metanorma-bsi 0.0.1

data/lib/asciidoctor/bsi/reqt.rng

@@ -0,0 +1,194 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<grammar xmlns="http://relaxng.org/ns/structure/1.0" datatypeLibrary="http://www.w3.org/2001/XMLSchema-datatypes">
+  <!--
+    Presupposes isodoc.rnc, is included in it
+    include "isodoc.rnc" { }
+  -->
+  <define name="requirement">
+    <element name="requirement">
+      <ref name="RequirementType"/>
+    </element>
+  </define>
+  <define name="recommendation">
+    <element name="recommendation">
+      <ref name="RequirementType"/>
+    </element>
+  </define>
+  <define name="permission">
+    <element name="permission">
+      <ref name="RequirementType"/>
+    </element>
+  </define>
+  <define name="RequirementType">
+    <optional>
+      <attribute name="obligation">
+        <ref name="ObligationType"/>
+      </attribute>
+    </optional>
+    <optional>
+      <attribute name="unnumbered">
+        <data type="boolean"/>
+      </attribute>
+    </optional>
+    <optional>
+      <attribute name="number"/>
+    </optional>
+    <optional>
+      <attribute name="subsequence"/>
+    </optional>
+    <optional>
+      <attribute name="keep-with-next">
+        <data type="boolean"/>
+      </attribute>
+    </optional>
+    <optional>
+      <attribute name="keep-lines-together">
+        <data type="boolean"/>
+      </attribute>
+    </optional>
+    <attribute name="id">
+      <data type="ID"/>
+    </attribute>
+    <optional>
+      <attribute name="filename"/>
+    </optional>
+    <optional>
+      <attribute name="model"/>
+    </optional>
+    <optional>
+      <attribute name="type"/>
+    </optional>
+    <optional>
+      <ref name="reqtitle"/>
+    </optional>
+    <optional>
+      <ref name="label"/>
+    </optional>
+    <optional>
+      <ref name="subject"/>
+    </optional>
+    <zeroOrMore>
+      <ref name="reqinherit"/>
+    </zeroOrMore>
+    <zeroOrMore>
+      <ref name="classification"/>
+    </zeroOrMore>
+    <zeroOrMore>
+      <choice>
+        <ref name="measurementtarget"/>
+        <ref name="specification"/>
+        <ref name="verification"/>
+        <ref name="import"/>
+        <ref name="description"/>
+      </choice>
+    </zeroOrMore>
+    <optional>
+      <ref name="reqt_references"/>
+    </optional>
+    <zeroOrMore>
+      <choice>
+        <ref name="requirement"/>
+        <ref name="recommendation"/>
+        <ref name="permission"/>
+      </choice>
+    </zeroOrMore>
+  </define>
+  <define name="reqtitle">
+    <element name="title">
+      <ref name="FormattedString"/>
+    </element>
+  </define>
+  <define name="label">
+    <element name="label">
+      <text/>
+    </element>
+  </define>
+  <define name="subject">
+    <element name="subject">
+      <text/>
+    </element>
+  </define>
+  <define name="reqinherit">
+    <element name="inherit">
+      <text/>
+    </element>
+  </define>
+  <define name="measurementtarget">
+    <element name="measurement-target">
+      <ref name="RequirementSubpart"/>
+    </element>
+  </define>
+  <define name="specification">
+    <element name="specification">
+      <ref name="RequirementSubpart"/>
+    </element>
+  </define>
+  <define name="verification">
+    <element name="verification">
+      <ref name="RequirementSubpart"/>
+    </element>
+  </define>
+  <define name="import">
+    <element name="import">
+      <ref name="RequirementSubpart"/>
+    </element>
+  </define>
+  <define name="description">
+    <element name="description">
+      <ref name="RequirementSubpart"/>
+    </element>
+  </define>
+  <define name="reqt_references">
+    <element name="references">
+      <oneOrMore>
+        <ref name="bibitem"/>
+      </oneOrMore>
+    </element>
+  </define>
+  <define name="RequirementSubpart">
+    <optional>
+      <attribute name="type"/>
+    </optional>
+    <optional>
+      <attribute name="exclude">
+        <data type="boolean"/>
+      </attribute>
+    </optional>
+    <optional>
