Merge pull request #1160 from metanorma/fix/linebreak-empty-xref

fix behaviour of carriage returns before empty eref, xref, link in te…

Author: opoudjis

lib/metanorma/cleanup/text.rb

@@ -3,7 +3,7 @@ module Standoc
     module Text
       def ancestor_include?(elem, ancestors)
         path = elem.path.gsub(/\[\d+\]/, "").split(%r{/})[1..-2]
-        !path.intersection(ancestors).empty?
+        path.intersect?(ancestors)
       end
 
       # process example/p, example/sourcecode, not example on its own:
@@ -30,29 +30,41 @@ def linebreak_cleanup_block(block)
       end
 
       def lines_strip_textspan(span, nextspan)
-        lines = span[:text].lines[0..-2].map(&:rstrip) <<
-          span[:text].lines[-1]&.sub(/\n$/, "")
+        #AAA
+        lines = []
+        span[:text] and
+          lines = span[:text].lines[0..-2].map(&:rstrip) <<
+            span[:text].lines[-1]&.sub(/\n$/, "")
+        #lines = span[:text].lines[0..-2].map(&:rstrip) <<
+  #span[:text].lines[-1]&.sub(/\n$/, "")
         # no final line rstrip: can be space linking to next line
         span[:last] or lines << nextspan[:text].lines.first # next token context
         lines
       end
 
-      # TODO: we are not counting empty xref, eref here
       def gather_text_for_linebreak_cleanup(block)
-        x = block.xpath(".//text()").map do |e|
-          { elem: e, text: e.text, stem: ancestor_include?(e, %w(stem)),
-            skip: ancestor_include?(e, PRESERVE_LINEBREAK_ELEMENTS) }
-        end
+        x = gather_text_for_linebreak_cleanup1(block)
         x.empty? and return x
-        x.each { |e| e[:skip] ||= !e[:text].include?("\n") }
         x.each_with_index do |e, i|
+          e[:skip] ||= !e[:text].include?("\n")
           # do not treat stem linebreaks as meaningful
           e[:skip] ||= x[i + 1]&.dig(:stem)
+          e[:skip] ||= !e[:elem].text?
         end
         x[-1][:last] = true
         x
       end
 
+      def gather_text_for_linebreak_cleanup1(block)
+        block.xpath(".//text() | .//eref[not(text())] |  " \
+                        ".//xref[not(text())] | .//termref[not(text())] | " \
+                        ".//link[not(text())] ").map do |e|
+                          #x = block.xpath(".//text()").map do |e|
+          { elem: e, text: e.text, stem: ancestor_include?(e, %w(stem)),
+            skip: ancestor_include?(e, PRESERVE_LINEBREAK_ELEMENTS) }
+        end
+      end
+
       def smartquotes_cleanup(xmldoc)
         xmldoc.xpath("//date").each { |d| Metanorma::Utils::endash_date(d) }
         if @smartquotes then smartquotes_cleanup1(xmldoc)
@@ -79,14 +91,16 @@ def uninterrupt_quotes_around_xml(xmldoc)
         end
       end
 
-      # "abc<tag/>", def => "abc",<tag/> def
-      # TODO?
-      def uninterrupt_quotes_around_xml1(xmldoc)
-        xmldoc.xpath("//text()[preceding-sibling::*[1]]").each do |n|
-          uninterrupt_quotes_around_xml_skip(n) and next
-          uninterrupt_quotes_around_xml1(n.previous)
-        end
-      end
+#AAA
+# "abc<tag/>", def => "abc",<tag/> def
+# TODO?
+#def uninterrupt_quotes_around_xml1(xmldoc)
+  #xmldoc.xpath("//text()[preceding-sibling::*[1]]").each do |n|
+    #uninterrupt_quotes_around_xml_skip(n) and next
+    #uninterrupt_quotes_around_xml1(n.previous)
+  #end
+#end
+
 
       IGNORE_QUOTES_ELEMENTS =
         %w(pre tt sourcecode stem asciimath figure bibdata passthrough
@@ -111,6 +125,7 @@ def uninterrupt_quotes_around_xml_skip(elem)
            ignoretext?(elem.previous)))
       end
 
+      # "abc<tag/>", def => "abc",<tag/> def
       def uninterrupt_quotes_around_xml1(elem)
         prev = elem.at(".//preceding::text()[1]") or return
         /\S\Z/.match?(prev.text) or return

spec/cleanup/blocks_spec.rb

@@ -1438,6 +1438,9 @@
       ABC
       *JSA*)
 
+      ABC
+      link:example.com[]
+
     INPUT
     output = <<~OUTPUT
       <sections><clause id="_" type="scope" inline-header="false" obligation="normative">
@@ -1461,6 +1464,8 @@
        <p id="_">ABC (<strong>JSA</strong>)</p>
 
        <p id="_">ABC <strong>JSA</strong>)</p>
+
+       <p id="_">ABC <link target="example.com"/></p>
       </clause>
       </sections>
     OUTPUT