rj-action-library/Runtime/Text/JSON/JSONLexer.cs

using System.Collections;
using System.Collections.Generic;

using System.Text;




namespace Rokojori
{

  public enum JSONLexerEventType
  {
    DONE_SUCCESS,
    DONE_ERROR,
    
    OBJECT_START,
    OBJECT_END,
    ARRAY_START,
    ARRAY_END,
    IDENTIFIER,
    STRING,
    NUMBER,
    TRUE,
    FALSE,
    NULL,
    ARRAY_SEPERATOR,
    IDENTIFIER_SEPERATOR,


    ERROR_UNEXPECTED_SYMBOL,
    ERROR_UNCLOSED_STRING,
    ERROR_UNCLOSED_OBJECT,
    ERROR_UNCLOSED_ARRAY,
    ERROR_EXPECTED_FALSE,
    ERROR_EXPECTED_TRUE,
    ERROR_EXPECTED_NULL,
    ERROR_INFINITY_PREVENTION
  }

  public class JSONLexer
  {
    public static bool IsErrorType( JSONLexerEventType type )
    {
      switch( type )
      {
        case JSONLexerEventType.ERROR_UNEXPECTED_SYMBOL:
        case JSONLexerEventType.ERROR_UNCLOSED_STRING:
        case JSONLexerEventType.ERROR_UNCLOSED_OBJECT:
        case JSONLexerEventType.ERROR_UNCLOSED_ARRAY:
        case JSONLexerEventType.ERROR_EXPECTED_FALSE:
        case JSONLexerEventType.ERROR_EXPECTED_TRUE:
        case JSONLexerEventType.ERROR_EXPECTED_NULL:
        case JSONLexerEventType.ERROR_INFINITY_PREVENTION:
        case JSONLexerEventType.DONE_ERROR:
        {
          return true;
        }
      }

      return false;
    }

    public delegate void OnParse( JSONLexerEventType type, int offset, int length );
    
    string source;
    OnParse onParseCallback;

  

    public void Lex( string source, OnParse onParseCallback = null )
    {
      this.source = source;
      this.onParseCallback = onParseCallback;

      Call( LexJSON(), 0 );

    }

    void Call( JSONLexerEventType type, int offset, int length = -1 )
    {
      if ( onParseCallback != null )
      {
        onParseCallback( type, offset, length );
      }
    }

    

    JSONLexerEventType LexJSON()
    {
      var offset = SkipWhiteSpace( 0 );
      
      while ( offset < source.Length )
      {
        var offsetStart = offset;

        var character = source[ offset ];

        if ( character == '{' )
        {
          Call( JSONLexerEventType.OBJECT_START, offset , 1 );
        }
        else if ( character == '}' )
        {
          Call( JSONLexerEventType.OBJECT_END, offset , 1 );
        }
        else if ( character == '[' )
        {
          Call( JSONLexerEventType.ARRAY_START, offset , 1 );
        }
        else if ( character == ']' )
        {
          Call( JSONLexerEventType.ARRAY_END, offset , 1 );
        }
        else if ( character == ',' )
        {
          Call( JSONLexerEventType.ARRAY_SEPERATOR, offset , 1 );
        }
        else if ( character == ':' )
        {
          Call( JSONLexerEventType.IDENTIFIER_SEPERATOR, offset , 1 );          
        }
        else if ( character == '\"' )
        {
          offset = LexString( offset );
        }
        else if ( character == 't' )
        {
          offset = LexTrue( offset );
        }
        else if ( character == 'f' )
        {
          offset = LexFalse( offset );
        }
        else if ( character == 'n' )
        {
          offset = LexNull( offset );
        }
        else if ( System.Char.IsDigit( character ) || character == '-' )
        {
          offset = LexNumber( offset );
        }
        
        if ( offset == -1 )
        {
          return JSONLexerEventType.DONE_ERROR;
        }

        offset ++;
        offset = SkipWhiteSpace( offset );
        
        if ( offsetStart >= offset )
        {
          Call( JSONLexerEventType.ERROR_INFINITY_PREVENTION, offsetStart );
          return JSONLexerEventType.DONE_ERROR;
        }
      }
      
      return JSONLexerEventType.DONE_SUCCESS;
    }
    
    int SkipWhiteSpace( int offset )
    {

      while ( offset < source.Length && System.Char.IsWhiteSpace( source[ offset ] ) )
      {
        offset ++;
      }

