aedict-notepad-to-anki.rb

#!/usr/bin/env ruby

require 'pathname'
require 'uri'


module HTML
  def table(txt)
    node('table', txt)
  end

  def tr(txt)
    node('tr', txt)
  end

  def td(txt)
    node('td', txt)
  end

  def ul(txt)
    node('ul', txt)
  end

  def li(txt)
    node('li', txt)
  end

  def em(txt)
    node('em', txt)
  end

  def a(txt, attrs)
    node('a', txt, attrs)
  end

  def nbsp
    " "
  end

  def br
    "<br/>"
  end

  def node(tag, txt, attrs={})
    attrs = attrs.empty? ? "" : " #{attrs.map {|k,v| %[#{k}="#{v}"]}.join(' ')}"
    "<#{tag}#{attrs}>#{txt}</#{tag}>"
  end
end

AEDICT_KEYWORDS = %w[gikun go ik jouyou kan kanyou kun name oik ok on rad tou uK ateji iK ik io oK gA gai ichi news nf spec abbr aphorism arch chn col company derog eK fam fem given hon hum id joc m-sl male obs obsc on-mim organization person place poet pol product proverb quote rare sens sl station surname uk unclass vulg work X yoji adj-f adj-i adj-ix adj-kari adj-ku adj-na adj-nari adj-no adj-pn adj-shiku adj-t adv adv-to aux aux-adj aux-v conj cop-da ctr exp int n n-adv n-pr n-pref n-suf n-t num pn pref prt suf unc v-unspec v1 v1-s v2a-s v2b-k v2b-s v2d-k v2d-s v2g-k v2g-s v2h-k v2h-s v2k-k v2k-s v2m-k v2m-s v2n-s v2r-k v2r-s v2s-s v2t-k v2t-s v2w-s v2y-k v2y-s v2z-s v4b v4g v4h v4k v4m v4n v4r v4s v4t v5aru v5b v5g v5k v5k-s v5m v5n v5r v5r-i v5s v5t v5u v5u-s v5uru vi vk vn vr vs vs-c vs-i vs-s vt vz anat archit astron baseb biol bot Buddh bus chem comp econ engr finc food geol geom law ling MA mahj math med mil music physics Shinto shogi sports sumo zool hob ksb ktb kyb kyu nab osb rkb std thb tsb tsug eng afr ain alg amh ara arn bnt bre bul bur chi chn cze dan dut epo est fil fin fre geo ger glg grc gre haw heb hin hun ice ind ita khm kor kur lat mal mao may mnc mol mon nor per pol por rum rus san scr slo slv som spa swa swe tah tam tha tib tur urd vie yid equ expl fig lit jmdict jmnedict test A D R ant cf ex kvar pref see syn uses]

class AedictNotepad
  def self.to_anki_tsv(filepath)
    Pathname.new(filepath).read
      .sub(/^\[(\w|\s)+\]\n?$/,'')
      .split(/\n/)
      .reject(&:empty?)
      .map {|line| AnkiCard.new(entry: Entry.parse(line)).tab_separated}
      .join("\n")
  end

  class Entry
    %i[kanji kana common defns].each do |sym| attr_reader sym end
    def initialize(kanji:, kana:, common:, defns:)
      @kanji  = kanji
      @kana   = kana
      @common = common
      @defns  = defns
    end

    def self.parse(str)
      line_re = %r{^(.+?)\s*\[(.+)\]:\s*(.+?)([(]P[)])?$}
      multi_gloss_initial_re = %r{([(][^/]+?[)] )?[(]\d+[)]}

      m = str.match line_re

      raw_defns = m[3].gsub(multi_gloss_initial_re, "\n\\0").sub(/^\n/,'').split("\n").map {|g| g.chomp '/'}
      defns =
        if raw_defns.length == 1
          [ Definition.parse(Definition::SINGLE_RE, raw_defns.first, num: 1) ]
        else
          raw_defns.map {|d| Definition.parse Definition::MULTI_RE, d}
        end

      new \
        kanji:    m[1] == "null" ? nil : m[1],
        kana:     m[2],
        common: !!m[4],
        defns:     defns
    end

    class Definition
      %i[num glosses pos tags].each do |sym| attr_reader sym end
      def initialize(num:, glosses:, pos:, tags:)
        @num     = num.nil? ? 1 : num.to_i
        @glosses = glosses
        @pos     = pos
        @tags    = tags
      end

      SINGLE_RE = %r{([(](?<pos>[^)]+?)[)] )?(?<rest>.+)}
      MULTI_RE  = %r{([(](?<pos>[^)]+?)[)] )?[(](?<num>\d+)[)] (?<rest>.+)}

      def self.parse(regex, str, num: nil)
        m1 = str.match regex
        m2 = m1[:rest].match %r{(?<parens>([(](#{AEDICT_KEYWORDS.join('|')})[)] )+)?(?<glosses>.+)}

        tags = m2[:parens]&.scan(%r{[(]([^)]+)[)]}).flatten.map{|t| t.split(',')}.flatten if m2 && m2[:parens]

        new \
          num:     num || m1[:num],
          pos:     m1[:pos]&.split(','),
          tags:    tags,
          glosses: m2[:glosses].split('/')
      end
    end
  end
end

class AnkiCard
  include HTML

  attr_reader :entry
  def initialize(entry:)
    @entry = entry
  end

  def tab_separated
    "#{front}\t#{back}"
  end

  def front
    entry.kanji || entry.kana
  end

  def back
    readings_and_common = [
      entry.kanji,
      "[#{entry.kana}]",
      (entry.common ? em("(P)") : nil),
    ].reject(&:nil?).join(nbsp)

    lines = [
      readings_and_common,
      defns_table,
      dict_links,
    ].reject(&:nil?).join(br*2)
  end

  def defns_table
    table entry.defns.map {|defn|
      pos  = defn.pos.nil?  ? "" : "(#{defn.pos.join(',')})"
      tags = defn.tags.nil? ? "" : "[#{defn.tags.join(',')}]"

      glosses = ul(defn.glosses.map{|g| li g}.join(''))

      tr [
        td(defn.num),
        td(pos),
        td(tags),
        td(glosses),
      ].join('')
    }.join('')
  end

  def dict_links
    txt = entry.kanji || entry.kana

    [
      a("Jisho",     href: jisho(txt)),
      a("goo",       href: goo(txt)),
      a("Weblio JJ", href: weblio_jj(txt)),
      a("Weblio JE", href: weblio_je(txt)),
    ].join(nbsp)
  end

  def weblio_je(txt)
    "https://ejje.weblio.jp/content/#{URI.encode(txt)}"
  end
  def weblio_jj(txt)
    "https://www.weblio.jp/content/#{URI.encode(txt)}"
  end
  def goo(txt)
    "https://dictionary.goo.ne.jp/srch/jn/#{URI.encode(txt)}/m0u/"
  end
  def jisho(txt)
    "https://jisho.org/search/#{URI.encode(txt)}"
  end
end

input  = Pathname.new(ARGV[0])
output = Pathname.new("#{input.sub_ext ''}-#{Time.now.to_i}.tsv")

output.write AedictNotepad.to_anki_tsv(input)