metanorma-utils 1.9.0 → 1.10.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 717014b944b65d1e4f5b7c7ed0c71b3ac3314a3fb9ac740ff0027e9b5088cea8
-  data.tar.gz: c5bc40281e230a272deaf0469d4b9b6dd85e4f6b124de1e64d4c5c16347b2e8c
+  metadata.gz: fe2af0c9b9521b3ae1953545fb434d08e45d028dd93c02fa509a8251e14e1c8e
+  data.tar.gz: c551aa4d4c08dba83273ce0be945a171cfbbdc4a9a2c02fd09044b3c97a8b317
 SHA512:
-  metadata.gz: ef77c94f8be1071eead01e24fdbdbf1367ea2fc19ede7f1b3bf0149127863e38d56fea39a42a4685b3c65112431c251fab1507602a54b0307dae108a7ab563c9
-  data.tar.gz: 6a86994cc199526918abbb80cb79f0c099a3aa2eb3480b4d763e381a351cea0184ccce285eef28b49cb9d441037469ea4f21df91258a3b09c64bc889c76bf80f
+  metadata.gz: f6a23dde225425211a6ac35bb0e9b718189f0bc071eeef526a36ecebd3f8572f78addf7b93c4390374511642c360b6543e88e98e459ea1bb76849d077e008730
+  data.tar.gz: b537b6828491a4cc6f495c462e0c0810a8b1fcd61effb326232980f07325fe70e2ddc6b9f3c09d60b1f8c6af250c38e21064617acd294f8fa3542e141d647f40

data/lib/utils/linestatus.rb

@@ -29,7 +29,7 @@ module Metanorma
           @middoc_docattr = false
           @pass = false
         elsif (@is_delim && /^(--+|\*\*+|==+|__+)$/.match?(text)) ||
-            (!@is_delim && !@delimln && /^-----*$|^\.\.\.\.\.*$|\/\/\/\/\/*$/
+            (!@is_delim && !@delimln && /^-----*$|^\.\.\.\.\.*$|^\/\/\/\/\/*$/
           .match?(text))
           @delimln = text
           @pass = true

data/lib/utils/log.rb

@@ -91,7 +91,7 @@ module Metanorma
         if node.respond_to?(:line) && !node.line.nil?
           "#{'%06d' % node.line}"
         elsif /^XML Line /.match?(msg)
-          msg.sub(/^XML Line /, "").sub(/:.*$/, "")
+          msg.sub(/^XML Line /, "").sub(/(^[^:]+):.*$/, "\\1")
         else
           "000000"
         end

data/lib/utils/version.rb

@@ -1,5 +1,5 @@
 module Metanorma
   module Utils
-    VERSION = "1.9.0".freeze
+    VERSION = "1.10.1".freeze
   end
 end

data/lib/utils/xml.rb

@@ -48,49 +48,37 @@ module Metanorma
       # block for processing XML document fragments as XHTML,
       # to allow for HTMLentities
       # Unescape special chars used in Asciidoctor substitution processing
-      def noko(script = "Latn", &block)
-        doc = ::Nokogiri::XML.parse(NOKOHEAD)
-        fragment = doc.fragment("")
-        ::Nokogiri::XML::Builder.with fragment, &block
-        eoln = %w(Hans Hant Jpan).include?(script) ? "" : " "
-        fragment.to_xml(encoding: "US-ASCII", indent: 0,
-                        save_with: Nokogiri::XML::Node::SaveOptions::AS_XML)
-          .lines.map do |l|
-          l.gsub(/>\n$/, ">").gsub(/\n$/m, eoln).gsub("–", "\u0096")
-            .gsub("—", "\u0097").gsub("–", "\u0096")
-            .gsub("—", "\u0097")
-        end
-      end
-
       def noko(_script = "Latn", &block)
         fragment = ::Nokogiri::XML.parse(NOKOHEAD).fragment("")
         ::Nokogiri::XML::Builder.with fragment, &block
-        ret = fragment
+        fragment
           .to_xml(encoding: "UTF-8", indent: 0,
                   save_with: Nokogiri::XML::Node::SaveOptions::AS_XML)
-          .lines.map do |l|
-            l.rstrip.gsub("–", "\u0096").gsub("—", "\u0097")
-              .gsub("–", "\u0096").gsub("—", "\u0097")
-          end
-        line_sanitise(ret)
+          .gsub("–", "\u0096").gsub("—", "\u0097")
+          .gsub("–", "\u0096").gsub("—", "\u0097")
       end
 
