pandoc_refeq_mathml 0.1

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: c025068f2083e35c2d82be9dadf5260b70f2823ce4ab9607238395388d4b84dc
+  data.tar.gz: bb47e488a24a8a23b1f93be4c001225492c26465ec77fc425aab9d34876f1113
+SHA512:
+  metadata.gz: 53f285f6113a7173c112f06bb06bbe27d7aa8ab6a2288ca98059fd33550c5dc7ff932a62de8ad7dbe7f11cddad818fbec795959a9c569717ae6df515e41e8ea2
+  data.tar.gz: eec508b49714a0a76fde53883459d41940a3814c763902aee73887603a4883a6d1d378461504af9fb01ccd31ab62bbe270770e3663abee4d8ae34cfcdddbb2f5

data/.gitignore

@@ -0,0 +1,51 @@
+# See https://help.github.com/articles/ignoring-files for more about ignoring files.
+#
+# If you find yourself ignoring temporary files generated by your text editor
+# or operating system, you probably want to add a global ignore instead:
+#   git config --global core.excludesfile '~/.gitignore_global'
+
+# Ignore bundler config.
+/.bundle
+/vendor/bundle
+
+# Ignore all logfiles and tempfiles.
+/log/*
+/tmp/*
+!/log/.keep
+!/tmp/.keep
+
+.rbenv-version
+# unless supporting rvm < 1.11.0 or doing something fancy, ignore this:
+.rvmrc
+
+/node_modules
+/yarn-error.log
+
+.byebug_history
+
+*.[oa]
+*.so
+*~
+*.nogem
+*nogem.*
+*.bak
+*.BAK
+*.backup
+*.org
+*.orig
+*.elc
+*.pyc
+\#*\#
+
+# Elastic Beanstalk Files
+.elasticbeanstalk/*
+!.elasticbeanstalk/*.cfg.yml
+!.elasticbeanstalk/*.global.yml
+
+# yard
+*.yardoc
+
+# Ruby Gem doc
+*.gem
+doc/*
+

data/ChangeLog

@@ -0,0 +1,5 @@
+-----
+(Version: 1.0)
+2022-08-26  Masa Sakano
+
+ * Initial commit.

data/LICENSE.txt

@@ -0,0 +1,20 @@
+Copyright (c) 2022 Masa Sakano
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

data/Makefile

@@ -0,0 +1,26 @@
+ALL	= 
+
+objs	= 
+
+.SUFFIXES:	.so .o .c .f
+
+#.o.so:
+#	${LD} ${LFLAGS} -o $@ $< ${LINK_LIB}
+
+all: ${ALL}
+
+
+.PHONY: clean test doc
+clean:
+	$(RM) bin/*~
+
+## You may need RUBYLIB=`pwd`/lib:$RUBYLIB
+test:
+	rake test
+
+## yard2md_afterclean in Gem plain_text https://rubygems.org/gems/plain_text
+doc:
+	yard doc; [[ -x ".github" && ( "README.en.rdoc" -nt ".github/README.md" ) ]] && ( ruby -r rdoc -e 'puts RDoc::Markup::ToMarkdown.new.convert ARGF.read' < README.en.rdoc | /Users/alpin/Rubygems/Project/plain_text/bin/yard2md_afterclean --lang= > .github/README.md.$$ && ( mv -f .github/README.md.$$ .github/README.md && echo ".github/README.md is updated." ) || ( echo "ERROR: failed to create .github/README.md" >&2 ) ) || exit 0
+#	yard doc; [[ -x ".github" && ( "README.en.rdoc" -nt ".github/README.md" ) ]] && ( ruby -r rdoc -e 'puts RDoc::Markup::ToMarkdown.new.convert ARGF.read' < README.en.rdoc | yard2md_afterclean --lang=bash > .github/README.md.$$ && ( mv -f .github/README.md.$$ .github/README.md && echo ".github/README.md is updated." ) || ( echo "ERROR: failed to create .github/README.md" >&2 ) ) || exit 0
+
+