+      <attribute name="keep-with-next">
+        <data type="boolean"/>
+      </attribute>
+    </optional>
+    <optional>
+      <attribute name="keep-lines-together">
+        <data type="boolean"/>
+      </attribute>
+    </optional>
+    <oneOrMore>
+      <ref name="BasicBlock"/>
+    </oneOrMore>
+  </define>
+  <define name="ObligationType">
+    <choice>
+      <value>requirement</value>
+      <value>recommendation</value>
+      <value>permission</value>
+    </choice>
+  </define>
+  <define name="classification">
+    <element name="classification">
+      <ref name="classification_tag"/>
+      <ref name="classification_value"/>
+    </element>
+  </define>
+  <define name="classification_tag">
+    <element name="tag">
+      <text/>
+    </element>
+  </define>
+  <define name="classification_value">
+    <element name="value">
+      <text/>
+    </element>
+  </define>
+</grammar>

data/lib/asciidoctor/bsi/validate.rb

@@ -0,0 +1,224 @@
+require_relative "./validate_requirement"
+require_relative "./validate_list"
+
+module Asciidoctor
+  module BSI
+    class Converter < ISO::Converter
+      # BSI 0:2016 9.4.1
+      def doctype_validate(xmldoc)
+        doctype = xmldoc&.at("//bibdata/ext/doctype")&.text
+        %w(specification management-systems-standard code-of-practice
+           guide method-of-test method-of-specifying vocabulary
+           classification).include? doctype or
+          @log.add("Document Attributes", nil,
+                   "#{doctype} is not a recognised document type")
+      end
+
+      def validate(doc)
+        content_validate(doc)
+        schema_validate(formattedstr_strip(doc.dup),
+                        File.join(File.dirname(__FILE__), "bsi.rng"))
+      end
+
+      def image_name_validate(xmldoc); end
+
+      def norm_ref_validate(doc); end
+
+      def style(node, text)
+        super
+        style_plusminus(node, text)
+        style_subscript(node)
+      end
+
+      # Rules for Structure 27.5
+      def style_plusminus(node, text)
+        style_regex(/\s*(?<num>\S+\u00b1\s*\S+)/,
+                    "no space before plus-minus sign", node, text)
+        style_regex(/\s*(?<num>\S+\s*\u00b1\s+\S+)/,
+                    "space after plus-minus sign", node, text)
+      end
+
+      # Rules for Structure 27.6
+      def style_subscript(node)
+        node.xpath(".//sub[.//sub]").each do |r|
+          style_warning(node, "Nested subscript", r)
+        end
+        node.xpath(".//m:msub[.//m:msub]", "m" => MATHML_NS).each do |r|
+          style_warning(node, "Nested subscript", r)
+        end
+        node.xpath(".//sup[.//sup]").each do |r|
+          style_warning(node, "Nested superscript", r)
+        end
+        node.xpath(".//m:msup[.//m:msup]", "m" => MATHML_NS).each do |r|
+          style_warning(node, "Nested superscript", r)
+        end
+      end
+
+      # protest decimal comma not decimal point
+      # TODO we will not yet override decimal comma warning if this is a National Annex to Eurocode (Rules for Structure Annex B)
+      def style_number(node, text)
+        style_two_regex_not_prev(
+          node, text, /^(?<num>-?[0-9]{4,}[,0-9]*)$/,
+          %r{\b(ISO|IEC|IEEE/|(in|January|February|March|April|May|June|August|September|October|November|December)\b)$},
+          "number not broken up in threes"
+        )
+        style_regex(/\b(?<num>[0-9]+,[0-9]+)/i,
+                    "possible decimal comma", node, text)
+        style_regex(/\b(?<num>billions?)\b/i,
+                    "ambiguous number", node, text)
+        ordinal_check(node)
+      end
+
+      # Rules for Structure 9.1
+      def ordinal_check(node)
+        node.xpath(".//sup").each do |s|
+          next unless %w(st nd rd th).include? s.text
+
+          style_warning(s, "text contains ordinal superscript", s.text)
+        end
+      end
+
+      def section_style(root)
+        super
+        @doctype = root.at("//bibdata/ext/doctype")&.text
+        @type = @doctype.gsub(/-/, " ").split.map(&:capitalize).join(" ")