      return offset;
    }

    bool IsIdentifier( int offset )
    {
      offset = SkipWhiteSpace( offset );
      return offset < source.Length && source[ offset ] == ':'; 
    }

    int LexString( int offset )
    {
      var offsetStart = offset;
      offset++;

      var maxTries = 10000; var currentTries = 0;      
      // currentTries ++; if ( currentTries == maxTries ) { throw new System.Exception(); }

      while ( offset < source.Length )  
      {
        currentTries ++; if ( currentTries == maxTries ) { throw new System.Exception(); }
        
        var character = source[ offset ];

        if ( character == '\"' )
        {
          var previousCharacter = source[ offset - 1 ];

          if ( previousCharacter != '\\' )
          {
            var length = ( offset - offsetStart ) + 1;
            var isIdentifier = IsIdentifier( offset + 1 );
            var type = isIdentifier ?  JSONLexerEventType.IDENTIFIER :
                                       JSONLexerEventType.STRING;
            Call( type, offsetStart, length );
            return ( offsetStart + length ) - 1;
          }
        }
        
        offset ++;

      }

      Call( JSONLexerEventType.ERROR_UNCLOSED_STRING, offsetStart );  
      return -1;
    }

    int LexTrue( int offset )
    {
      if ( MatchNextThree( offset, 'r', 'u', 'e' ) )
      {
        Call( JSONLexerEventType.TRUE, offset, 4 );
        return offset + 3;
      }

      Call( JSONLexerEventType.ERROR_EXPECTED_TRUE, offset );
      return -1;
    }

    int LexFalse( int offset )
    {
      if ( MatchNextFour( offset, 'a', 'l', 's', 'e' ) )
      {
        Call( JSONLexerEventType.FALSE, offset, 5 );
        return offset + 4;
      }
      Call( JSONLexerEventType.ERROR_EXPECTED_FALSE, offset );
      return -1;
    }

    int LexNull( int offset )
    {
      if ( MatchNextThree( offset, 'u', 'l', 'l' ) )
      {
        Call( JSONLexerEventType.NULL, offset, 4 );
        return offset + 3;
      }

      Call( JSONLexerEventType.ERROR_EXPECTED_NULL, offset );

      return -1;
    }

    int LexNumber( int offset )
    {
      var hasDot = false;
      var hasE = false;
      var hasMinus = false;
      var parsing = true;
      
      var offsetStart = offset;
      
      var maxTries = 100; var currentTries = 0;      
      // currentTries ++; if ( currentTries == maxTries ) { throw new System.Exception(); }

      while ( parsing && offset < source.Length )
      {
        currentTries ++; if ( currentTries == maxTries ) { throw new System.Exception(); }
        var character = source[ offset ];
        
        parsing = false;

        if ( System.Char.IsDigit( character ) )
        {
          parsing = true;
        }
        else if ( character == '-' && ! hasMinus )
        {
          hasMinus = true;
          parsing = true;
        }
        else if ( character == '.' && ! hasDot )
        {
          hasDot = true;
          parsing = true;
        }
        else if ( ( character == 'e' || character == 'E' ) && ! hasE )
        {
          var hasNext = offset + 2 < source.Length;

          if ( hasNext )
          {
            var nextCharacter = source[ offset + 1 ];
            hasE = nextCharacter == '+' || nextCharacter == '-';
            parsing = hasE;
          }

          if ( parsing )
          {
            offset ++;
          }

        }

        if ( parsing )
        {
          offset ++;
        }

      }

      var length = ( offset - offsetStart );
      Call( JSONLexerEventType.NUMBER, offsetStart, length );

      return ( offsetStart + length ) - 1;
    }

    bool MatchNextThree( int offset, char next, char second, char third )
    {
      if ( offset + 3 >= source.Length )   { return false; }
      if ( source[ offset + 1 ] != next )  { return false; }
      if ( source[ offset + 2 ] != second ){ return false; }
      if ( source[ offset + 3 ] != third ) { return false; }

      return true;
    }

    bool MatchNextFour( int offset, char next, char second, char third, char fourth )
    {
      if ( offset + 4 >= source.Length )   { return false; }
      if ( source[ offset + 1 ] != next )  { return false; }
      if ( source[ offset + 2 ] != second ){ return false; }
      if ( source[ offset + 3 ] != third ) { return false; }
      if ( source[ offset + 4 ] != fourth ){ return false; }

      return true;
    }
   

  }
}