data/README.en.rdoc

@@ -0,0 +1,227 @@
+
+= PandocRefeqMathml - ad hoc tool to modify pandoc-converted MathML from LaTeX
+
+== Summary
+
+This Ruby command-line command modifies a MathML file converted with <tt>pandoc</tt> from LaTeX.
+
+Whereas <tt>pandoc</tt> is a great text-ish file converter, there are a
+few caveats, at the time of writing, in converting a LaTeX file to MathML.
+
+A major caveat is the generated MathML does not display
+the equation numbers that are auto-generated by LaTeX in default for
+the <tt>equation</tt> and <tt>eqnarray</tt> environments, nor their (LaTeX) labels.
+All the (LaTeX) <tt>ref</tt> remain as they are, which is a coded message
+for readers.
+
+Another caveat is the alignments of equations in the <tt>eqnarray</tt> environment.
+
+This tool is a bit of ad hoc (dirty) hack to correct these points
+*in some basic situations*.  "Basic" here means just the standard
+LaTeX commands, not some package-specific commands. Also, this may not
+correctly handle complicated formats of equations, using arrays etc.
+
+The full package of this module is found in
+{PandocRefeqMathml Ruby Gems page}[http://rubygems.org/gems/pandoc_refeq_mathml]
+(with document created from source annotation with yard) and
+in {Github}[https://github.com/masasakano/pandoc_refeq_mathml]
+
+== Background and constraints
+
+Pandoc-converted MathML.html from LaTeX lacks equation numbers that
+are present in the original LaTeX.  This tool offers a very crude fix,
+adding equation numbers based on the annotation fields in +<math>+
+and LaTeX aux file (which is automatically generated as a byproduct
+when you compile a LaTeX document).  Not all the numbers are recovered
+but only those that are referenced somewhere in the MathML/LaTeX
+file.
+
+(Note that in principle, it should not be too difficult to modify the program so
+that all the labelled equations in LaTeX are labelled again in
+MathML.  However, it would be tricky to label equations that are not
+explicitly labelled in LaTeX because implicit numbering information is not
+available in the LaTeX aux file.)
+
+The algorithm assumes a LaTeX standard aux file-format, the MathML having
+a link tag +<a>+ with the attributes "data-reference-type=ref" and href to the label
+of the exact reference label in LaTeX (and the label should have no duplicates
+in the MathML) and also having the +'annotation[ encoding="application/x-tex"]'+ tag
+in each math tag containing the original LaTeX code. The LaTeX code must have
+either the standard "equation" or "eqnarray" structures associated with
+the standard "label" tag with a simple content (if it contains, apart from the label string,
+something more than preceding or trailing white spaces, such as a comment,
+this algorithm would likely fail).  If equations in an eqnarray environment have
+complicated nested structures like a matrix, I do not know how the algorithm
+of this routine handles them.  Also, the LaTeX section numbering must
+be combinations of Arabic numbers, full-stops, and maybe capital
+letters (for Appendix) only.
+
+Essentially, LaTeX has a huge amount of freedom and so I am afraid it would be a somewhat
+futile effort to deal with every possibility...
+
+=== Output MathML by pandoc-2.19 converted from LaTeX
+
+Ordinary LaTeX inline maths expressions (e.g., +$5^2$+) are expressed
+as follows:
+
+  <math display="inline" xmlns="http://w..."><semantics>
+   <mrow><mn>5</mn><mi>π</mi></mrow>
+   <annotation encoding="application/x-tex">5\pi</annotation>
+  </semantics></math>
+
+LaTeX's +begin{equation}+ is as follows (n.b., the +<p>+ tag may not
+be closed immediately after +</math>+ but another ordinary sentences
+may follow):
+
+  <p><math display="block" xmlns="http://w..."><semantics>
+   <mrow><mi>x</mi><mo>±</mo><mi>ϵ</mi></mrow></mrow>
+   <annotation encoding="application/x-tex">x \pm \epsilon \label{my_xe}</annotation>
+  </semantics></math>
+
+LaTeX's +begin{eqnarray}+ is as follows:
+
+  <p><math display="block" xmlns="http://w..."><semantics><mtable>
+   <mtr><mtd columnalign="right"><mrow><mn>1</mn><mo>+</mo><mi>x</mi></mrow></mtd>
+        <mtd columnalign="left"><mo>=</mo></mtd>
+        <mtd columnalign="right"><mrow><mn>1</mn><mo>−</mo><mi>x</mi></mrow></mtd></mtr>
+   <mtr><mtd columnalign="right"></mtd>
+        <mtd columnalign="left"><mo>=</mo></mtd>
+        <mtd columnalign="right"><mfrac><mn>2</mn><mrow><mn>1</mn><mi>x</mi></mrow></mfrac></mtd></mtr>
+   </mtable><annotation encoding="application/x-tex">\begin{aligned}
+    1+x &amp; = &amp; 1-x \nonumber\\
+        &amp; = &amp; \frac{2}{1x} \label{eq_trivial}
+    \end{aligned}</annotation></semantics></math></p>
+
+They are referred to as from another text follows:
+
+  <p>Eq.<a href="#eq_trivial" data-reference-type="ref"
+     data-reference="eq_trivial">[eq_trivial]</a> was easy...
+
+=== Algorithm
+
+For fixing the alignments to follow the standard eqnarray alignments
+(right, centre, and left in this order), the program searches for
++<mtable>+ and rewrites the <tt>columnalign</tt> attributes in the +<mtd>+ tags.
+