+        informative_text_style(root)
+        normative_text_style(root)
+        avoid_must(root)
+        annex_validate(root)
+      end
+
+      def annex_validate(root)
+        annex_normative_validate(root)
+        annex_ordering_validate(root)
+      end
+
+      # Rules for Structure 20.2
+      def annex_normative_validate(root)
+        @doctype == "guide" and root.xpath("//annex[@obligation = 'normative']")
+          .each do |s|
+          style_warning(s, "Guide contains normative annex",
+                        s&.at("./title")&.text)
+        end
+      end
+
+      # Rules for Structure 20.4
+      def annex_ordering_validate(root)
+        ids = citation_order(root.xpath("//annex"))
+        cite_order = ids.keys.sort { |a, b| ids[a] <=> ids[b] }
+        annex_order = root.xpath("//annex").map { |x| x["id"] }
+        cite_order.each_with_index do |a, i|
+          next if annex_order[i] == a
+
+          annex = root.at("//annex[@id = '#{a}']")
+          title = annex&.at("./title")&.text || "Annex #{a}"
+          style_warning(annex, "#{title} should be annex ##{i + 1}", nil)
+        end
+      end
+
+      BSI_PUBLISHER_XPATH =
+        "./contributor[role/@type = 'publisher']/"\
+        "organization[abbreviation = 'ISO' or abbreviation = 'IEC' or "\
+        "abbreviation = 'BSI' or abbreviation = 'CEN' or "\
+        "abbreviation = 'CENELEC']".freeze
+
+      NORM_BSI_WARN = "reference other than BSI, CEN, CENELEC, ISO, IEC "\
+        "not expected as normative".freeze
+
+      # Rules for Structure 10.2
+      def norm_bibitem_style(root)
+        root.xpath(NORM_BIBITEMS).each do |b|
+          b.at(BSI_PUBLISHER_XPATH).nil? and
+            @log.add("Style", b, "#{NORM_BSI_WARN}: #{b.text}")
+          status = b&.at("./status/stage")&.text
+          /^9.$/.match?(status) and
+            @log.add("Style", b,
+                     "Do not cite withdrawn standards as normative: #{b.text}")
+          /^[0-5].$/.match?(status) and @log.add(
+            "Style", b,
+            "Do not cite unpublished standards as normative: #{b.text}"
+          )
+        end
+      end
+
+      # Rules for Structure 11.5.2
+      def title_names_type_validate(root)
+        super
+        doctypes = /British\sStandard/xi
+        title_main_en = title_lang_part(root, "main", "en")
+        !title_main_en.nil? && doctypes.match(title_main_en.text) and
+          @log.add("Style", title_main_en, "Main Title may name document type")
+        title_intro_en = title_lang_part(root, "intro", "en")
+        !title_intro_en.nil? && doctypes.match(title_intro_en.text) and
+          @log.add("Style", title_intro_en,
+                   "Title Intro may name document type")
+      end
+
+      # Rules for Structure 16.5.5
+      def termdef_style(xmldoc)
+        super
+        xmldoc.xpath("//term").each do |t|
+          para = t.at("./definition") || return
+          term = t.at("./preferred").text.strip
+          termdef_warn(para.text, /^[A-Z]/, t, term,
+                       "term definition starts with upper case character")
+          termdef_warn(para.text, /\b#{term.downcase}\b/i, t, term,
+                       "term definition may be circular")
+        end
+      end
+
+      def content_validate(doc)
+        super
+        symbol_style(doc)
+        table_validate(doc)
+        listcount_validate(doc)
+        list_punctuation(doc)
+      end
+
+      # Rules for Structure 17.5
+      def symbol_style(xmldoc)
+        return if @novalid
+
+        xmldoc.xpath("//definitions//dd").each do |para|
+          term = para&.at("./preceding::dt[1]")&.text&.strip || "[symbol]"
+          termdef_warn(para.text, /^(the|a)\b/i, para, term,
+                       "symbol definition starts with article")
+          termdef_warn(para.text, /\.$/i, para, term,
+                       "symbol definition ends with period")
+          termdef_warn(para.text, /^[A-Z]/, para, term,
+                       "symbol definition starts with upper case character")
