metanorma 2.0.6 → 2.0.9

data/lib/metanorma/collection/renderer/fileparse.rb

@@ -19,6 +19,7 @@ module Metanorma
         @nested || sso or
           Metanorma::Collection::XrefProcess::xref_process(xml, xml, nil, docid,
                                                            @isodoc)
+        @ncnames = {}
         @nested or update_indirect_refs_to_docs(xml, docid, internal_refs)
         @files.add_document_suffix(docid, xml)
         @nested or update_sectionsplit_refs_to_docs(xml, internal_refs)
@@ -49,9 +50,11 @@ module Metanorma
         end
       end
 
+      ANCHOR_XPATH = "xmlns:locality[@type = 'anchor']/xmlns:referenceFrom"
+        .freeze
+
       def update_sectionsplit_eref_to_doc(eref, internal_refs, doclist, opts)
-        a = eref.at(ns("./localityStack/locality[@type = 'anchor']/" \
-                       "referenceFrom")) or return
+        a = eref.at("./xmlns:localityStack/#{ANCHOR_XPATH}") or return
         doc = internal_refs[opts[:key]]["#{a.text}_#{opts[:target_suffix]}"]
         bibitemid = Metanorma::Utils::to_ncname("#{doc}_#{opts[:source_suffix]}")
         eref["bibitemid"] = bibitemid
@@ -59,12 +62,6 @@ module Metanorma
         doclist
       end
 
-      def eref2link(docxml)
-        isodoc = IsoDoc::PresentationXMLConvert.new({})
-        isodoc.bibitem_lookup(docxml)
-        isodoc.eref2link(docxml)
-      end
-
       BIBITEM_NOT_REPO_XPATH = "//bibitem[not(ancestor::bibitem)]" \
         "[not(./docidentifier[@type = 'repository'])]".freeze
 
@@ -79,64 +76,37 @@ module Metanorma
         end
       end
 
-      def svg_datauri(docxml, docid)
-        rel = @files.get(docid, :rel_path)
-        parent = @files.get(docid, :parentid) and
-          rel = @files.get(parent, :rel_path)
-        # if sectionsplit, use orig file dir
-        dir = File.join(@dirname, File.dirname(rel))
-        datauri_encode(docxml, dir)
-      end
-
-      def svgmap_resolve(docxml, docid)
-        ids = @files.get(docid, :ids)
-        docxml = svg_unnest(svg_datauri(docxml, docid))
-        isodoc = IsoDoc::PresentationXMLConvert.new({})
-        isodoc.bibitem_lookup(docxml)
-        docxml.xpath(ns("//svgmap//eref")).each do |e|
-          svgmap_resolve_eref(e, isodoc, docxml, ids)
-        end
-        Vectory::SvgMapping.new(docxml, "").call
-        docxml.xpath(ns("//svgmap")).each { |s| isodoc.svgmap_extract(s) }
-      end
-
-      def svg_unnest(docxml)
-        docxml.xpath(ns("//svgmap//image[.//*[name() = 'image']]")).each do |i|
-          s = i.elements.detect { |e| e.name == "svg" } and
-            i.replace(s)
-        end
-        docxml
-      end
-
-      def svgmap_resolve_eref(eref, isodoc, _docxml, ids)
-        href = isodoc.eref_target(eref) or return
-        href = href[:link]
-        href == "##{eref['bibitemid']}" ||
-          (href =~ /^#/ && !ids[href.sub(/^#/, "")]) and return
-        eref["target"] = href.strip
-        eref.name = "link"
-        eref.elements&.remove
-      end
-
       # repo(current-metanorma-collection/ISO 17301-1:2016)
       # replaced by bibdata of "ISO 17301-1:2016" in situ as bibitem.
       # Any erefs to that bibitem id are replaced with relative URL
       # Preferably with anchor, and is a job to realise dynamic lookup
       # of localities.
       def update_direct_refs_to_docs(docxml, identifier)
-        erefs, erefs1 = update_direct_refs_to_docs_prep(docxml)
+        erefs, erefs_no_anchor, anchors, erefs1 =
+          update_direct_refs_to_docs_prep(docxml)
         docxml.xpath(ns("//bibitem")).each do |b|
           docid = b.at(ns("./docidentifier[@type = 'repository']")) or next
           strip_unresolved_repo_erefs(identifier, docid, erefs1, b) or next
           update_bibitem(b, identifier)
           docid = docid_to_citeas(b) or next
-          erefs[docid] and update_anchors(b, docid, erefs[docid])
+          erefs[docid] and
+            update_anchors(b, docid, erefs[docid], erefs_no_anchor[docid],
+                           anchors[docid])
         end
       end
 