+For fixing the equation numbers and links, the program
+
+1. first reads a LaTeX aux file and lists all the labels for equations and their numbers.
+2. Then, it picks up an internally-pointing HTML anchor,
+3. matches it with the list generated from the LaTeX aux file and identifies the equation number,
+4. searches labels in +<annotation>+ tags for the identical string for the HTML/MathML-anchor,
+5. identifies the exact equation corresponding to the label (if in the eqnarray environment),
+6. inserts the identified equation number next to the MathML equation,
+7. and finally modifies the plain text for the HTML anchor.
+
+Each of the inserted equation number next to the corresponding
+equation is inside the +<mtext>+ tags. In +<mtable>+ (for LaTeX +\eqnarray{}+),
+it is inserted as a new +<mtd>+ cell.
+In both cases, the text is right-aligned with some padding to the
+left.  However, the position is relative to either the equation or the set of the
+equations that contains the relevant equation (for LaTeX
++\eqnarray{}+) and is not like the original LaTeX, where equation
+numbers inside a pair of parentheses are always located at the right
+edge of a page in default.
+
+== How to use this command
+
+Once you have installed it according to the standard RubyGems procedure
+(see section Install), the main Ruby executable (command) <tt>pandoc_refeq_mathml</tt>
+should be in your command-search path.
+
+It basically reads a MathML file from either the first command-line
+argument or STDIN and also a LaTeX aux file specified in a
+command-line, and then outputs the modified (corrected) MathML to STDOUT.
+
+Any warnings are printed to either STDERR or a log-file specified in a
+command-line as an option.
+
+Failure in matching the labels from an HTML tag with any of the MathML
+equations are printed as a warning (to STDERR in default).
+Although it may genuinely mean the non-existent labels in the original
+LaTeX source, it is far more likely that the labels belong to one of
+the sections (or tables of figures), because the algorithm cannot tell
+what the type (section, table, figure, or equation or else) of each label's origin is.
+
+=== Help doc
+
+The help doc for the command-line interface is displayed with +-h+ (or +--help+) option:
+
+  % pandoc_refeq_mathml -h
+  Usage: pandoc_refeq_mathml [options] [--] [MathML.html] > STDOUT
+         pandoc_refeq_mathml [options] [--] < STDIN > STDOUT
+  
+  Description (Version=0.1):
+     This fixes issues, label-references of equations and eqnarray alignments, of pandoc-converted MathML from LaTeX.
+  
+  Specific options:
+      -a, --aux [FILENAME]             (mandatory) LaTeX aux filename
+          --log [FILENAME]             Log filename (Default: STDERR). /dev/null to disable it.
+          --[no-]fixalign              Fix eqnarray-alignment problems? (Def: true)
+      -v, --[no-]verbose               Run verbosely (Def: true)
+  
+  Common options:
+      -h, --help                       Show this message
+          --version                    Show version
+
+=== Examples
+
+    % pandoc_refeq_mathml --aux=mydoc.aux --log=error.log mydoc.html > revised1.html
+    % head -n 90 mydoc.html | pandoc_refeq_mathml --aux=mydoc.aux --no-fixalign > revised2.html
+
+== Install
+
+Standard Ruby-gem install procedure is suffice
+
+    % gem install pandoc_refeq_mathml
+
+which should also install the dependant 
+{Nokogiri gem}[https://rubygems.org/gems/nokogiri/].
+
+Alternatively, it is possible to download the library file
+<tt>lib/pandoc_refeq_mathml.rb</tt> somewhere in your local directory,
+set the environmental variable <tt>RUBYLIB</tt> to also point to the
+directory for the library, and execute 
+
+    % ruby bin/pandoc_refeq_mathml
+
+where <tt>ruby</tt> is optional.  Note that {Nokogiri gem}[https://rubygems.org/gems/nokogiri/]
+must be available in your RUBY library path.
+
+In the developer's environment {diff-lcs gem}[https://rubygems.org/gems/diff-lcs] is also required.
+
+This tool requires {Ruby}[http://www.ruby-lang.org] Version 2.0
+or above.
+
+== Developer's note
+
+The source code is maintained also in
+{Github}[https://github.com/masasakano/pandoc_refeq_mathml]
+with no intuitive interface for annotations.
+
+=== Tests
+
+The Ruby codes under the directory <tt>test/</tt> are the test scripts.
+You can run them from the top directory as <tt>ruby test/test_****.rb</tt>
+or simply run <tt>make test</tt> or <tt>rake test</tt>.
+
+== Known bugs and ToDo items
+
+* pandoc-generated HTMLs do not contain Table/Figure numbers in their +<caption>+, even
+  though each anchored text refers to the corresponding number, such
+  as, +see Table "2"+, where "2" is the anchor.
+* In fact, pandoc-generated HTMLs do not generate +<figure>+ tags, let
+  alone +<figurecaption>+ for
+  the LaTeX figure environments that contain more than one figure
+  (with +\includegraphics+)...
+
+
+== Copyright
+
+Author::  Masa Sakano < info a_t wisebabel dot com >
+Versions:: The versions of this package follow Semantic Versioning (2.0.0) http://semver.org/
+License:: MIT
+Warranty:: No warranty.
+
+----------
+

