token.c

Calc Software Package for Number Calc

/*
 * token - read input file characters into tokens
 *
 * Copyright (C) 1999  David I. Bell and Ernest Bowen
 *
 * Primary author:  David I. Bell
 *
 * Calc is open software; you can redistribute it and/or modify it under
 * the terms of the version 2.1 of the GNU Lesser General Public License
 * as published by the Free Software Foundation.
 *
 * Calc is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
 * or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU Lesser General
 * Public License for more details.
 *
 * A copy of version 2.1 of the GNU Lesser General Public License is
 * distributed with calc under the filename COPYING-LGPL.  You should have
 * received a copy with calc; if not, write to Free Software Foundation, Inc.
 * 59 Temple Place, Suite 330, Boston, MA  02111-1307, USA.
 *
 * @(#) $Revision: 29.10 $
 * @(#) $Id: token.c,v 29.10 2006/06/24 18:43:05 chongo Exp $
 * @(#) $Source: /usr/local/src/cmd/calc/RCS/token.c,v $
 *
 * Under source code control:	1990/02/15 01:48:25
 * File existed as early as:	before 1990
 *
 * Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/
 */


#include <stdio.h>
#include <setjmp.h>

#include "calc.h"
#include "token.h"
#include "string.h"
#include "args.h"
#include "math_error.h"


#define isletter(ch)	((((ch) >= 'a') && ((ch) <= 'z')) || \
				(((ch) >= 'A') && ((ch) <= 'Z')))
#define isdigit(ch)	(((ch) >= '0') && ((ch) <= '9'))
#define issymbol(ch)	(isletter(ch) || isdigit(ch) || ((ch) == '_'))
#define isoctal(ch)	(((ch) >= '0') && ((ch) <= '7'))

#define STRBUFSIZE	1024


/*
 * Current token.
 */
static struct {
	short t_type;		/* type of token */
	char *t_sym;		/* symbol name */
	long t_strindex;	/* index of string value */
	long t_numindex;	/* index of numeric value */
} curtoken;


static BOOL rescan;		/* TRUE to reread current token */
static BOOL newlines;		/* TRUE to return newlines as tokens */
static BOOL allsyms;		/* TRUE if always want a symbol token */
static STRINGHEAD strings;	/* list of constant strings */
static char *numbuf;		/* buffer for numeric tokens */
static long numbufsize;		/* current size of numeric buffer */

long errorcount = 0;			/* number of compilation errors */


/*
 * Table of keywords
 */
struct keyword {
	char *k_name;	/* keyword name */
	int k_token;	/* token number */
};

static struct keyword keywords[] = {
	{"if",		T_IF},
	{"else",	T_ELSE},
	{"for",		T_FOR},
	{"while",	T_WHILE},
	{"do",		T_DO},
	{"continue",	T_CONTINUE},
	{"break",	T_BREAK},
	{"goto",	T_GOTO},
	{"return",	T_RETURN},
	{"local",	T_LOCAL},
	{"global",	T_GLOBAL},
	{"static",	T_STATIC},
	{"switch",	T_SWITCH},
	{"case",	T_CASE},
	{"default",	T_DEFAULT},
	{"quit",	T_QUIT},
	{"exit",	T_QUIT},
	{"define",	T_DEFINE},
	{"read",	T_READ},
	{"show",	T_SHOW},
	{"help",	T_HELP},
	{"write",	T_WRITE},
	{"mat",		T_MAT},
	{"obj",		T_OBJ},
	{"print",	T_PRINT},
	{"cd",		T_CD},
	{"undefine",	T_UNDEFINE},
	{"abort",	T_ABORT},
	{NULL,		0}
};


static void eatcomment(void);
static void eatstring(int quotechar);
static void eatline(void);
static int eatsymbol(void);
static int eatnumber(void);


/*
 * Initialize all token information.
 */
void
inittokens(void)
{
	initstr(&strings);
	newlines = FALSE;
	allsyms = FALSE;
	rescan = FALSE;
	setprompt(conf->prompt1);
}


/*
 * Set the new token mode according to the specified flag, and return the
 * previous value of the flag.
 */
int
tokenmode(int flag)
{
	int	oldflag;

	oldflag = TM_DEFAULT;
	if (newlines)
		oldflag |= TM_NEWLINES;
	if (allsyms)
		oldflag |= TM_ALLSYMS;
	newlines = FALSE;
	allsyms = FALSE;
	if (flag & TM_NEWLINES)
		newlines = TRUE;
	if (flag & TM_ALLSYMS)
		allsyms = TRUE;
	setprompt(newlines ? conf->prompt1 : conf->prompt2);
	return oldflag;
}