+      # Hash(docid) of arrays
       def update_direct_refs_to_docs_prep(docxml)
-        @ncnames = {}
-        [Util::gather_citeases(docxml), Util::gather_bibitemids(docxml)]
+        erefs = Util::gather_citeases(docxml)
+        no_anchor = erefs.keys.each_with_object({}) { |k, m| m[k] = [] }
+        anchors = erefs.keys.each_with_object({}) { |k, m| m[k] = [] }
+        erefs.each do |k, v|
+          v.each do |e|
+            if loc = e.at(".//#{ANCHOR_XPATH}") then anchors[k] << loc
+            else no_anchor[k] << e end
+          end
+        end
+        [erefs, no_anchor, anchors, Util::gather_bibitemids(docxml)]
       end
 
       # strip erefs if they are repository erefs, but do not point to a document
@@ -155,56 +125,78 @@ module Metanorma
       # Resolve erefs to a container of ids in another doc,
       # to an anchor eref (direct link)
       def update_indirect_refs_to_docs(docxml, _docidentifier, internal_refs)
-        bibitems, erefs = update_indirect_refs_to_docs_prep(docxml)
+        bib, erefs, doc_suffix, doc_type, f = update_indirect_refs_prep(docxml)
         internal_refs.each do |schema, ids|
-          s = "#{schema}_"
+          add_suffix = doc_suffix && doc_type && doc_type != schema
           ids.each do |id, file|
-            k = indirect_ref_key(s, schema, id, docxml)
-            update_indirect_refs_to_docs1(docxml, k, file, bibitems, erefs)
+            f[file] ||= { url: @files.url?(file),
+                          parentid: @files.get(file) && @files.get(file,
+                                                                   :parentid) }
+            k = indirect_ref_key(schema, id, doc_suffix, add_suffix)
+            update_indirect_refs_to_docs1(f[file], k, file, bib, erefs)
           end
         end
       end
 
-      def update_indirect_refs_to_docs_prep(docxml)
+      def update_indirect_refs_prep(docxml)
         @updated_anchors = {}
-        [Util::gather_bibitems(docxml), Util::gather_bibitemids(docxml)]
+        @indirect_keys = {}
+        [Util::gather_bibitems(docxml), Util::gather_bibitemids(docxml),
+         docxml.root["document_suffix"], docxml.root["type"], {}]
+      end
+
+      def indirect_ref_key(schema, id, doc_suffix, doc_type)
+        /^#{schema}_/.match?(id) and return id
+        key = [schema, id, doc_suffix, doc_type].join("::")
+        x = @indirect_keys[key] and return x
+        ret = "#{schema}_#{id}"
+        doc_suffix && doc_type && doc_type != schema and
+          ret = "#{ret}_#{doc_suffix}"
+        @indirect_keys[key] = ret
+        ret
       end
 
-      def indirect_ref_key(schema_, schema, id, docxml)
-        /^#{schema_}/.match?(id) and return id
-        ret = schema_ + id
-        suffix = docxml.root["document_suffix"] or return ret
-        (k = docxml.root["type"]) && k != schema or return ret
-        "#{ret}_#{suffix}"
+      def indirect_ref_key(schema, id, doc_suffix, add_suffix)
+        /^#{schema}_/.match?(id) and return id
+        #key = "#{schema}_#{id}"
+        x = @indirect_keys.dig(schema, id) and return x
+        @indirect_keys[schema] ||= {}
+        @indirect_keys[schema][id] = if add_suffix
+                                  "#{schema}_#{id}_#{doc_suffix}"
+                                else
+                                   "#{schema}_#{id}"
+                                end
       end
 