data/Rakefile

@@ -0,0 +1,9 @@
+require 'rake/testtask'
+
+Rake::TestTask.new do |t|
+  t.libs << 'test'
+end
+
+desc "Run tests"
+task :default => :test
+

data/bin/pandoc_refeq_mathml

@@ -0,0 +1,195 @@
+#!/usr/bin/env ruby
+# coding: utf-8
+
+require 'optparse'
+require 'nokogiri'
+
+require 'pandoc_refeq_mathml'
+
+# Optparse-handling Template: <https://gist.github.com/masasakano/45ee0d737f2d33122e8ff44007693b40>
+class OptparsePandocRefeqMathml
+  begin
+    ver = nil
+    gemfile = __dir__+'/../pandoc_refeq_mathml.gemspec'
+    open(gemfile){ |ioin|
+      while line=ioin.gets
+        if /^\s*[^.]+.version\s*=\s*(['"])([\d\._a-zA-Z\-]+)\1/ =~ line
+          ver = $2
+          break $2
+        end
+      end
+    }
+  rescue
+    warn "gemspec file (#{gemfile}) is not found or its format is unexpected." #if $DEBUG
+  ensure
+    # Reads Version from the gemspec file.
+    # Make sure the format of the version in the gemspec file is standard,
+    # using quotations (not like "%q@1.2.3@").
+    Version = (ver || "0.1")  # Just in case, update this hard-coded fallback version number, too!
+  end
+
+
+  # Default values for the command-line options.
+  # If nil, it means they are mandatory AS DEFINED in "def parse(args)" below,
+  # except for the 2 arguments of "inplace" and "extension".
+  # The Hash keyword (and attribute) names should be identical to the command-line
+  # option names, except "inplace" and "extension" in this case as demonstrated
+  # (in *parse()* etc).
+  DEF_OPTS = {
+    aux: nil,    # mandatory LaTeX aux filename
+    auxstr: "",  # Contents (String) of the LaTeX aux (extracted from --aux option)
+    fixalign: true, # Fix eqnarray-originating alignment issues, if true
+    #fixref: true, # Fix equation-labels and refs (Def: true); this option is not implemented.
+    log: "",     # Log filename, where warning messages are recorded. Default is "", meaning STDERR
+    verbose: true,
+  }
+
+  class ScriptOptions
+
+    # attr_accessor :aux, :auxstr, :verbose
+    DEF_OPTS.each_key do |ek|
+      attr_accessor ek
+    end
+
+    def initialize
+      DEF_OPTS.each_pair do |ek, ev|
+        self.public_send(ek.to_s+"=", ev)
+      end
+    end
+
+    def define_options(parser)
+      parser.banner = "Usage: pandoc_refeq_mathml [options] [--] [MathML.html] > STDOUT"
+      parser.separator "       pandoc_refeq_mathml [options] [--] < STDIN > STDOUT"
+
+      descriptions = []
+      descriptions.push <<EOF
+   This fixes issues, label-references of equations and eqnarray alignments, of pandoc-converted MathML from LaTeX.
+EOF
+
+      parser.separator ""
+      parser.separator "Description (Version=#{Version}):"
+      descriptions.each do |et|
+        parser.separator et.gsub(/\s*\n\*/, " ")
+        parser.separator ""
+      end
+      parser.separator "Specific options:"
+
+      # add additional options
+      read_aux_option(parser)  # mandatory
+      read_log_option(parser)  # mandatory
+      fixalign_option(parser)
+      boolean_verbose_option(parser)
+
+      parser.separator ""
+      parser.separator "Common options:"
+      # No argument, shows at tail.  This will print an options summary.
+      parser.on_tail("-h", "--help", "Show this message") do
+        puts parser
+        exit
+      end
+      # Another typical switch to print the version.
+      parser.on_tail("--version", "Show version") do
+        puts Version
+        exit
+      end
+    end
+
+    # Specifies a LaTeX aux filename
+    def read_aux_option(parser)  # mandatory option
+      parser.on("-a", "--aux [FILENAME]",  # Do not change this into "--aux FILENAME".
+                "(mandatory) LaTeX aux filename") do |fname|
+        raise OptionParser::MissingArgument if !fname  # This is necessary as fname is mandatory!
+        self.aux = fname
+        self.auxstr = File.read fname  # may raise an Exception!
+      end
+    end
+
+    # Specifies a log file
+    #
+    # if self.log.empty? is true, it should be treated as STDERR later. 
+    def read_log_option(parser)
+      parser.on("--log [FILENAME]",  # Do not change this into "--aux FILENAME".
+                "Log filename (Default: STDERR). /dev/null to disable it.") do |fname|
+        raise OptionParser::MissingArgument if !fname  # This is necessary as fname is mandatory!
+        self.log = ((fname == "STDERR") ? "" : fname)
+      end
+    end
+
+    # Fix alignment-problems? (Boolean)
+    def fixalign_option(parser)
+      parser.on("--[no-]fixalign", "Fix eqnarray-alignment problems? (Def: #{self.fixalign.inspect})") do |v|
+        self.fixalign = v
+      end
+    end
+
+    # Boolean switch.
+    def boolean_verbose_option(parser)
+      parser.on("-v", "--[no-]verbose", "Run verbosely (Def: #{self.verbose.inspect})") do |v|
+        self.verbose = v
+      end
+    end
+  end  # class ScriptOptions
+
+  #
+  # Return a structure describing the options.
+  #
+  def parse(args)
+    # The options specified on the command line will be collected in
+    # *options*.
+    #
+    # NOTE: if undefined options are specified, OptionParser::InvalidOption is raised.
+
+    @options = ScriptOptions.new
+    @args = OptionParser.new do |parser|
+      @options.define_options(parser)
+      begin
+        parser.parse!(args)
+
+        # Error if mandatory options are not specified.
+        DEF_OPTS.keys.each do |ek|  # 2 arguments are exceptions
+          if @options.public_send(ek).nil?  # *.nil? is used b/c "false" must be accepted.
+            # Assuming the Hash keyword (and attribute) name is identical to the option name!
+            warn "ERROR: Keyword argument --#{ek.to_s} is mandatory."
+            exit 1
+          end
+        end
+
+      rescue OptionParser::MissingArgument => err
+        # Missing argument for optional arguments.
+        warn sprintf "ERROR: %s: %s", err.reason, err.args.join(" ")
+        exit 1
+
+      rescue OptionParser::ParseError => err
+        # Other types of argument-handling errors
+        raise
+      end
+    end
+    @options
+  end
+
+  attr_reader :parser, :options
+end  # class OptparsePandocRefeqMathml
+
+####################################
+# MAIN
+####################################
+
+if $0 == __FILE__
+  cmdarg = OptparsePandocRefeqMathml.new
+  cmdopts = cmdarg.parse(ARGV)  # Optional command-line options, apart from ARGV
+    # cmdopts.auxstr : String of *.aux
+
+  logio = (cmdopts.log.empty? ? $stderr : open(cmdopts.log, "w"))
+  logio.sync = true
+
+  begin
+    apmre = PandocRefeqMathml.new( Nokogiri::HTML(ARGF.read), cmdopts.auxstr, logio: logio, is_verbose: cmdopts.verbose)
+    apmre.alter_html!(fixalign: cmdopts.fixalign)
+
+    # outputs the alterned HTML to STDOUT
+    puts apmre.page.to_s
+  ensure
+    logio.close if logio != $stderr 
+  end
+end
+

