hwp_script_to_latex 1.4.0 → 1.5.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8a1f7260a320ad6b3be51ded4f0a61b2da7d0143ed340715e37a0ebdcf89f2e1
-  data.tar.gz: 045d8409baaeff28e2803f13f7c5fe152e8f7100749d40a253fef6f2f21f6467
+  metadata.gz: 5e948d02ae83f17c99fa96f38d16c3d0b257e7aa0f7ba4cea964fc300bb24975
+  data.tar.gz: '0825cfac358661df69d6cfa0f487a89428129ffa60c1af6f8e4b182162ea5e8f'
 SHA512:
-  metadata.gz: 020d0e2857c38dbf1aa4ae5051150631bcb4033f075fef06c2229923fce601ca74c9d3058ccaa5e106b3eb2c923528af7b3993802681997a4869d29087333de5
-  data.tar.gz: 5addae0db39cea048c66807b8f1141a31c8206f63dd3788a995bedb3bfb59331096463924eca23f3e2c1c8ba934fab1df9cb007fa5457c185c9e7f0369627b07
+  metadata.gz: 35732e0afd39f4818ce2aab81458fa7366202631bea5800391df1954ba0f850d7631f57276273126ef6b293269d44239ffde63e155804629f100962b21ab7432
+  data.tar.gz: f7647a86a3de1f04b9e503b0b34300824c29f2ce535c7b109aec3bfe1600ff37fa9edb7a504f96d11cda322697b9a9a0249587d4ff459d0b51c1825e045f3adc

data/lib/hwp_script_to_latex/converter.rb

@@ -1,33 +1,34 @@
 require 'json'
+require 'hwp_script_to_latex/syntax'
+require 'hwp_script_to_latex/processor'
 
 module HwpScriptToLatex
   # 한글 수식스크립트를 LaTeX 문법으로 변환
   class Converter
+    include Syntax
+
     # 명령어에 사용되는 좌, 우항 정규표현식
     LEFT_TERM_REGEX = "(?:[^{}\\s`]{9})*?\\K(?<%s>{(?>[^{}]+|(?:\\g<%s>))*}|(?:[^{}\\s`]{0,9}))"
     RIGHT_TERM_REGEX = "(?<%s>{(?>[^{}]+|(?:\\g<%s>))*}|(?:[^{}\\s`]{0,9}))"
 
     def initialize
-      f = File.open(File.join(File.dirname(__FILE__), '..', '..', 'rules.json'), 'r')
-      file_content = f.read
-      f.close
-
-      @rules= JSON.parse(file_content, symbolize_names: true)
-      @simple_commands = @rules[:command][:simple].sort_by { |hash| -1 * hash[:script].length }
-      @block_commands = @rules[:command][:block]
-      @meta_keywords = @rules[:keyword][:meta].sort_by { |hash| -1 * hash[:regex].inspect.length }
-      @command_keywords = @rules[:keyword][:command].sort_by { |hash| -1 * hash[:regex].inspect.length }
-      @symbol_keywords = @rules[:keyword][:symbol].sort_by { |hash| -1 * hash[:regex].inspect.length }
-      @reserved_keywords = @rules[:keyword][:reserved].sort_by { |hash| -1 * hash[:regex].inspect.length }
+      @processor = Processor.new
+      @keyword_commands = KEYWORD_COMMANDS
+      @default_commands = DEFAULT_COMMANDS
+      @block_commands = BLOCK_COMMANDS
+
+      @meta = META.sort_by { |hash| -1 * hash[:regex].inspect.length }
+      @symbols = SYMBOL.sort_by { |hash| -1 * hash[:regex].inspect.length }
+      @reserved_words = RESERVED_WORD.sort_by { |hash| -1 * hash[:regex].inspect.length }
     end
 
     def convert(script, math_mode: false, display_mode: false)
-      # Sanitize for starting convert
-      result = pre_sanitize(script)
-      # 단순 치환 키워드
-      result = replace_keywords(result)
-      # 1개의 우항을 가지는 간단한 명령어
-      result = replace_simple_commands(result)
+      # Data pre processing
+      result = @processor.pre_process(script)
+      # 파라미터가 없는 명령어
+      result = replace_keyword_commands(result)
+      # 1개의 우항을 가지는 명령어
+      result = replace_default_commands(result)
       # 행렬, 케이스등 블록 명령어
       #   cases {...} => \begin{cases}...\end{cases}
       #   dmatrix {...} => \begin{vmatrix}...\end{vmatrix}
@@ -45,11 +46,14 @@ module HwpScriptToLatex
       result = replace_sqrt(result) # Case 1
       result = replace_fractions(result) # Case 2
 
+      # 단순 치환 키워드
+      result = replace_keywords(result)
+
       # 전체 수식에 디스플레이 스타일 적용
       result = decorate_displaystyle(result) if display_mode
 
-      # Sanitize result for removing dirty spaces
-      result = post_sanitize(result)
+      # Data post processing
+      result = @processor.post_process(result)
 