-      #OLD
-      def indirect_ref_key1(schema, id, docxml)
+      def indirect_ref_keyx(schema, id, doc_suffix, doc_type)
         /^#{schema}_/.match?(id) and return id
         ret = "#{schema}_#{id}"
-        suffix = docxml.root["document_suffix"]
-        (k = docxml.root["type"]) && k != schema && suffix and
-          ret = "#{ret}_#{suffix}"
+        doc_suffix && doc_type && doc_type != schema and
+          ret = "#{ret}_#{doc_suffix}"
         ret
       end
 
-      def update_indirect_refs_to_docs1(_docxml, key, file, bibitems, erefs)
+      def update_indirect_refs_to_docs1(filec, key, file, bibitems, erefs)
         erefs[key]&.each do |e|
           e["citeas"] = file
-          update_indirect_refs_to_docs_anchor(e, file)
+          update_indirect_refs_to_docs_anchor(e, file, filec[:url],
+                                              filec[:parentid])
         end
         update_indirect_refs_to_docs_docid(bibitems[key], file)
       end
 
-      def update_indirect_refs_to_docs_anchor(eref, file)
-        a = eref.at(ns(".//locality[@type = 'anchor']/referenceFrom")) or return
-        suffix = file
-        @files.get(file) && p = @files.get(file, :parentid) and
-          suffix = "#{p}_#{suffix}"
+      def update_indirect_refs_to_docs_anchor(eref, file, url, parentid)
+        a = eref.at(".//#{ANCHOR_XPATH}") or return
+        parentid and file = "#{parentid}_#{file}"
         existing = a.text
-        anchor = existing
-        @files.url?(file) or
-          anchor = Metanorma::Utils::to_ncname("#{anchor}_#{suffix}")
+        anchor = if url then existing
+                   else 
+                     #suffix_anchor_indirect(existing, suffix)
+                     #k = "#{existing}_#{file}"
+        #@ncnames[k] ||= Metanorma::Utils::to_ncname(k)
+        @indirect_keys[existing] ||= {}
+        @indirect_keys[existing][file] ||= Metanorma::Utils::to_ncname("#{existing}_#{file}")
+                   end
         @updated_anchors[existing] or a.children = anchor
         @updated_anchors[anchor] = true
       end
@@ -216,55 +208,36 @@ module Metanorma
           "<docidentifier type='metanorma-collection'>#{file}</docidentifier>"
       end
 
-      # update crossrefences to other documents, to include
-      # disambiguating document suffix on id
-      def update_anchors(bib, docid, erefs)
-        erefs.each do |e|
-          if @files.get(docid) then update_anchor_loc(bib, e, docid)
-          else
-            msg = "<strong>** Unresolved reference to document #{docid} " \
-                  "from eref</strong>"
-            e << msg
-            strip_eref(e)
-            @log&.add("Cross-References", e, msg)
-          end
+      # bottleneck
+      def update_anchors(bib, docid, erefs, erefs_no_anchor, erefs_anchors)
+        @files.get(docid) or error_anchor(erefs, docid)
+        has_anchors, url, ncn_docid = update_anchors_prep(docid)
+        erefs_no_anchor.each do |e|
+          update_anchor_create_loc(bib, e, docid)
+        end
+        !url && has_anchors or return
+        erefs_anchors.each do |e|
+          update_anchors1(docid, ncn_docid, e)
         end
       end
 
-      def update_anchor_loc(bib, eref, docid)
-        loc = eref.at(".//xmlns:locality[@type = 'anchor']") or
-          return update_anchor_create_loc(bib, eref, docid)
-        a = @files.get(docid, :anchors) or return
-        ref = loc.elements&.first or return
-        anchor = suffix_anchor(ref.text, docid)
-        a.values.detect { |x| x.value?(anchor) } or return
-        ref.content = anchor
-      end
-
-      #OLD
-      def update_anchor_loc1(bib, eref, docid)
-        loc = eref.at(".//xmlns:locality[@type = 'anchor']") or
-          return update_anchor_create_loc(bib, eref, docid)
-        ref = loc.at("./xmlns:referenceFrom") or return
-        anchor = suffix_anchor(ref, docid)
-        a = @files.get(docid, :anchors) or return
-        a.inject([]) { |m, (_, x)| m + x.values }
-          .include?(anchor) or return
-        ref.content = anchor
+      def update_anchors1(docid, ncn_docid, anchor)
+        @concat_anchors[anchor.text] ||= "#{ncn_docid}_#{anchor.text}"
+        if @files.get(docid).dig(:anchors_lookup, @concat_anchors[anchor.text])
+          anchor.content = @concat_anchors[anchor.text]
+        end
       end
 