data/lib/pandoc_refeq_mathml.rb

@@ -0,0 +1,189 @@
+# coding: utf-8
+
+require 'nokogiri'
+
+# Class to handle MathML and LaTeX aux
+#
+class PandocRefeqMathml
+
+  attr_reader :page, :aux
+
+  # Alignment for LaTeX eqnarray.
+  EQNARRAY_ALIGNS = %w(right center left)
+
+  # @param page [Nokogiri::HTML4::Document]
+  # @param auxstr [String] String of the contents of a LaTeX aux file
+  # @param logio [IO] Output IO for logs. You may give +IO.open("/dev/null", "w")+. Def: $stderr
+  # @param is_verbose [Boolean] If true (Default), verbose.
+  def initialize(page, auxstr, logio: $stderr, is_verbose: true)
+    @page = page
+    @auxstr = auxstr
+    @logio = logio
+    @is_verbose = is_verbose
+    @hslabel = {}  # String(label) => String(EqNumber), which have been detected.
+  end
+
+  # @param taga [Nokogiri::HTML4::Document] Nokogiri <A> object.
+  # @return [String, NilClass] Label string used in +<a href=..>+. Nil if something is wrong.
+  def get_label(taga)
+    kwdlink = taga['href'].split('#')[1]
+    return kwdlink if kwdlink == taga['data-reference']
+
+    @logio.puts "WARNING: Inconsistent href and data-reference: "+a.to_s if @is_verbose
+    return nil
+  end
+
+  # Return the equation number.
+  #
+  # The number is assumed to contain only numbers and maybe full-stops +[\d.]+.
+  #
+  # @return [String, NilClass] Equation number guessed from the Aux file. nil if something goes wrong.
+  def get_eq_num(kwdlink)
+    mat = /^\\newlabel\{#{kwdlink}\}\{\{([\d\.]+)\}\{(\d+)\}\{(.*)\}\{equation.([a-zA-Z\d\.]+)\}/.match @auxstr  # it may be like equation.B.193 (for Appendix B).
+       # => #<MatchData "\\newlabel{eq_my_lab}{{65}{35}{割り算}{equation.4.62}" 1:"65" 2:"35" 3:"割り算" 4:"4.62">
+    return mat[1] if mat && mat[1] && !mat[1].empty?
+
+    ## Something is wrong. 
+    str = sprintf 'WARNING: Not found equation number for label="%s" (maybe it is for a section etc?): MatchData=%s', kwdlink, mat.inspect
+    @logio.puts str
+    return nil
+  end
+
+  # @param kwdlink [String] label
+  # @param n_eq [String] Equation number like "58", maybe "52.3"
+  # @return [Integer, NilClass] If something goes wrong, 
+  def find_insert_n_eq(kwdlink, n_eq)
+    # Select the <math> tag component that hs the kwdlink
+    maths = @page.css('math').select{|ep|
+      /\\label\{\s*#{Regexp.quote(kwdlink)}\s*\}/ =~ (ep.css('annotation[encoding="application/x-tex"]').children[0].text.strip rescue "X")
+    }
+    if maths.size != 1
+      if maths.size == 0
+        @logio.puts 'WARNING: no math tag contains label="#{kwdlink}"'
+      else
+        @logio.print 'WARNING: Multiple math tags contain label="#{kwdlink}"'
+        @logio.puts (@is_verbose ? ": maths="+maths.inspect : "")
+      end
+      return nil
+    end
+
+    mtext = sprintf '<mtext id="%s" style="padding-left:1em; text-align:right;">(%s)</mtext>', kwdlink, n_eq
+    
+    if maths[0].css('mtable').empty?
+      # \begin{equation}
+      #
+      # Insert the new node (<mrow><mtext...>(65)</mtext></mrow>) as the last child of
+      # the last top-level existing <mrow>; if it is added AFTER the <mrow>
+      # the <mtext> number would not be displayed!
+      newnode = '<mrow>' + mtext + '</mrow>'
+      begin
+        maths[0].css("mrow")[0].parent.css("> mrow")[-1].add_child(newnode)
+          # Between the last <mrow> and <annotation>
+          # n.b., css('mrow')[-1] would give an mrow inside another mrow!
+      rescue
+        msg = "FATAL: contact the code developer: equation: maths[0]="+maths[0].inspect
+        @logio.puts msg
+        raise msg
+      end
+      return 0
+    else
+      # \begin{eqnarray}
+      newnode = '<mtd columnalign="right">' + mtext + '</mtd>'
+      annot_node= maths[0].css('annotation[encoding="application/x-tex"]')[0]
+      i_eq_annot = annot_node.children[0].text.split(/\\\\\s*(?:\%[^\n]*)?\n?/).find_index{|ev| /\\label\{#{kwdlink}\}/ =~ ev}  # Index of the equation (starting from 0) in the eqnarray
+      raise "FATAL: contact the code developer: eqnarray: "+annot_node.inspect if !i_eq_annot
+
+      # Insert the new node (<mtd><mtext...>(65)</mtext></mtd>)
+      mtrnode = maths[0].css('mtable mtr')[i_eq_annot]
+      # mtrnode.css('mtd')[-1].add_next_sibling(newnode)  # not considering multi-layer tables
+      find_last_shallowest(mtrnode, 'mtd').add_next_sibling(newnode)
+    end
+    return i_eq_annot+1
+  end
+
+  # @param kwdlink [String] label
+  # @param taga [Nokogiri::HTML4::Document] Nokogiri <A> object.
+  # @param n_eq_str [String] equation number string
+  # @return [void]
+  def alter_link_text(kwdlink, taga, n_eq_str)
+    textnode = taga.children[0]
+    if !textnode.text?
+      @logio.puts "WARNING: Inconsistent text inside href: "+taga.to_s
+      return nil
+    end
+
+    if /\A\[?#{Regexp.quote(kwdlink)}\]?\z/ !~ textnode.to_s.strip
+      @logio.puts "WARNING: Strange linked-text=(#{textnode.to_s}) inside <a>: "+a.to_s if @is_verbose
+    end
+
+    taga.content=n_eq_str
+  end
+
+  # Alter the alignments of mtable originating from eqnarray
+  #
+  # Original is all right-aligned.
+  # After alteration, it will be right, center, left (which is the specification of eqnarray).
+  def alter_align_eqnarray!
+    @page.css("math mtable mtr").each do |ea_mtr|
+      ea_mtr.css("mtd").each_with_index do |ea_mtd, i|
+        break if i >= EQNARRAY_ALIGNS.size
+        ea_mtd["columnalign"]=EQNARRAY_ALIGNS[i]
+      end
+    end
+  end
+
+  # Alter the existing HTML Nokogiri content
+  def alter_reflinks!
+    all_ref_href = @page.css("a[data-reference-type=ref]")
+    all_ref_href.each do |ea_nodes|
+      # Gets a label from MathML
+      (kwdlink = get_label(ea_nodes)) || next
+
+      if !@hslabel.keys.include? kwdlink
+        # Gets the number of the equation from Aux
+        n_eq = get_eq_num(kwdlink)
+        next if !n_eq
+        @hslabel[kwdlink] = n_eq
+
+        # Finds the equation in MathML and adds the number of the equation.
+        find_insert_n_eq(kwdlink, n_eq) # this returns nil if something goes wrong
+      elsif !@hslabel[kwdlink]
+        # the label "kwdlink" has been detected, but no Equation-number was found.
+        next
+      end
+
+      # Alter the original link text in MathML to Equation number.
+      alter_link_text(kwdlink, ea_nodes, @hslabel[kwdlink])
+    end
+  end
+
+  # @param fixalign [Boolean] fix alignments if true
+  def alter_html!(fixalign: true)
+    alter_align_eqnarray! if fixalign
+    alter_reflinks!
+  end
+
+  # Returns the last shallowest node with the given tag-name
+  #
+  # @see https://stackoverflow.com/a/73459162/3577922
+  #
+  # @param root [NokogiriXmlNode] root node
+  # @param tagname [String] Tag-name like "mrow" for which the last-shallowest is looked
+  def find_last_shallowest(root, tagname)
+    queue = [root]
+    while queue.any?
+      element = queue.shift
+      return element if node_matching?(element, tagname)
+      queue.concat element.children.reverse
+    end
+  end
+  private :find_last_shallowest
+
+  # @param element [NokogiriXmlNode] including Nokogiri::XML::Element
+  def node_matching?(element, tagname)
+    # Put your matching logic here
+    element.name == tagname
+  end
+  private :node_matching?
+end # PandocRefeqMathml
+

