Defines
#define	STRING(x) ((x)->unode.c[2])
#define	WORD(x) ((x)->unode.a[3])
#define	LENGTH(x) ((x)->unode.u[1] >> 8)
#define	IDLENGTH(x) ((x)->unode.u[1])
#define	IS_WORD(x) (ID(x)==WORD_ID)
#define	IS_SUBWORD(x) (ID(x)==SUBWORD_ID)
#define	IS_ANY_WORD(x) ((ID(x)==WORD_ID)\|\|(ID(x)==SUBWORD_ID))
#define	WORD_SIZE 256256256
	max word length (in chars)
Functions
atom_t	new_word (chars_t string, uint_t length)
	creates a new word atom
atom_t	new_subword (atom_t word, chars_t string, uint_t length)
	creates a new subword atom
atom_t	create_word (int buff_len)
	Creates word with len buff_len and empty data.
void	delete_word (atom_t a)
	deletes word atom
void	delete_subword (atom_t a)
	deletes subword atom
void	dump_word (atom_t a, int level)
	dumps (sub)word atom
atom_t	read_word (chars_t filename)
	reads a word from ASCII, UTF-8 or UTF16-LE encoded file
void	write_word (atom_t word, chars_t filename)
	writes a word to ASCII or UTF-16LE encoded file
int	same_words (atom_t a, atom_t b)
	compares two words
atom_t	decode_word (unsigned char *buffer, int size, int dealc)
	decodes ASCII, UTF-8 or UTF16-LE encoded text
int	atom_to_boolean (atom_t a, int *np)
	convert atom to a boolean
int	atom_to_string (atom_t a, chars_t buff, int *buff_len)
	convert atom to string
atom_t	atom_to_word (atom_t data)
	convert atom to word or subword atom
atom_t	atom_to_real_word (atom_t data)
	convert atom to word
int	same_strings (int ci, chars_t sa, chars_t sb, int n)
	compares two strings
Variables
int	full_print_p
	integer of FULLPRINTP
atom_t	false_true [2]
	array holding words with FALSE and TRUE

Define Documentation

#define STRING ( x ) ((x)->unode.c[2])

#define WORD ( x ) ((x)->unode.a[3])

#define LENGTH ( x ) ((x)->unode.u[1] >> 8)

#define IDLENGTH ( x ) ((x)->unode.u[1])

#define IS_WORD ( x ) (ID(x)==WORD_ID)

#define IS_SUBWORD ( x ) (ID(x)==SUBWORD_ID)

#define IS_ANY_WORD ( x ) ((ID(x)==WORD_ID)||(ID(x)==SUBWORD_ID))

#define WORD_SIZE 256*256*256

Function Documentation

atom_t new_word	(	chars_t	string,
		uint_t	length
	)

Parameters:

string	pointer to the first character of the word
length	length of the word

Returns:: word atom

Creates a word atom with reference count 1. A NULL_CHAR is automatically appended to the end of the word.

{
  atom_t a = (atom_t)take_from_pool( &data_pool );

  if( length==UNKNOWN ) length = STRLEN(string);
  REF(a)      = 1;
  IDLENGTH(a) = WORD_ID | (length<<8);
  STRING(a)   = (chars_t)ALLOC( (size_t)CHAR_SIZE*(length+1) );

  #ifdef SAFEMODE
    assert( length<WORD_SIZE );
    assert( STRING(a) );
  #endif //SAFEMODE

  STRNCPY (STRING(a),string,length);
  *(STRING(a)+length) = NULL_CHAR;

  #ifdef DEBUG_ATOM
    printf("<ATOM>  [%08x] word={"STR"}\n",(int)a,string);
  #endif //DEBUG_ATOM

  #ifdef ADVANCED
    stats[ID(a)].allocs++;
    if( stats[ID(a)].max<(stats[ID(a)].allocs-stats[ID(a)].deallocs) )
      stats[ID(a)].max=(stats[ID(a)].allocs-stats[ID(a)].deallocs);
    stats_free--;
  #endif //ADVANCED

