html2doc 1.5.5 → 1.6.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 46856bf56ad5dd95f8f5781dc11049bb4600060c28c49715a262837ece8028bf
-  data.tar.gz: 866ba19867f233b45aeee436df719679623d671902b30476b61952f7a6357e1f
+  metadata.gz: 6b08c4ee810280f83835a9884c17d54a04837195c1d133bedd48f8a103780316
+  data.tar.gz: f47560825288a3297700d048c3fff8c90ca61ba6bcb2b2e8fef3ac9749d9e094
 SHA512:
-  metadata.gz: b949f47c356437ce418f65ce7fd1c497648d0d0e960fe1e05d7318d280ddf6de23ddad8e5ab94a18f447b3eaba948b2a4db69ca2d00f0dcfebd692933a64c1da
-  data.tar.gz: 953999bd39aa1c1b6a0e1a34c939dcbdba01242c898f5a587eab546b4c9a4578e8051c9c68b49900390fcb3ab5d2de4ad6f0f1bce1af5a8ee6e7bd4daf300966
+  metadata.gz: 5a94da368aa84ae4abcbcf6c3bc401b349811b32b1d8097b82e96b479cd684d545ac66fc9fb6f41367db85bfeb86590a15bbac7af0e989c41a0e1af0e2c79966
+  data.tar.gz: 4dd0add251285b7c23a82b3b5d709d5421542b96ef0025357672616003d5e31ed35c4c064deb652925666a4a6e33962d6e73d72b137c8cea3592e5d4bb131574

data/lib/html2doc/base.rb

@@ -53,7 +53,7 @@ class Html2Doc
   end
 
   def process_html(result)
-    docxml = to_xhtml(asciimath_to_mathml(result, @asciimathdelims))
+    docxml = to_xhtml(result)
     define_head(cleanup(docxml))
     msword_fix(from_xhtml(docxml))
   end

data/lib/html2doc/math.rb

@@ -5,29 +5,6 @@ require "nokogiri"
 require "plane1converter"
 
 class Html2Doc
-  def asciimath_to_mathml1(expr, retain_asciimath)
-    ret = Plurimath::Math.parse(HTMLEntities.new.decode(expr), "asciimath").to_mathml
-      .gsub(/<math>/, "<math xmlns='http://www.w3.org/1998/Math/MathML'>").strip
-    retain_asciimath and
-      ret += "<asciimath>#{@c.encode(@c.decode(expr), :basic)}</asciimath>"
-    ret
-  rescue StandardError => e
-    puts "parsing: #{expr}"
-    puts e.message
-    raise e
-  end
-
-  def asciimath_to_mathml(doc, delims, retain_asciimath: false)
-    return doc if delims.nil? || delims.size < 2
-
-    m = doc.split(/(#{Regexp.escape(delims[0])}|#{Regexp.escape(delims[1])})/)
-    m.each_slice(4).map.with_index do |(*a), i|
-      progress_conv(i, 500, (m.size / 4).floor, 1000, "AsciiMath")
-      a[2].nil? or a[2] = asciimath_to_mathml1(a[2], retain_asciimath)
-      a.size > 1 ? a[0] + a[2] : a[0]
-    end.join
-  end
-
   def progress_conv(idx, step, total, threshold, msg)
     return unless (idx % step).zero? && total > threshold && idx.positive?
 
@@ -43,6 +20,8 @@ class Html2Doc
     doc
   end
 
+  MATHML_NS = "http://www.w3.org/1998/Math/MathML".freeze
+
   # random fixes to MathML input that OOXML needs to render properly
   def ooxml_cleanup(math, docnamespaces)
     math = unwrap_accents(
@@ -50,7 +29,7 @@ class Html2Doc
         mathml_insert_rows(math, docnamespaces), docnamespaces
       ),
     )
-    math.add_namespace(nil, "http://www.w3.org/1998/Math/MathML")
+    math.add_namespace(nil, MATHML_NS)
     math
   end
 
@@ -151,8 +130,8 @@ class Html2Doc
   def mathml_to_ooml1(xml, docnamespaces)
     doc = Nokogiri::XML::Document::new
     doc.root = ooxml_cleanup(xml, docnamespaces)
-    ooxml = ooml_clean(unitalic(esc_space(accent_tr(@xsltemplate.transform(doc)))))
-    ooxml = uncenter(xml, ooxml)
+    ooxml = unitalic(esc_space(accent_tr(@xsltemplate.transform(doc))))
+    ooxml = ooml_clean(uncenter(xml, ooxml))
     xml.swap(ooxml)
   end
 
@@ -184,19 +163,33 @@ class Html2Doc
     xml
   end
 
+  OOXML_NS = "http://schemas.microsoft.com/office/2004/12/omml".freeze
+
+  def math_only_para?(node)
+    x = node.dup
+    x.xpath(".//m:math", "m" => MATHML_NS).each(&:remove)
+    x.xpath(".//m:oMathPara | .//m:oMath", "m" => OOXML_NS).each(&:remove)
+    x.text.strip.empty?
+  end
+
+  def math_block?(ooxml, mathml)
+    ooxml.name == "oMathPara" || mathml["displaystyle"] == "true"
+  end
+
+  STYLE_BEARING_NODE =
+    %w(p div td th li).map { |x| ".//ancestor::#{x}" }.join(" | ").freeze
+
   # if oomml has no siblings, by default it is centered; override this with
   # left/right if parent is so tagged
+  # also if ooml has mathPara already, or is in para with only oMath content
   def uncenter(math, ooxml)
-    alignnode = math.at(".//ancestor::*[@style][local-name() = 'p' or "\
-                        "local-name() = 'div' or local-name() = 'td']/@style")
-    return ooxml unless alignnode && (math.next == nil && math.previous == nil)
-
-    %w(left right).each do |dir|
-      if alignnode.text.include? ("text-align:#{dir}")
-        ooxml = "<m:oMathPara><m:oMathParaPr><m:jc "\
-                "m:val='#{dir}'/></m:oMathParaPr>#{ooxml}</m:oMathPara>"
-      end
-    end
-    ooxml
+    alignnode = math.xpath(STYLE_BEARING_NODE).last
+    ret = ooxml.root.to_xml(indent: 0)
+    (math_block?(ooxml, math) ||
+      !alignnode) || !math_only_para?(alignnode) and return ret
+    dir = "left"
+    alignnode["style"]&.include?("text-align:right") and dir = "right"
+    "<oMathPara><oMathParaPr><jc " \
+      "m:val='#{dir}'/></oMathParaPr>#{ret}</oMathPara>"
   end
 end

data/lib/html2doc/version.rb

@@ -1,3 +1,3 @@
 class Html2Doc
-  VERSION = "1.5.5".freeze
+  VERSION = "1.6.1".freeze
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: html2doc
 version: !ruby/object:Gem::Version
-  version: 1.5.5
+  version: 1.6.1
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2023-06-29 00:00:00.000000000 Z
+date: 2023-08-12 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: htmlentities