data/pandoc_refeq_mathml.gemspec

@@ -0,0 +1,59 @@
+# -*- encoding: utf-8 -*-
+
+require 'rake'
+require 'date'
+
+Gem::Specification.new do |s|
+  s.name = 'pandoc_refeq_mathml'.sub(/.*/){|c| (c == File.basename(Dir.pwd)) ? c : raise("ERROR: s.name=(#{c}) in gemspec seems wrong!")}
+  s.version = "0.1".sub(/.*/){|c| fs = Dir.glob('changelog{,.*}', File::FNM_CASEFOLD); raise('More than one ChangeLog exist!') if fs.size > 1; warn("WARNING: Version(s.version=#{c}) already exists in #{fs[0]} - ok?") if fs.size == 1 && !IO.readlines(fs[0]).grep(/^\(Version: #{Regexp.quote c}\)$/).empty? ; c }  # n.b., In macOS, changelog and ChangeLog are identical in default.
+  # s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
+  # s.bindir = 'bin'
+  # %w(slim_string).each do |f|
+  #   path = s.bindir+'/'+f
+  #   File.executable?(path) ? s.executables << f : raise("ERROR: Executable (#{path}) is not executable!")
+  # end
+  s.authors = ["Masa Sakano"]
+  s.date = %q{2022-08-26}.sub(/.*/){|c| (Date.parse(c) == Date.today) ? c : raise("ERROR: s.date=(#{c}) is not today!")}
+  s.summary = %q{Add equation numbers to a pandoc-output MathML converted from LaTeX}
+  s.description = <<-EOF
+Add equation numbers in a crude way to a pandoc-output MathML converted from LaTeX, utilising its LaTeX aux file, and also adjust math-table alignments.
+  EOF
+  # s.email = %q{abc@example.com}
+  s.extra_rdoc_files = [
+     #"LICENSE.txt",
+     "README.en.rdoc",
+  ]
+  s.license = 'MIT'
+  s.files = FileList['.gitignore','lib/**/*.rb','[A-Z]*','test/**/*.rb', '*.gemspec', 'bin/pandoc_refeq_mathml'].to_a.delete_if{ |f|
+    ret = false
+    arignore = IO.readlines('.gitignore')
+    arignore.map{|i| i.chomp}.each do |suffix|
+      if File.fnmatch(suffix, File.basename(f))
+        ret = true
+        break
+      end
+    end
+    ret
+  }
+  s.files.reject! { |fn| File.symlink? fn }
+
+  s.add_runtime_dependency 'nokogiri', '>= 1.13'
+  s.add_development_dependency 'diff-lcs', '>= 1.5'
+
+  s.homepage = "https://www.wisebabel.com"
+  # s.rdoc_options = ["--charset=UTF-8"]  # "-e UTF-8" is now Default...
+
+  # s.require_paths = ["lib"]	# Default "lib"
+  s.required_ruby_version = '>= 2.0'  # respond_to_missing?
+  s.test_files = Dir['test/**/*.rb']
+  s.test_files.reject! { |fn| File.symlink? fn }
+  # s.requirements << 'libmagick, v6.0' # Simply, info to users.
+  # s.rubygems_version = %q{1.3.5}      # This is always set automatically!!
+
+  ## cf. https://guides.rubygems.org/specification-reference/#metadata
+  s.metadata["yard.run"] = "yri" # use "yard" to build full HTML docs.
+  # s.metadata["changelog_uri"]     = "https://github.com/masasakano/slim_string/blob/master/ChangeLog"
+  # s.metadata["source_code_uri"]   = "https://github.com/masasakano/slim_string"
+  # s.metadata["documentation_uri"] = "https://www.example.info/gems/bestgemever/0.0.1"
+end
+