-      def suffix_anchor(ref, docid)
-        @ncnames[docid] ||= "#{Metanorma::Utils::to_ncname(docid)}_"
-        @files.url?(docid) or return ref
-        @ncnames[docid] + ref
+      def update_anchors_prep(docid)
+        @concat_anchors = {}
+        [@files.get(docid)&.key?(:anchors_lookup), @files.url?(docid),
+         Metanorma::Utils::to_ncname(docid)]
       end
 
-      #OLD
-       def suffix_anchor1(ref, docid)
-        @ncnames[docid] ||= Metanorma::Utils::to_ncname(docid)
-        anchor = ref.text
-        @files.url?(docid) or anchor = "#{@ncnames[docid]}_#{anchor}"
-        anchor
+      # encode both prefix and suffix to NCName
+      def suffix_anchor_indirect(prefix, suffix)
+        k = "#{prefix}_#{suffix}"
+        @ncnames[k] ||= Metanorma::Utils::to_ncname(k)
       end
 
       # if there is a crossref to another document, with no anchor, retrieve the
@@ -276,7 +249,7 @@ module Metanorma
         ref = ins.at(ns("./locality/referenceFrom"))&.text
         a = @files.get(docid, :anchors).dig(type, ref) or return
         ins << "<locality type='anchor'><referenceFrom>#{a.sub(/^_/, '')}" \
-               "</referenceFrom></locality>"
+          "</referenceFrom></locality>"
       end
     end
   end

data/lib/metanorma/collection/renderer/navigation.rb

@@ -114,7 +114,7 @@ module Metanorma
             files.each { |f| docrefs(f, u) }
           end
         end
-        r.doc.root&.to_html&.gsub("\n", " ")
+        r.doc.root&.to_html&.tr("\n", " ")
       end
 
       def liquid_docrefs(mnfs)

data/lib/metanorma/collection/renderer/renderer.rb

@@ -8,6 +8,7 @@ require_relative "../multilingual/multilingual"
 require_relative "utils"
 require_relative "render_word"
 require_relative "navigation"
+require_relative "svg"
 
 module Metanorma
   class Collection
@@ -42,7 +43,8 @@ module Metanorma
         @compile = Compile.new
         @compile.load_flavor(@flavor)
 
-        @isodoc = isodoc_create # output processor for flavour
+        # output processor for flavour
+        @isodoc = Util::isodoc_create(@flavor, @lang, @scropt, @xml)
         @outdir = dir_name_cleanse(options[:output_folder])
         @coverpage = options[:coverpage] || collection.coverpage
         @format = ::Metanorma::Util.sort_extensions_execution(options[:format])
@@ -74,6 +76,7 @@ module Metanorma
 
       def flush_files
         warn "\n\n\n\n\nDone: #{DateTime.now.strftime('%H:%M:%S')}"
+        warn "\nFiles to delete:\n"
         warn @files.files_to_delete
         @files.files_to_delete.each { |f| FileUtils.rm_f(f) }
         @files_to_delete.each { |f| FileUtils.rm_f(f) }

