/* -*- LPC -*- */
/*
* $Locker: $
* $Id: tokeniser.c,v 1.3 2000/05/11 23:54:51 pinkfish Exp $
* $Log: tokeniser.c,v $
* Revision 1.3 2000/05/11 23:54:51 pinkfish
* Make sure it detects floats correctly.
*
* Revision 1.2 1998/04/03 19:46:40 pinkfish
* Add floating point support.
*
* Revision 1.1 1998/01/06 05:03:33 ceres
* Initial revision
*
*/
/**
* Does tokenising the the parser used by the mudlib when someone uses
* the ';' command.
* @author Pinkfish
*/
/* Ok, simple (hopefully) recursive descent parser. */
#include "tokenise.h"
mapping token;
void create() {
token = ([
'+' : TOK_PLUS,
'-' : TOK_MINUS,
'*' : TOK_MULT,
'/' : TOK_DIV,
'(' : TOK_OBRAC,
')' : TOK_CBRAC,
'{' : TOK_SARRAY,
'}' : TOK_EARRAY,
'[' : TOK_SMAPPING,
']' : TOK_EMAPPING,
':' : TOK_COLON,
',' : TOK_COMMA,
'.' : TOK_DOT,
'\t' : TOK_TAB,
'=' : TOK_ASSIGN,
#ifdef BOOLEAN
'>' : TOK_GREATOR,
'<' : TOK_LESS,
"==" : TOK_EQUAL,
">=" : TOK_GREATOREQ,
"<=" : TOK_LESSEQUAL,
#endif
"->" : TOK_CALL,
".." : TOK_DOTDOT,
]);
} /* create() */
/**
* Tokenise, we rip out strings and make them as seperate enties
* Otherwise things inside strings will get processed elsewhere
* which could be bad...
* @param inp the input string
* @return the tokenised string
*/
mixed tokenise(string inp) {
string *bits, *cur;
int pos, num;
int start_pos;
cur = ({ });
while (pos < strlen(inp))
switch (inp[pos++]) {
case ' ' :
case '\t' :
break;
case '\'' :
case '"' :
case '`' :
bits = explode("#"+inp[pos-1..]+"#", inp[pos-1..pos-1]);
if (!bits || sizeof(bits) < 2) {
write("Error processing the string.\n");
return 0;
}
pos += strlen(bits[1])+1;
cur += ({ ({ bits[1] }) });
break;
case '.' :
case '0' :
case '1' :
case '2' :
case '3' :
case '4' :
case '5' :
case '6' :
case '7' :
case '8' :
case '9' :
start_pos = pos;
while ((inp[pos] >= '0' && inp[pos] <= '9') || (inp[pos] == '.')) {
pos++;
}
if (strsrch(inp[start_pos..pos], ".") != -1) {
sscanf(inp[start_pos-1..pos], "%f", num);
} else {
sscanf(inp[start_pos-1..pos], "%d", num);
}
cur += ({ ({ num }) });
break;
case '+' :
case '*' :
case ':' :
case ',' :
case '{' :
case '}' :
case '(' :
case ')' :
case '[' :
case ']' :
cur += ({ token[inp[pos-1]] });
break;
case '>' :
case '<' :
case '=' :
case '-' :
case '/' :
case '.' :
if (inp[pos-1] != '/') {
if ((num = token[inp[pos-1..pos]])) {
cur += ({ num });
pos++;
break;
}
num = token[inp[pos-1]];
if (num) {
cur += ({ num });
break;
}
} else {
if (sizeof(cur) && !stringp(cur[sizeof(cur)-1])) {
cur += ({ token[inp[pos-1]] });
break;
}
}
default :
num = pos-1;
while (pos < strlen(inp) && (!token[inp[pos++]] || inp[pos-1] == '/'));
if (pos == num+1)
break;
if (token[inp[pos-1]])
pos--;
cur += ({ inp[num..pos-1] });
break;
}
return cur;
} /* tokenise() */