  #ifdef DEBUG_RUNTIME_ATOMS
  if( running_compiled_code )
    {
      outter( TEXT("<RUNTIME> new  "), -1 );
      dump_atom_address( a );
      dump_atom( a, 1 );
      outter( TEXT("\n"), -1 );
    }
  #endif
  #ifdef DEBUG_COMPILETIME_ATOMS
  if( compiling_code )
    {
      outter( TEXT("<COMPILETIME> new  "), -1 );
      dump_atom_address( a );
      dump_atom( a, 1 );
      outter( TEXT("\n"), -1 );
    }
  #endif

  return a;
}

atom_t new_subword	(	atom_t	word,
		chars_t	string,
		uint_t	length
	)

Parameters:

word	word atom containing the host word of the word
string	pointer to the first character of the word
length	length of the word

Returns:: subword atom

Creates a subword atom containing a substring of the string in a word atom. The reference count of the subword is 1, the reference count of the host word atom is incremented.

If word parameter is subword, then it is replaced by its hosting word.

{
  atom_t a = (atom_t)take_from_pool( &data_pool );

  #ifdef SAFEMODE
    assert( word );
    assert( length<WORD_SIZE );
  #endif //SAFEMODE

  REF(a)      = 1;
  IDLENGTH(a) = SUBWORD_ID | (length << 8);
  if( IS_SUBWORD(word) ) word = WORD(word);
  WORD(a)     = USE(word);
  STRING(a)   = string;

  #ifdef DEBUG_ATOM
    printf("<ATOM>  [%08x] subword=",(int)a);
    char_t x;
    x = *(string+length);
    *(string+length) = NULL_CHAR;
    printf("{"STR"}\n",string);
    *(string+length) = x;
  #endif //DEBUG_ATOM

  #ifdef ADVANCED
    stats[ID(a)].allocs++;
    if( stats[ID(a)].max<(stats[ID(a)].allocs-stats[ID(a)].deallocs) )
      stats[ID(a)].max=(stats[ID(a)].allocs-stats[ID(a)].deallocs);
    stats_free--;
  #endif //ADVANCED

  #ifdef DEBUG_RUNTIME_ATOMS
  if( running_compiled_code )
    {
      outter( TEXT("<RUNTIME> new  "), -1 );
      dump_atom_address( a );
      dump_atom( a, 1 );
      outter( TEXT("\n"), -1 );
    }
  #endif
  #ifdef DEBUG_COMPILETIME_ATOMS
  if( compiling_code )
    {
      outter( TEXT("<COMPILETIME> new  "), -1 );
      dump_atom_address( a );
      dump_atom( a, 1 );
      outter( TEXT("\n"), -1 );
    }
  #endif

  return a;
}

atom_t create_word ( int buff_len )

Returns:: new word

Creates new word with given length and no data set.

{
    atom_t word;

#ifdef SAFEMODE
    assert (buff_len < WORD_SIZE);
#endif //SAFEMODE

    word = (atom_t)take_from_pool(&data_pool);

    REF(word)      = 1;
    IDLENGTH(word) = WORD_ID | (buff_len << 8);
    STRING(word)   = (chars_t)ALLOC((size_t)CHAR_SIZE * (buff_len+1));

  #ifdef SAFEMODE
    assert (STRING(word));
  #endif //SAFEMODE

  #ifdef DEBUG_ATOM
    printf("<ATOM>  [%08x] word with len %d \n", (int)word, buff_len);
  #endif //DEBUG_ATOM

  #ifdef ADVANCED
    stats[ID(word)].allocs++;
    stats_free--;
  #endif //ADVANCED

  return word;
}

void delete_word ( atom_t a )

Parameters:

a	atom to delete

Deletes the word atom and deallocates memory used for the word's characters.

{
  DEALLOC( STRING(a) ); 
  return_to_pool( &data_pool, a );
}

void delete_subword ( atom_t a )

Parameters:

a	atom to delete

Deletes the subword atom and unlinks it from the host word.

{
  DEUSE( WORD(a) );
  return_to_pool( &data_pool, a );
}

void dump_word	(	atom_t	a,
		int	level
	)

Parameters:

a	atom to dump
level	dump level

Dumps word or subword atom through the current outter function.

