lib/metanorma/standoc/cleanup_review.rb
 - lib/metanorma/standoc/cleanup_section.rb
 - lib/metanorma/standoc/cleanup_section_names.rb
 - lib/metanorma/standoc/cleanup_symbols.rb
@@ -555,6 +555,7 @@ files:
 - lib/metanorma/standoc/ref_queue.rb
 - lib/metanorma/standoc/ref_sect.rb
 - lib/metanorma/standoc/ref_utility.rb
+- lib/metanorma/standoc/regex.rb
 - lib/metanorma/standoc/render.rb
 - lib/metanorma/standoc/reqt.rb
 - lib/metanorma/standoc/reqt.rng
@@ -566,6 +567,7 @@ files:
 - lib/metanorma/standoc/terms.rb
 - lib/metanorma/standoc/utils.rb
 - lib/metanorma/standoc/validate.rb
+- lib/metanorma/standoc/validate_schema.rb
 - lib/metanorma/standoc/validate_section.rb
 - lib/metanorma/standoc/validate_table.rb
 - lib/metanorma/standoc/validate_term.rb

data/Gemfile.devel

@@ -1 +0,0 @@
-gem "html2doc", git: "https://github.com/metanorma/html2doc", branch: "main"