Method: ConverterBase#insert_word_separator

Defined in:
lib/converterbase.rb

#insert_word_separator(str) ⇒ Object

単語単位でzwsを挿入する



1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
 
# File 'lib/converterbase.rb', line 1164

def insert_word_separator(str)
  buffer = ""
  ss = StringScanner.new(str)
  before_symbol = false

  if @text_type == "textfile"
    buffer << ss.scan(/(.+\n){2}/)
  end

  while char = ss.getch
    symbol = false
    case char
    when "｜"
      ss.scan(/.+?》/)
    when "［"
      buffer << char
      if ss.scan(/^＃.+?］/)
        buffer << "#{ss.matched}"
      else
        before_symbol = false
      end
      next
    when "<"
      if ss.scan(/.+?>/)
        buffer << "<#{ss.matched}"
        next
      end
      symbol = true
    when /[\d０-９]/
      ss.scan(/[\d０-９]+/)
    when /[ぁ-んゝゞ]/
      ss.scan(/[ぁ-んゝゞー]+/)
    when /[ァ-ヶ]/
      ss.scan(/[ァ-ヶー・]+/)
    when /[Ａ-Ｚａ-ｚA-Za-z]/
      ss.scan(/[Ａ-Ｚａ-ｚA-Za-z ]+/)
    when /[一-龥朗-鶴]/
      ss.scan(/[一-龥朗-鶴]+/)
    when /[〔「『\(（【〈《≪〝]/
      buffer << char
      before_symbol = false
      next
    else
      symbol = true
    end
    if before_symbol && !symbol
      buffer << WORD_SEPARATOR
    end
    buffer << char
    unless symbol
      buffer << ss.matched if ss.matched?
      buffer << WORD_SEPARATOR
    end
    before_symbol = symbol
  end
  buffer
end