{
  // test for full print
  if( full_print_p )
    {
      // count the number of dangerous characters
      int length = LENGTH(a);
      int dirty = (length == 0);

      chars_t chs;
      for( chs=STRING(a); length && !dirty ; chs++,length-- )
      {
        char_t ch = DEBAR(*chs);
        if( ch==TEXT('[') || ch==TEXT(']') ||
            ch==TEXT('(') || ch==TEXT(')') ||
            ch==TEXT('+') || ch==TEXT('-') ||
            ch==TEXT('*') || ch==TEXT('\\') ||
            ch==TEXT('<') || ch==TEXT('>') ||
            ch==TEXT('=') || ch==TEXT(';') ||
            ch==TEXT('|') || ch==TEXT('~') ||
            ch<=TEXT(' ') )
          dirty = 1;
      }

      if( dirty )
   {
     length = LENGTH(a);
     outter( TEXT("|"), 1 );
     outter_size += 1;
     for( chs=STRING(a); length; chs++,length-- )
       {
         char_t ch = DEBAR(*chs);
         if( ch==TEXT('|') || ch==TEXT('\\') )
      {
        outter( TEXT("\\"), 1 );
        outter_size += 1;
      }
         outter( chs, 1 );
         outter_size += 1;
       }
     outter( TEXT("|"), 1 );
     outter_size += 1;
   }
      else
   {
     outter( STRING(a), LENGTH(a) );
     outter_size += LENGTH(a);
   }
      return;
    }


  // no full print
  int limit = print_width_limit;

  if( 0<=limit && limit<10 ) limit=10;

  if( 0<=limit && limit<LENGTH(a) )
    {
      outter( STRING(a), limit );
      outter( TEXT("..."), 3 );
      outter_size += limit+3;
    }
  else
    {
      outter( STRING(a), LENGTH(a) );
      outter_size += LENGTH(a);
    }
}

atom_t read_word ( chars_t filename )

Parameters:

filename file name

Returns:: word atom containing text from the file

Reads the contents of text file into a word atom. If Unicode mode is switched off, the file is assumed to be in ASCII, otherwise it is UTF-8 or UTF-16LE (little endian) depending on the BOM (byte order mark). In case of error its code is stored in global variable errno.

{
  int filesize;
  unsigned char* buffer = (unsigned char*)load_file( filename, &filesize );

  if( errno ) return new_os_error( filename );

  return decode_word( buffer, filesize, 1 );
}

void write_word	(	atom_t	word,
		chars_t	wfilename
	)

Parameters:

word	word to be written
wfilename	file name

Writes the contents of a word atom to a file determined by its name. If Unicode mode is switched of, the encoding is ASCII, otherwise it is UTF-16LE (little endian) and the BOM (byte order mark) is included. In case of error its code is stored in global variable errno.

{
  FILE* file;
  char* filename = FILENAME(wfilename);

  // open file
  errno = 0;
  file = fopen( filename, "wb" );

  // read file into the buffer
  #ifdef UNICODE_CHARS
    unsigned char buffer[2] = {0xFF,0xFE};
    if( !errno ) fwrite( buffer, 2, 1, file );
  #endif //UNICODE_CHARS
  if( !errno ) fwrite( STRING(word), LENGTH(word), CHAR_SIZE, file );
  if( !errno ) fclose( file );

  DEALLOC( filename );
}

int same_words	(	atom_t	a,
		atom_t	b
	)

Parameters:

a	first word
b	second word

Returns:: 0 if words are different, non zero - otherwise

Compares two word atoms case sensitive or insensitive depending on option_case_insensitive global flag.

{
#ifdef SAFEMODE
  assert( IS_WORD(a) || IS_SUBWORD(a) );
  assert( IS_WORD(b) || IS_SUBWORD(b) );
#endif

  unsigned int n = LENGTH(a);
  if( n!=LENGTH(b) ) return 0; // different sizes

  return same_strings( OPTION_CASE_INSENSITIVE, STRING(a), STRING(b), n );
}

atom_t decode_word	(	unsigned char *	buffer,
		int	size,
		int	dealc
	)

Parameters:

buffer	buffer containing raw data
size	size of buffer
dealc	deallocate (=1) or not (=0) the input buffer

Returns:: word atom containing decoded text

If Unicode mode is switched off, the raw data are assumed to be in ASCII, otherwise they are UTF-8 or UTF-16LE (little endian) depending on the BOM (byte order mark).