       # Math mode
       result = "$#{result}$" if math_mode
@@ -59,37 +63,40 @@ module HwpScriptToLatex
 
     private
 
-    # 변환 전 sanitize
-    def pre_sanitize(script)
-      # 족보닷컴 텍스트 제거
-      jokbo_regex = %r(from\s*=+\s*(?:족보닷컴[\s\S]*?)=+)
-      script = script.gsub(jokbo_regex, "")
-      # 2개 이상의 공백을 하나의 공백으로 치환
-      script = script.gsub(/\s+/, " ").strip
-      # 백슬래시(로만체)를 삭제
-      script = script.gsub(/\\/, "")
-      # 꺽쇠 치환
-      script = script.gsub(/&lt;/, "<")
-      script = script.gsub(/&gt;/, ">")
-      # 위, 아래 첨자 명령어로 변경
-      script = script.gsub(/_/, " sub ")
-      script = script.gsub(/\^/, " sup ")
+    def replace_keyword_commands(script)
+      commands = @keyword_commands
+      matched_count = 0
+      
+      # 1star(2star(3starcdotsstar(99star100))cdots)
+      # 위와 같은 경우 이중 루프 cdots 부터 매치되기 시작하면 star 때문에
+      # 제대로 변환되지 않음.
+      # 그러므로 전체 키워드를 검사하는 루프(loop 2)를
+      # 이중 루프로 변환이 완료될 때 까지 반복함(loop 1)
+      # 속도는 느려지겠지만 대안을 찾을 때 까지 이대로 사용하려고 함
+      loop do # loop 1
+        matched_count = 0
+        commands.each do |command|
+          command_regex = rule_regex(command)
+          before_script = script
+          script = script.gsub(command_regex, command[:latex])
+          while before_script != script # loop 2
+            matched_count += 1
+            before_script = script
+            script = script.gsub(command_regex, command[:latex])
+          end
+        end
+        break if matched_count == 0
+      end
 
       return script
     end
 
-    # 변환 후 sanitize
-    def post_sanitize(script)
-      # 2개 이상의 공백을 하나의 공백으로 치환
-      script = script.gsub(/\s+/, " ").strip
-    end
-
-    def replace_simple_commands(script)
+    def replace_default_commands(script)
       right_term_group_name = "rt"
 
-      @simple_commands.each do |command|
+      @default_commands.each do |command|
         right_term_regex = RIGHT_TERM_REGEX % [right_term_group_name, right_term_group_name]
-        command_regex = %r((?<![a-zA-Z\\])(?i:#{command[:script]})\s*#{right_term_regex})
+        command_regex = %r((?<![a-zA-Z\\])(?i:#{command[:regex]})\s*#{right_term_regex})
 
         match_data = script.match(command_regex)
         while match_data
@@ -106,7 +113,7 @@ module HwpScriptToLatex
 
     def replace_block_commands(script)
       @block_commands.each do |command|
-        command_regex = %r((?<![a-zA-Z])(?i:#{command[:script]})\s*(?<block_content>{(?>[^{}]+|(?:\g<block_content>))*}))
+        command_regex = %r((?<![a-zA-Z])(?i:#{command[:regex]})\s*(?<block_content>{(?>[^{}]+|(?:\g<block_content>))*}))
 
         match_data = script.match(command_regex)
         while match_data
@@ -123,7 +130,7 @@ module HwpScriptToLatex
     end
 
     def replace_keywords(script)
-      keywords = @meta_keywords + @command_keywords + @symbol_keywords + @reserved_keywords
+      keywords = @meta + @symbols + @reserved_words
       matched_count = 0
       
       # 1star(2star(3starcdotsstar(99star100))cdots)
@@ -135,7 +142,7 @@ module HwpScriptToLatex
       loop do # loop 1
         matched_count = 0
         keywords.each do |keyword|
-          keyword_regex = get_keyword_regex(keyword)
+          keyword_regex = rule_regex(keyword)
           before_script = script
           script = script.gsub(keyword_regex, keyword[:latex])
           while before_script != script # loop 2
@@ -201,22 +208,6 @@ module HwpScriptToLatex
       return script
     end
 
-    def get_keyword_regex(keyword)
-      regexes = []
-      if keyword[:regex].class == Array
-        regexes = keyword[:regex]
-      else
-        regexes << keyword[:regex]
-      end
-
-      is_alphabetic = keyword[:alphabetic].nil? ? true : keyword[:alphabetic]
-      if is_alphabetic
-        return %r((?<![a-zA-Z\\])(#{regexes.join('|')}))
-      else
-        return %r((?<![\\])(#{regexes.join('|')}))
-      end
-    end
-
     def decorate_displaystyle(script)
       script = script.gsub(/\\sum/, "\\displaystyle \\sum")
       script = script.gsub(/\\int/, "\\displaystyle \\int")

data/lib/hwp_script_to_latex/processor.rb

@@ -0,0 +1,34 @@
+#
+#  Author: osh
+#  Created: 2019-05-27
+#  Last modified: 2019-05-27
+
+module HwpScriptToLatex
+  class Processor
+
+    # 수식 문자열 전처리
+    def pre_process(script)
+      # 족보닷컴 텍스트 제거
+      jokbo_regex = %r(from\s*=+\s*(?:족보닷컴[\s\S]*?)=+)
+      script = script.gsub(jokbo_regex, "")
+      # 2개 이상의 공백을 하나의 공백으로 치환
+      script = script.gsub(/\s+/, " ").strip
+      # 백슬래시(로만체)를 로만체 명령어로 변환
+      script = script.gsub(/^\\| \\/, " \\rm ")
+      # 꺽쇠 치환
+      script = script.gsub(/&lt;/, "<")
+      script = script.gsub(/&gt;/, ">")
+      # 위, 아래 첨자 명령어로 변경
+      script = script.gsub(/_/, " sub ")
+      script = script.gsub(/\^/, " sup ")
+
+      return script
+    end
+
+    # 수식 문자열 후처리
+    def post_process(script)
+      # 2개 이상의 공백을 하나의 공백으로 치환
+      script = script.gsub(/\s+/, " ").strip
+    end
+  end
+end