+          termdef_warn(para.text, /\b#{term.downcase}\b/i, para, term,
+                       "symbol definition may be circular")
+        end
+      end
+
+      # Rules for Structure 22.3.2
+      def subclause_validate(root)
+        return if @novalid
+
+        root.xpath("//clause/clause/clause/clause/clause/clause")
+          .each do |c|
+          style_warning(c, "Exceeds the maximum clause depth of 5", nil)
+        end
+      end
+
+      # Rules for Structure 29.6
+      def table_validate(xmldoc)
+        xmldoc.xpath("//td[not(./*) and normalize-space(.)=''] | "\
+                     "//th[not(./*) and normalize-space(.)='']").each do |td|
+          style_warning(td, "Empty table cell", nil)
+        end
+      end
+
+      # TODO: we will not block foreword from National Annex to Eurocode (Rules for Structure G.6.1)
+      #
+      # TODO: we will not test location of commentaries (Rules for Structure 24.7)
+      #
+      # TODO: we will not check for space either side of mathematical sign (Rules for Structure 27.5)
+      #
+      # TODO: we will not check for conflict of ln and log_e or log lg log_10 (Rules for Structure 27.8)
+      #
+      # TODO: we will not check order of brackets in stem as {[( (Rules for Structure 27.9)
+    end
+  end
+end

data/lib/asciidoctor/bsi/validate_list.rb

@@ -0,0 +1,72 @@
+module Asciidoctor
+  module BSI
+    class Converter < ISO::Converter
+      # Rules for Structure 23.3
+      def listcount_validate(doc)
+        return if @novalid
+
+        doc.xpath("//clause[not(.//clause)] | //annex").each do |c|
+          next if c.xpath(".//ol").empty?
+
+          ols = c.xpath(".//ol") - c.xpath(".//ul//ol | .//ol//ol")
+          ols.size > 3 and
+            style_warning(c, "More than 3 ordered lists in a numbered clause",
+                          nil)
+        end
+      end
+
+      # Rules for Structure 23.5
+      def list_punctuation(doc)
+        return if @novalid
+
+        ((doc.xpath("//ol") - doc.xpath("//ul//ol | //ol//ol")) +
+         (doc.xpath("//ul") - doc.xpath("//ul//ul | //ol//ul"))).each do |list|
+          prec = list.previous_element
+          prec&.name == "p" or
+            style_warning(list, "All lists must be preceded by "\
+                          "introductory phrase", nil)
+          list_punctuation1(list, prec.text)
+        end
+      end
+
+      def list_punctuation1(list, prectext)
+        entries = list.xpath(".//li")
+        case prectext.sub(/^.*?(\S)\s*$/, "\\1")
+        when ":"
+          list.xpath(".//li").each_with_index do |li, i|
+            list_semicolon_phrase(li, i == entries.size - 1)
+          end
+        when "." then entries.each { |li| list_full_sentence(li) }
+        else style_warning(list, "All lists must be preceded by "\
+                          "colon or full stop", prectext)
+        end
+      end
+
+      def list_semicolon_phrase(elem, last)
+        text = elem.text.strip
+        text.match?(/^[^A-Za-z]*[a-z]/) or
+          style_warning(elem, "List entry after colon must start with "\
+                        "lowercase letter", text)
+        punct = text.sub(/^.*?(\S)\s*$/, "\\1")
+        if last
+          punct == "." or style_warning(elem, "Final list entry after colon "\
+                                        "must end with full stop", text)
+        else
+          punct == ";" or style_warning(elem, "List entry after colon must "\
+                                        "end with semicolon", text)
+        end
+      end
+
+      def list_full_sentence(elem)
+        text = elem.text.strip
+        text.match?(/^[^A-Za-z]*[A-Z]/) or
+          style_warning(elem, "List entry after full stop must start with "\
+                        "uppercase letter", text)
+        punct = text.sub(/^.*?(\S)\s*$/, "\\1")
+        punct == "." or
+          style_warning(elem, "List entry after full stop must "\
+                        "end with full stop", text)
+      end
+    end
+  end
+end