data/lib/metanorma/collection/renderer/svg.rb

@@ -0,0 +1,44 @@
+module Metanorma
+  class Collection
+    class Renderer
+      def svg_datauri(docxml, docid)
+        rel = @files.get(docid, :rel_path)
+        parent = @files.get(docid, :parentid) and
+          rel = @files.get(parent, :rel_path)
+        # if sectionsplit, use orig file dir
+        dir = File.join(@dirname, File.dirname(rel))
+        datauri_encode(docxml, dir)
+      end
+
+      def svgmap_resolve(docxml, docid)
+        ids = @files.get(docid, :ids)
+        docxml = svg_unnest(svg_datauri(docxml, docid))
+        isodoc = IsoDoc::PresentationXMLConvert.new({})
+        isodoc.bibitem_lookup(docxml)
+        docxml.xpath(ns("//svgmap//eref")).each do |e|
+          svgmap_resolve_eref(e, isodoc, docxml, ids)
+        end
+        Vectory::SvgMapping.new(docxml, "").call
+        docxml.xpath(ns("//svgmap")).each { |s| isodoc.svgmap_extract(s) }
+      end
+
+      def svg_unnest(docxml)
+        docxml.xpath(ns("//svgmap//image[.//*[name() = 'image']]")).each do |i|
+          s = i.elements.detect { |e| e.name == "svg" } and
+            i.replace(s)
+        end
+        docxml
+      end
+
+      def svgmap_resolve_eref(eref, isodoc, _docxml, ids)
+        href = isodoc.eref_target(eref) or return
+        href = href[:link]
+        href == "##{eref['bibitemid']}" ||
+          (href =~ /^#/ && !ids[href.sub(/^#/, "")]) and return
+        eref["target"] = href.strip
+        eref.name = "link"
+        eref.elements&.remove
+      end
+    end
+  end
+end

data/lib/metanorma/collection/renderer/utils.rb

@@ -55,7 +55,7 @@ module Metanorma
       def docid_to_citeas(bib)
         docid = bib.at(ns("./docidentifier[@primary = 'true']")) ||
           bib.at(ns("./docidentifier")) or return
-        docid_prefix(docid)
+        ::Metanorma::Collection::Util::key(docid_prefix(docid))
       end
 
       def collect_erefs(docxml)
@@ -168,21 +168,13 @@ module Metanorma
         end
       end
 
-      def isodoc_create
-        isodoc = Util::load_isodoc(@flavor)
-        isodoc.i18n_init(@lang, @script, @locale) # read in internationalisation
-        isodoc.metadata_init(@lang, @script, @locale, isodoc.i18n)
-        isodoc.info(@xml, nil)
-        isodoc
-      end
-
       # create the @meta class of isodoc, for populating Liquid,
       # with "navigation" set to the index bar.
       # extracted from the manifest
       def isodoc_populate
         @isodoc.info(@xml, nil)
         { navigation: indexfile(@manifest), nav_object: index_object(@manifest),
-          docrefs: liquid_docrefs(@manifest),
+          bibdata: @bibdata.to_hash, docrefs: liquid_docrefs(@manifest),
           "prefatory-content": isodoc_builder(@xml.at("//prefatory-content")),
           "final-content": isodoc_builder(@xml.at("//final-content")),
           doctitle: @bibdata.title.first.title.content,
@@ -203,6 +195,22 @@ module Metanorma
         end.doc.root.to_html
       end
 
+      def eref2link(docxml)
+        isodoc = IsoDoc::PresentationXMLConvert.new({})
+        isodoc.bibitem_lookup(docxml)
+        isodoc.eref2link(docxml)
+      end
+
+      def error_anchor(erefs, docid)
+        erefs.each do |e|
+          msg = "<strong>** Unresolved reference to document #{docid} " \
+            "from eref</strong>"
+          e << msg
+          strip_eref(e)
+          @log&.add("Cross-References", e, msg)
+        end
+      end
+
       def ns(xpath)
         @isodoc.ns(xpath)
       end

data/lib/metanorma/collection/sectionsplit/collection.rb

@@ -76,22 +76,22 @@ module Metanorma
         File.basename(ret)
       end
 
-      def section_split_cover(col, ident, _one_doc_coll)
+      def section_split_cover(col, ident, one_doc_coll)
         dir = File.dirname(col.file)
         collection_setup(nil, dir)
         r = ::Metanorma::Collection::Renderer
           .new(col, dir, output_folder: "#{ident}_collection",
-                         format: %i(html), coverpage: File.join(dir, "cover.html"))
+                         format: %i(html),
+                         coverpage: File.join(dir, "cover.html"))
         r.coverpage
-        section_split_cover1(ident, r, dir, _one_doc_coll)
+        section_split_cover1(ident, r, dir, one_doc_coll)
       end
 