BOM signatures are:

UTF-16LE: FF FE
UTF-16BE: FE FF
UTF-8: EF BB BF

The input buffer is deallocated if dealc!=0

{
  #ifdef SAFEMODE
    assert( buffer );
  #endif //SAFEMODE
  atom_t a = (atom_t)take_from_pool( &data_pool );
  REF(a) = 1;

  #ifdef ADVANCED
    stats[WORD_ID].allocs++;
    stats_free--;
  #endif //ADVANCED

  #ifdef UNICODE_CHARS
    // UTF-16LE
    if( buffer[0]==0xFF && buffer[1]==0xFE )
    { // reusing the buffer from load_file
      memcpy( &buffer[0], &buffer[2], size-2 );
      buffer[size-2] = 0;
      buffer[size-1] = 0;
      IDLENGTH(a) = WORD_ID | (((size-2)>>1)<<8);
      STRING(a) = (chars_t)buffer;
      return a;
    }
    // UTF-8
    if( buffer[0]==0xEF && buffer[1]==0xBB && buffer[2]==0xBF )
    { // UTF-8 (header EF BB BF)
      memcpy( &buffer[0], &buffer[3], size-3 );
      buffer[size-3] = 0;
      STRING(a) = UTF8_to_UTF16( buffer );
      IDLENGTH(a) = WORD_ID | (wcslen(STRING(a)))<<8;
      if( dealc ) DEALLOC( buffer );
      return a;
    }
    // ASCII
    buffer[size] = 0;
    STRING(a) = ASCII_to_UTF16( (char*)buffer );
    IDLENGTH(a) = WORD_ID | (STRLEN(STRING(a)))<<8;
    if( dealc ) DEALLOC( buffer );

  #else //UNICODE_CHARS

    // ASCII
    buffer[size] = 0;
    STRING(a) = ASCII_to_ASCII( (char*)buffer );
    IDLENGTH(a) = WORD_ID | (STRLEN(STRING(a)))<<8;
    if( dealc ) DEALLOC( buffer );
  #endif //UNICODE_CHARS

  return a;
}

int atom_to_boolean	(	atom_t	a,
		int *	np
	)

Parameters:

a	atom to convert
np	pointer where to put the result

Returns:: returns 1 if conversion was successful.

Converts an atom to a boolean integer number, i.e. 0 indicates FALSE and 1 (or non-zero) indicates TRUE. If conversion is possible, then *np will contain the result and the return value will be 1. Otherwise *np will be set to 0 and the return value will be set to 0.

{
  *np = 0;

  // check for FALSE atom
  if( a == false_true[0] )
    {
      *np = 0;
      return 1;
    }

  // check for TRUE atom
  if( a == false_true[1] )
    {
      *np = 1;
      return 1;
    }

  // eliminate non-words
  if( !IS_ANY_WORD(a) )
    {
      return 0;
    }

  // check for FALSE word
  if( LENGTH(a)==LENGTH(false_true[0]) && 
      same_strings( 1, STRING(a), STRING(false_true[0]), LENGTH(a)) )
    {
      *np = 0;
      return 1;
    }

  // check for TRUE word
  if( LENGTH(a)==LENGTH(false_true[1]) && 
      same_strings( 1, STRING(a), STRING(false_true[1]), LENGTH(a)) )
    {
      *np = 1;
      return 1;
    }

  return 0;
}

int atom_to_string	(	atom_t	a,
		chars_t	buff,
		int *	buff_len
	)

Parameters:

a	atom to convert
buff	buffer where string will be written
buff_len	length of the buffer

Returns:: returns 1 if conversion was successful.

