html2doc 1.8.9 → 1.9.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: dd0f1be48de070c84ac6a59fb206218df94e8a73a6a1148d3b65ec7bb498fa70
-  data.tar.gz: 7478925f96c8a85c52ca9a4b9e48417425877f08afbbf2b847a09a72e061dd5d
+  metadata.gz: abee9d2b61164e6a1ca6a93832d2f0883fd7f759287590e0a3995287f1711964
+  data.tar.gz: efaee47d719f40b076dbc851299e7fd59594a77b3af1654fe3827ba1cce28e6a
 SHA512:
-  metadata.gz: 5b2f0e8ed50629189f149bbd4a06b0f2f72973e68280d1e895eb16e0a638bcd99d608d5b6572e3487e1d9f7264936afc711065fb8d6c2e7dee63c88009914724
-  data.tar.gz: c3f980b31ed8e10654a7a7c3326bb25ccebb00e58f30cc63d1a83244a0dd699bfe05d7527b2c7746b235edf075bcb1f84ce69db0a6ec5d4a861cabb4a5415a8f
+  metadata.gz: 7898834a57a94b995beea51a93270efa90b19f247f8d3bb4d8ee95d9216e9c3828ff8991432a7ac561fe7257fe5ae43760733f035d5562e4f6fc2953c5a9f597
+  data.tar.gz: 78ac742fe5ca906728541c21973739ea6b22256762fc361a9fc7264e3f966fd14f0f29e8f30e2f5a77b8c739201089decf88f3b4c8361f01ecea0413b1c3f464

data/html2doc.gemspec

@@ -28,9 +28,10 @@ Gem::Specification.new do |spec|
 
   spec.add_dependency "base64"
   spec.add_dependency "htmlentities", "~> 4.3.4"
+  spec.add_dependency "lutaml-model", "~> 0.7.0"
   spec.add_dependency "metanorma-utils", ">= 1.9.0"
   spec.add_dependency "mime-types"
-  spec.add_dependency "nokogiri", "~> 1.16.8"
+  spec.add_dependency "nokogiri", "~> 1.18.3"
   spec.add_dependency "plane1converter", "~> 0.0.1"
   spec.add_dependency "plurimath", "~> 0.9.0"
   spec.add_dependency "thread_safe"

data/lib/html2doc/math.rb

@@ -167,7 +167,8 @@ class Html2Doc
     doc.root = ooxml_cleanup(xml, docnamespaces)
     # d = xml.parent["block"] != "false" # display_style
     ooxml = Nokogiri::XML(Plurimath::Math
-      .parse(doc.root.to_xml(indent: 0), :mathml).to_omml(split_on_linebreak: true))
+      .parse(doc.root.to_xml(indent: 0), :mathml)
+      .to_omml(split_on_linebreak: true))
     ooxml = unitalic(accent_tr(ooxml))
     ooxml = ooml_clean(uncenter(xml, ooxml))
     xml.swap(ooxml)
@@ -217,15 +218,20 @@ class Html2Doc
     alignnode = math.xpath(STYLE_BEARING_NODE).last
     ooxml.document? and ooxml = ooxml.root
     ret = uncenter_unneeded(math, ooxml, alignnode) and return ret
-    dir = "left"
-    alignnode["style"]&.include?("text-align:right") and dir = "right"
-    ooxml.name == "oMathPara" or
-      ooxml.wrap("<m:oMathPara></m:oMathPara>")
+    dir = ooxml_alignment(alignnode)
+    ooxml.name == "oMathPara" or ooxml.wrap("<m:oMathPara></m:oMathPara>")
     ooxml.elements.first.previous =
       "<m:oMathParaPr><m:jc m:val='#{dir}'/></m:oMathParaPr>"
     ooxml
   end
 
+  def ooxml_alignment(alignnode)
+    dir = "left"
+    /text-align:\s*right/.match?(alignnode["style"]) and dir = "right"
+    /text-align:\s*center/.match?(alignnode["style"]) and dir = "center"
+    dir
+  end
+
   def uncenter_unneeded(math, ooxml, alignnode)
     (math_block?(ooxml, math) || !alignnode) and return ooxml
     math_only_para?(alignnode) and return nil

data/lib/html2doc/notes.rb

@@ -2,22 +2,27 @@ require "uuidtools"
 
 class Html2Doc
   def footnotes(docxml)
-    i = 1
+    #i = 1
+    indexes = {}
+    @footnote_idx = 1
     fn = []
     docxml.xpath("//a").each do |a|
-      next unless process_footnote_link(docxml, a, i, fn)
-
-      i += 1
+      process_footnote_link(docxml, a, indexes, fn) or next
+      #i += 1
     end
-    process_footnote_texts(docxml, fn)
+    process_footnote_texts(docxml, fn, indexes)
   end
 
-  def process_footnote_texts(docxml, footnotes)
+  # Currently cannot deal with separate footnote containers in each chapter
+  # We may eventually need to support that
+  def process_footnote_texts(docxml, footnotes, indexes)
     body = docxml.at("//body")
     list = body.add_child("<div style='mso-element:footnote-list'/>")
-    footnotes.each_with_index do |f, i|
-      fn = list.first.add_child(footnote_container(docxml, i + 1))
+    footnotes.each do |f|
+      #require 'debug'; binding.b
+      fn = list.first.add_child(footnote_container(docxml, indexes[f["id"]]))
       f.parent = fn.first
+      f["id"] = ""
       footnote_div_to_p(f)
     end
     footnote_cleanup(docxml)
@@ -47,14 +52,17 @@ class Html2Doc
     DIV
   end
 
-  def process_footnote_link(docxml, elem, idx, footnote)
-    return false unless footnote?(elem)
-
+  def process_footnote_link(docxml, elem, indexes, footnote)
+    footnote?(elem) or return false
     href = elem["href"].gsub(/^#/, "")
+    #require "debug"; binding.b
     note = docxml.at("//*[@name = '#{href}' or @id = '#{href}']")
-    return false if note.nil?
-
-    set_footnote_link_attrs(elem, idx)
+    note.nil? and return false
+unless indexes[href] 
+  indexes[href] = @footnote_idx
+@footnote_idx += 1
+end
+    set_footnote_link_attrs(elem, indexes[href])
     if elem.at("./span[@class = 'MsoFootnoteReference']")
       process_footnote_link1(elem)
     else elem.children = FN
@@ -73,7 +81,7 @@ class Html2Doc
   end
 
   def transform_footnote_text(note)
-    note["id"] = ""
+    #note["id"] = ""
     note.xpath(".//div").each { |div| div.replace(div.children) }
     note.xpath(".//aside | .//p").each do |p|
       p.name = "p"

data/lib/html2doc/version.rb

@@ -1,3 +1,3 @@
 class Html2Doc
-  VERSION = "1.8.9".freeze
+  VERSION = "1.9.1".freeze
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: html2doc
 version: !ruby/object:Gem::Version
-  version: 1.8.9
+  version: 1.9.1
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-03-17 00:00:00.000000000 Z
+date: 2025-04-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: base64
@@ -38,6 +38,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: 4.3.4
+- !ruby/object:Gem::Dependency
+  name: lutaml-model
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.7.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.7.0
 - !ruby/object:Gem::Dependency
   name: metanorma-utils
   requirement: !ruby/object:Gem::Requirement
@@ -72,14 +86,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.16.8
+        version: 1.18.3
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.16.8
+        version: 1.18.3
 - !ruby/object:Gem::Dependency
   name: plane1converter
   requirement: !ruby/object:Gem::Requirement