       # By default, carriage return in source translates to whitespace;
-      # but in CJK, it does not.  We don't want carriage returns in the final
-      # output because of CJK complications
+      # but in CJK, it does not.  (Non-CJK text \n CJK)
       def line_sanitise(ret)
         ret.size == 1 and return ret
-        (0...ret.size).each do |i|
+        (0...(ret.size - 1)).each do |i|
           last = firstchar_xml(ret[i].reverse)
           nextfirst = firstchar_xml(ret[i + 1])
-          /#{CJK}/o.match?(last) && /#{CJK}/o.match?(nextfirst) or
-            ret[i] += " "
+          cjk1 = /#{CJK}/o.match?(last)
+          cjk2 = /#{CJK}/o.match?(nextfirst)
+          text1 = /[^\p{Z}\p{C}]/.match?(last)
+          text2 = /[^\p{Z}\p{C}]/.match?(nextfirst)
+          (cjk1 && (cjk2 || !text2)) and next
+          !text1 && cjk2 and next
+          ret[i] += " "
         end
         ret
       end
 
+      # need to deal with both <em> and its reverse string, >me<
       def firstchar_xml(line)
-        m = /^(<[^>]+>)*(.)/.match(line) or return ""
+        m = /^([<>][^<>]+[<>])*(.)/.match(line) or return ""
         m[2]
       end
 
@@ -111,15 +99,15 @@ module Metanorma
       end
 
       def ns(xpath)
-        xpath.gsub(%r{/([a-zA-z])}, "/xmlns:\\1")
-          .gsub(%r{::([a-zA-z])}, "::xmlns:\\1")
-          .gsub(%r{\[([a-zA-z][a-z0-9A-Z@/-]* ?=)}, "[xmlns:\\1")
-          .gsub(%r{\[([a-zA-z][a-z0-9A-Z@/-]*[/\[\]])}, "[xmlns:\\1")
+        xpath.gsub(%r{/([a-zA-Z])}, "/xmlns:\\1")
+          .gsub(%r{::([a-zA-Z])}, "::xmlns:\\1")
+          .gsub(%r{\[([a-zA-Z][a-z0-9A-Z@/-]* ?=)}, "[xmlns:\\1")
+          .gsub(%r{\[([a-zA-Z][a-z0-9A-Z@/-]*[/\[\]])}, "[xmlns:\\1")
       end
 
       def numeric_escapes(xml)
         c = HTMLEntities.new
-        xml.split(/(&[^ \r\n\t#;]+;)/).map do |t|
+        xml.split(/(&[^ \r\n\t#&;]+;)/).map do |t|
           if /^(&[^ \t\r\n#;]+;)/.match?(t)
             c.encode(c.decode(t), :hexadecimal)
           else t

data/metanorma-utils.gemspec

@@ -46,5 +46,6 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "timecop", "~> 0.9"
   spec.add_development_dependency "vcr", "~> 6.1.0"
   spec.add_development_dependency "webmock"
+  spec.add_development_dependency "xml-c14n"
   #spec.metadata["rubygems_mfa_required"] = "true"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: metanorma-utils
 version: !ruby/object:Gem::Version
-  version: 1.9.0
+  version: 1.10.1
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-07-21 00:00:00.000000000 Z
+date: 2024-08-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: asciidoctor
@@ -276,6 +276,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: xml-c14n
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: 'metanorma-utils provides utilities for the Metanorma stack
 
   '