data/test/test_pandoc_refeq_mathml.rb

@@ -0,0 +1,188 @@
+# -*- encoding: utf-8 -*-
+#
+# Usage(at-Parent-directory): RUBYLIB=$RUBYLIB:./lib ruby test/test_pandoc_refeq_mathml.rb
+
+require 'open3'
+require 'tempfile'
+require 'diff/lcs'
+require 'diff/lcs/string'
+
+$stdout.sync=true
+$stderr.sync=true
+# print '$LOAD_PATH=';p $LOAD_PATH
+arlibrelpath = []
+arlibbase = %w(pandoc_refeq_mathml)	# Write those that are loaded inside a library to make their absolute paths be displayed.
+
+arlibbase.each do |elibbase|
+  arAllPaths = []
+  er=nil
+  pathnow = nil
+  (['../lib/', 'lib/', ''].map{|i| i+elibbase+'/'} + ['']).each do |dir|
+    begin
+      s = dir+File.basename(elibbase)
+      arAllPaths.push(s)
+      require s
+      pathnow = s
+      break
+    rescue LoadError => er
+    end
+  end	# (['../lib/', 'lib/', ''].map{|i| i+elibbase+'/'} + '').each do |dir|
+
+  if pathnow.nil?
+    warn "Warning: All the attempts to load the following files have failed.  Abort..."
+    warn arAllPaths.inspect
+    warn " NOTE: It may be because a require statement in that file failed, 
+rather than requiring the file itself.
+ Check with  % ruby -r#{File.basename(elibbase)} -e p
+ or maybe add  env RUBYLIB=$RUBYLIB:`pwd`"
+    # p $LOADED_FEATURES.grep(/#{Regexp.quote(File.basename(elibbase)+'.rb')}$/)
+    raise er
+  else
+#print pathnow," is loaded!\n"
+    arlibrelpath.push pathnow
+  end
+end	# arlibbase.each do |elibbase|
+
+print "NOTE: Library relative paths: "; p arlibrelpath
+print "NOTE: Library full paths:\n"
+arlibbase.each do |elibbase|
+  p $LOADED_FEATURES.grep(/#{Regexp.quote(File.basename(elibbase)+'.rb')}$/)
+end
+
+
+
+#################################################
+# Unit Test
+#################################################
+
+gem "minitest"
+# require 'minitest/unit'
+require 'minitest/autorun'
+# MiniTest::Unit.autorun
+
+class TestUnitPandocRefeqMathml < MiniTest::Test
+  T = true
+  F = false
+
+  def setup
+    @exefile  = __dir__ + "/../bin/pandoc_refeq_mathml"
+    @auxfile  = __dir__ + "/data/try01_latex.aux"
+    @htmlfile = __dir__ + "/data/try01.html"
+
+    # Array of IOs for temporary files (automatically set in generate_tmpfile())
+    @tmpfiles = []
+  end
+
+  def teardown
+    @tmpfiles.each do |ef|
+      ef.close if !ef.closed? 
+      File.unlink(ef.path)
+    end
+  end
+
+  # @option root [#to_s] Root-name of the temporary filename
+  def generate_tmpfile(root=File.basename($0))
+    io_tmpfile = Tempfile.open(root.to_s)
+    $stderr.print "TEST: Tmpfile="+io_tmpfile.path if ENV.key?('PRINT_TMPFILE') # To display Filename (NOTE the file will be removed when the script ends anyway.)
+    @tmpfiles.push io_tmpfile
+    [io_tmpfile, io_tmpfile.path]
+  end
+
+  def test_pandoc_refeq_mathml
+    auxstr = File.read @auxfile
+    htmlstr = File.read @htmlfile
+    page00 = Nokogiri::HTML(htmlstr)
+    page   = Nokogiri::HTML(htmlstr)
+    io_tmp, path_tmp = generate_tmpfile(__method__)
+
+    prm = PandocRefeqMathml.new page, auxstr, logio: io_tmp, is_verbose: true
+    prm.alter_html!
+
+    # an Equation (LaTeX: \begin{equation})
+    math1_org = page00.css("math:first-of-type")[0]
+    math1_rev = prm.page.css("math:first-of-type")[0]
+    lcs = math1_org.to_s.diff(math1_rev.to_s)
+    assert_equal 1, lcs.size, 'Diff-size should be 1 (one continuous addition only)'
+    assert_operator 90, '<', lcs[0].size, 'Number of different characters should be larger than 90'
+    assert_operator 99, '>', lcs[0].size, 'Number of different characters should be smaller than 99: Diff='+mk_str_diff_chg(lcs).inspect  # "mrow><mtext id=\"square_pm\" style=\"padding-left:1em; text-align:right;\">(36)</mtext></mrow><"
+    assert((%r@</mtext>@ !~ math1_org.to_s), '</mtext> should not be included')
+    assert_match(%r@</mtext></mrow></mrow>@, math1_rev.to_s, '</mtext></mrow></mrow> should be included')
+
+    mtds = prm.page.css("math mtable mtr")[2].css("mtd")
+    assert_equal "right",  mtds[0]["columnalign"]
+    assert_equal "center", mtds[1]["columnalign"], "align should be center: "+mtds[1]
+    assert_equal "left",   mtds[2]["columnalign"]
+    # NOTE: --no-fixalign is tested in test_integrated()
+
+    io_tmp.rewind
+    msg_log = io_tmp.read
+    assert_match(%r@label=.?sec_@, msg_log, "Warning message should be present in the log file because Equation-ID is not found for a label for a Section: \n> "+msg_log)
+  end
+
+  # Integrated tests
+  def test_integrated
+    #com = sprintf "%s --aux=%s --log=%s ", @exefile, @auxfile, @logfilename
+    com = sprintf "%s --aux=%s --no-fixalign", @exefile, @auxfile  # Logfile => STDERR, fixalign=no
+
+    ## From STDIN, out to STDOUT, log-file to STDERR
+    out, err, stat = Open3.capture3(com, stdin_data: File.read(@htmlfile))
+    assert_equal 0, stat
+    assert_match(%r@label=.?sec_@, err, "Warning message should be present in STDERR because Equation-ID is not found for a label for a Section: \n> "+err)
+    assert_operator 5, '<=', out.scan(%r@(?=</mtext>)@).count, 'There should be many </mtext>. out[0..100]='+out[0..100]
+    assert_match(%r@\bcolumnalign="right"@,  out, 'Sanity check columnalign')
+    refute_match(%r@\bcolumnalign="center"@, out, 'With --no-fixalign center columnalign should not exist, but..')
+  end
+
+  # Read a 2-dim Array of Diff::LCS::Change and convert it to a single Array of them
+  #
+  # Each Diff may (or usually) have more than 1 character.
+  # And therefore, it should be far more readable for humans.
+  # Here is an example.
+  #
+  #   # [#<Diff::LCS::Change: ["+", 1, "x"]>, #<Diff::LCS::Change: ["+", 2, "y"]>]
+  #   # => <Diff::LCS::Change: ["+", 1, "xy"]>
+  #
+  # You can still patch it.
+  #
+  #   s2 == s1.patch( [mk_str_diff_chg(Diff::LCS.diff(s1, s2))] )
+  #
+  # However, +s2.unpatch [mk_str_diff_chg(...)]+ raises RuntimeError.
+  # I think it works by starting from the beginning, swapping "`+`" and "`-`",
+  # where interpreting "`-`"+0 as inserting before pos=0 and "`+`"+1 as deleting after pos=1.
+  #
+  # @param arlcs [Array<Array<<Diff::LCS::Change>>]
+  # @return [Array<Diff::LCS::Change>]
+  def mk_str_diff_chg(ar2lcs)
+    arlcs = []  # ar2lcs.flatten actually also flattens the contents of Diff::LCS::Change !
+                # Therefore, this is a custom Array#flatten
+    ar2lcs.each do |ea1|
+      ea1.each do |ea2|
+        arlcs.push ea2
+      end
+    end
+
+    return arlcs if arlcs.empty?
+
+    pos = pos_ini = arlcs[0].position - 99
+    strdiff       = nil
+    action_now    = nil
+
+    arret = []
+    arlcs.each do |ed|  # ed: EachDiff
+      if (pos != ed.position - 1) || (ed.action != action_now)
+        # The previous series has ended.
+        arret.push Diff::LCS::Change.new(action_now, pos_ini, strdiff) if action_now # unless the very first one
+        pos = pos_ini = ed.position
+        strdiff       = ed.element.dup
+        action_now    = ed.action
+        next
+      end
+
+      pos = ed.position
+      strdiff << ed.element
+    end
+    arret.push Diff::LCS::Change.new(action_now, pos_ini, strdiff)
+    arret
+  end
+end	# class TestUnitPandocRefeqMathml < MiniTest::Test
+