       def section_split_cover1(ident, renderer, dir, _one_doc_coll)
-        # filename = one_doc_coll ? "#{ident}_index.html" : "index.html"
         filename = File.basename("#{ident}_index.html")
         # ident can be a directory with YAML indirection
-        FileUtils.mv File.join(renderer.outdir, "index.html"),
-                     File.join(dir, filename)
+        dest = File.join(dir, filename)
+        FileUtils.mv File.join(renderer.outdir, "index.html"), dest
         FileUtils.rm_rf renderer.outdir
         filename
       end

data/lib/metanorma/collection/sectionsplit/sectionsplit.rb

@@ -2,6 +2,7 @@ require "yaml"
 require_relative "../../util/util"
 require_relative "../xrefprocess/xrefprocess"
 require_relative "collection"
+require "concurrent-ruby"
 
 module Metanorma
   class Collection
@@ -35,12 +36,32 @@ module Metanorma
       def sectionsplit
         xml = sectionsplit_prep(File.read(@input_filename), @base, @dir)
         @key = Metanorma::Collection::XrefProcess::xref_preprocess(xml, @isodoc)
-        SPLITSECTIONS.each_with_object([]) do |n, ret|
+        empty = empty_doc(xml)
+        empty1 = empty_attachments(empty)
+        @mutex = Mutex.new
+        @pool = Concurrent::FixedThreadPool.new(4)
+        sectionsplit1(xml, empty, empty1, 0)
+      end
+
+      def sectionsplit1(xml, empty, empty1, idx)
+        ret = SPLITSECTIONS.each_with_object([]) do |n, m|
           conflate_floatingtitles(xml.xpath(ns(n[0]))).each do |s|
-            ret << sectionfile(xml, emptydoc(xml, ret.size),
-                               "#{@base}.#{ret.size}", s, n[1])
+            sectionsplit2(xml, idx.zero? ? empty : empty1, s, n[1],
+                          { acc: m, idx: idx })
+            idx += 1
           end
         end
+        @pool.shutdown
+        @pool.wait_for_termination
+        ret
+      end
+
+      def sectionsplit2(xml, empty, chunks, parentnode, opt)
+        @pool.post do
+          a = sectionfile(xml, empty, "#{@base}.#{opt[:idx]}", chunks,
+                          parentnode)
+          @mutex.synchronize { opt[:acc] << a }
+        end
       end
 
       def block?(node)
@@ -61,13 +82,13 @@ module Metanorma
 
       def sectionsplit_prep(file, filename, dir)
         @splitdir = dir
-        xml1, type = sectionsplit_preprocess_semxml(file, filename)
+        xml, type = sectionsplit_preprocess_semxml(file, filename)
         flags = { format: :asciidoc, extension_keys: [:presentation],
                   type: type }.merge(@compile_opts)
-        Compile.new.compile(xml1, flags)
-        f = File.open(xml1.sub(/\.xml$/, ".presentation.xml"),
-                      encoding: "utf-8")
+        Compile.new.compile(xml, flags)
+        f = File.open(xml.sub(/\.xml$/, ".presentation.xml"), encoding: "utf-8")
         r = Nokogiri::XML(f, &:huge)
+        f.close
         r.xpath("//xmlns:svgmap1").each { |x| x.name = "svgmap" }
         r
       end
@@ -108,7 +129,25 @@ module Metanorma
              "//indexsect | //colophon"),
         ).each(&:remove)
         ordinal.zero? or out.xpath(ns("//metanorma-ext//attachment | " \
-                      "//semantic__metanorma-ext//semantic__attachment"))
+                                      "//semantic__metanorma-ext//semantic__attachment"))
+          .each(&:remove) # keep only one copy of attachments
+        out
+      end
+
+      def empty_doc(xml)
+        out = xml.dup
+        out.xpath(
+          ns("//preface | //sections | //annex | //bibliography/clause | " \
+             "//bibliography/references[not(@hidden = 'true')] | " \
+             "//indexsect | //colophon"),
+        ).each(&:remove)
+        out
+      end
+
+      def empty_attachments(xml)
+        out = xml.dup
+        out.xpath(ns("//metanorma-ext//attachment | " \
+                     "//semantic__metanorma-ext//semantic__attachment"))
           .each(&:remove) # keep only one copy of attachments
         out
       end
@@ -126,7 +165,9 @@ module Metanorma
                                                          @ident, @isodoc)
         truncate_semxml(out, chunks)
         outname = "#{file}.xml"
-        File.open(File.join(@splitdir, outname), "w:UTF-8") { |f| f.write(out) }
+        File.open(File.join(@splitdir, outname), "w:UTF-8") do |f|
+          f.write(out)
+        end
         outname
       end
 