Converts an atom to a string. If buffer is too small string is truncated. Conversion is successfull if atom is word, subword or number. Else return 0

{
  //  outter(TEXT("a2s atom=|"),-1);dump(a);outter(TEXT("|\n"),-1);

    if (IS_ANY_WORD(a))
    {
        int min = LENGTH(a) < *buff_len ? LENGTH(a) : *buff_len;
        int i;
        for(i = 0; i < min; ++ i)
        {
            buff[i] = STRING(a)[i];
        }
        *buff_len = min;
        return 1;
    }else
    {
        float64_t x;
        if (!atom_to_float (a, &x))
        {
            return 0;
        }
        // In Window infinities are printed as "1.#INF" instead
        // of just "inf", and "-1.#IND" instead of "nan". So let's fix this.
        #ifdef WINDOWS
        if(isinf(x)!=0 && x > 0)
        {
            STRNCPY(buff, TEXT("inf"), *buff_len-1);
            if (*buff_len > 4) *buff_len = 3;
            return 1;
        }
        if( isinf(x)!=0 && x < 0)
        {
            STRNCPY(buff, TEXT("-inf"), *buff_len-1);
            if (*buff_len > 5) *buff_len = 4;
            return 1;
        }
        if( isnan(x)!=0 )
        {
            STRNCPY(buff, TEXT("nan"), *buff_len-1);
            if (*buff_len > 4) *buff_len = 3;
            return 1;
        }
      #endif //WINDOWS

        if( (x>=1e23) || (x<=-1e23) || ((x>=-1e-23) && (x<=1e-23)) )
        {
          SPRINTF( buff, *buff_len, FORMAT_EXTRA_FP, x );
          *buff_len = STRLEN(buff);
        }
        else
        {
          SPRINTF(buff, *buff_len, FORMAT_FP, x);
          *buff_len = STRLEN(buff);
          //outter(TEXT("a2s=|"),-1); outter(buff,*buff_len); outter(TEXT("|\n"),-1);
          //atom_t nn=new_integer(*buff_len);
          //outter(TEXT("len="),-1); dumpln(nn);
          while (buff[*buff_len-1] == '0') --*buff_len;
          if (buff[*buff_len-1] == '.') --*buff_len;
          buff[*buff_len] = '\0';
        }
        return 1;
    }
}

atom_t atom_to_word ( atom_t data )

Parameters:

data	atom to convert

Returns:: returns word, subword or error atom

Converts an atom to a word or a subword atom. If the input is a (sub)word, then it is returned directly. If it is a number, then a new word is created. Otherwise an error atom is returned.

{
  if( IS_ERROR(data) ) return data;
  if( IS_ANY_WORD(data) ) return USE(data);
  if( IS_UNBOUND(data) ) return new_error( ERROR_NOT_A_WORD, data );
  if( IS_INTEGER(data) || IS_FLOAT(data) )
    {
      #define BUFFER_SIZE 128
      char_t ch[BUFFER_SIZE];
      int len = BUFFER_SIZE;
      if( atom_to_string(data,ch,&len) ) return new_word( ch, len );
      #undef BUFFER_SIZE
    }
  return new_error( ERROR_NOT_A_WORD, data );
}

atom_t atom_to_real_word ( atom_t data )

Parameters:

data	atom to convert

Returns:: returns word or error atom

Converts an atom to a word. This guarantees that the string in the word atom is zero-terminated. If the input is a word then it is reused. Otherwise a new word is created. Note that a new word is created even if the input is a subword.

{
  atom_t subword;
  atom_t word = atom_to_word( data );
  if( IS_SUBWORD(word) )
    {
      subword = word;
      word = new_word( STRING(subword), LENGTH(subword) );
      DEUSE( subword );
    }
  return word;
}

int same_strings	(	int	ci,
		chars_t	sa,
		chars_t	sb,
		int	n
	)

Parameters:

ci	1=case insensitive, 0=case sensitive
sa	first string
sb	second string
n	string length

Returns:: 0 if strings are different, non zero - otherwise

Compares two strings case sensitive or insensitive.

{
  if( ci )
  // case sensitive   A!=a
  while( n-- ) 
  {
    if( TOLOWER(DEBAR(*sa)) != TOLOWER(DEBAR(*sb)) ) return 0;
    sa++;
    sb++;
  }
  else
  // case sensitive   A!=a
  while( n-- )
  {
    if( DEBAR(*sa) != DEBAR(*sb) ) return 0;
    sa++;
    sb++;
  }
  return 1;
}

Variable Documentation

int full_print_p

atom_t false_true[2]

Defines

Functions

Variables

Define Documentation

Function Documentation

Variable Documentation