metadata

@@ -0,0 +1,86 @@
+--- !ruby/object:Gem::Specification
+name: pandoc_refeq_mathml
+version: !ruby/object:Gem::Version
+  version: '0.1'
+platform: ruby
+authors:
+- Masa Sakano
+autorequire:
+bindir: bin
+cert_chain: []
+date: 2022-08-26 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: nokogiri
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.13'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.13'
+- !ruby/object:Gem::Dependency
+  name: diff-lcs
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.5'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.5'
+description: 'Add equation numbers in a crude way to a pandoc-output MathML converted
+  from LaTeX, utilising its LaTeX aux file, and also adjust math-table alignments.
+
+  '
+email:
+executables: []
+extensions: []
+extra_rdoc_files:
+- README.en.rdoc
+files:
+- ".gitignore"
+- ChangeLog
+- LICENSE.txt
+- Makefile
+- README.en.rdoc
+- Rakefile
+- bin/pandoc_refeq_mathml
+- lib/pandoc_refeq_mathml.rb
+- pandoc_refeq_mathml.gemspec
+- test/test_pandoc_refeq_mathml.rb
+homepage: https://www.wisebabel.com
+licenses:
+- MIT
+metadata:
+  yard.run: yri
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '2.0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.3.7
+signing_key:
+specification_version: 4
+summary: Add equation numbers to a pandoc-output MathML converted from LaTeX
+test_files:
+- test/test_pandoc_refeq_mathml.rb