@@ -161,13 +202,11 @@ module Metanorma
       end
 
       def titlerender(section)
-        title = section.at(ns("./title")) or return "[Untitled]"
+        title = section.at(ns("./fmt-title")) or return "[Untitled]"
         t = title.dup
         t.xpath(ns(".//tab | .//br")).each { |x| x.replace(" ") }
         t.xpath(ns(".//bookmark")).each(&:remove)
-        t.xpath(ns(".//strong | .//span"))
-          .each { |x| x.replace(x.children) }
-        t.children.to_xml
+        t.xpath('.//text()').map(&:text).join
       end
     end
   end

data/lib/metanorma/collection/util/util.rb

@@ -25,8 +25,9 @@ module Metanorma
         def gather_citeases(xml)
           xml.xpath("//*[@citeas]").each_with_object({}) do |e, m|
             /^semantic__/.match?(e.name) and next
-            m[e["citeas"]] ||= []
-            m[e["citeas"]] << e
+            k = key(e["citeas"])
+            m[k] ||= []
+            m[k] << e
           end
         end
 
@@ -61,10 +62,23 @@ module Metanorma
           def attr(_key); end
         end
 
-        def load_isodoc(flavor)
+        def load_isodoc(flavor, presxml: false)
           x = Asciidoctor.load nil, backend: flavor.to_sym
+          if presxml
+          x.converter.presentation_xml_converter(Dummy.new)
+          else
           x.converter.html_converter(Dummy.new) # to obtain Isodoc class
+          end
         end
+
+        def isodoc_create(flavor, lang, script, xml, presxml: false)
+        isodoc = Util::load_isodoc(flavor, presxml: presxml)
+        isodoc.i18n_init(lang, script, nil) # read in internationalisation
+        # TODO locale?
+        isodoc.metadata_init(lang, script, nil, isodoc.i18n)
+        isodoc.info(xml, nil)
+        isodoc
+      end
       end
     end
   end

data/lib/metanorma/compile/compile.rb

@@ -114,6 +114,14 @@ module Metanorma
       fnames = { xml: f.sub(/\.[^.]+$/, ".xml"), f: f,
                  orig_filename: File.expand_path(filename),
                  presentationxml: f.sub(/\.[^.]+$/, ".presentation.xml") }
+      if extensions == %i(presentation)
+        process_ext(:presentation, file, isodoc, fnames, options)
+      else
+      process_exts_queue(fnames, extensions, file, isodoc, options)
+      end
+    end
+
+    def process_exts_queue(fnames, extensions, file, isodoc, options)
       @queue = ::Metanorma::Util::WorkersPool
         .new(ENV["METANORMA_PARALLEL"]&.to_i || 3)
       gather_and_install_fonts(file, options.dup, extensions)

data/lib/metanorma/compile/extract.rb

@@ -1,8 +1,7 @@
 module Metanorma
   class Compile
     def relaton_export(isodoc, options)
-      return unless options[:relaton]
-
+      options[:relaton] or return
       xml = Nokogiri::XML(isodoc, &:huge)
       bibdata = xml.at("//bibdata") || xml.at("//xmlns:bibdata")
       # docid = bibdata&.at("./xmlns:docidentifier")&.text || options[:filename]
@@ -18,8 +17,7 @@ module Metanorma
     end
 
     def extract(isodoc, dirname, extract_types)
-      return unless dirname
-
+      dirname or return
       extract_types.nil? || extract_types.empty? and
         extract_types = %i[sourcecode image requirement]
       FileUtils.rm_rf dirname