/*
 * Routine to read in the next token from the input stream.
 * The type of token is returned as a value.  If the token is a string or
 * symbol name, information is saved so that the value can be retrieved.
 */
int
gettoken(void)
{
	int ch;			/* current input character */
	int type;		/* token type */

	if (rescan) {		/* rescanning */
		rescan = FALSE;
		return curtoken.t_type;
	}
	curtoken.t_sym = NULL;
	curtoken.t_strindex = 0;
	curtoken.t_numindex = 0;
	type = T_NULL;
	while (type == T_NULL) {
		ch = nextchar();
		if (allsyms && ch!=' ' && ch!=';' && ch!='"' &&
				ch!='\'' && ch!='\n' && ch!=EOF) {
			reread();
			type = eatsymbol();
			break;
		}
		switch (ch) {
		case ' ':
		case '\t':
		case '\r':
		case '\v':
		case '\f':
		case '\0':
			break;
		case '\n':
			if (newlines)
				type = T_NEWLINE;
			break;
		case EOF: type = T_EOF; break;
		case '{': type = T_LEFTBRACE; break;
		case '}': type = T_RIGHTBRACE; break;
		case '(': type = T_LEFTPAREN; break;
		case ')': type = T_RIGHTPAREN; break;
		case '[': type = T_LEFTBRACKET; break;
		case ']': type = T_RIGHTBRACKET; break;
		case ';': type = T_SEMICOLON; break;
		case ':': type = T_COLON; break;
		case ',': type = T_COMMA; break;
		case '?': type = T_QUESTIONMARK; break;
		case '@': type = T_AT; break;
		case '`': type = T_BACKQUOTE; break;
		case '$': type = T_DOLLAR; break;
		case '"':
		case '\'':
			type = T_STRING;
			eatstring(ch);
			break;
		case '^':
			switch (nextchar()) {
			case '=': type = T_POWEREQUALS; break;
			default: type = T_POWER; reread();
			}
			break;
		case '=':
			switch (nextchar()) {
			case '=': type = T_EQ; break;
			default: type = T_ASSIGN; reread();
			}
			break;
		case '+':
			switch (nextchar()) {
			case '+': type = T_PLUSPLUS; break;
			case '=': type = T_PLUSEQUALS; break;
			default: type = T_PLUS; reread();
			}
			break;
		case '-':
			switch (nextchar()) {
			case '-': type = T_MINUSMINUS; break;
			case '=': type = T_MINUSEQUALS; break;
			case '>': type = T_ARROW; break;
			default: type = T_MINUS; reread();
			}
			break;
		case '*':
			switch (nextchar()) {
			case '=': type = T_MULTEQUALS; break;
			case '*':
				switch (nextchar()) {
				case '=':
					type = T_POWEREQUALS; break;
				default:
					type = T_POWER; reread();
				}
				break;
			default: type = T_MULT; reread();
			}
			break;
		case '/':
			switch (nextchar()) {
			case '/':
				switch (nextchar()) {
				case '=':
					type = T_SLASHSLASHEQUALS;
					break;
				default:
					reread();
					type = T_SLASHSLASH;
					break;
				}
				break;
			case '=': type = T_DIVEQUALS; break;
			case '*': eatcomment(); break;
			default: type = T_DIV; reread();
			}
			break;
		case '%':
			switch (nextchar()) {
			case '=': type = T_MODEQUALS; break;
			default: type = T_MOD; reread();
			}
			break;
		case '<':
			switch (nextchar()) {
			case '=': type = T_LE; break;
			case '<':
				switch (nextchar()) {
				case '=':
					type = T_LSHIFTEQUALS;
					break;
				default:
					reread();
					type = T_LEFTSHIFT;
					break;
				}
				break;
			default: type = T_LT; reread();
			}
			break;
		case '>':
			switch (nextchar()) {
			case '=': type = T_GE; break;
			case '>':
				switch (nextchar()) {
				case '=':
					type = T_RSHIFTEQUALS;
					break;
				default:
					reread();
					type = T_RIGHTSHIFT;
					break;
				}
				break;
			default: type = T_GT; reread();
			}
			break;
		case '&':
			switch (nextchar()) {
			case '&': type = T_ANDAND; break;
			case '=': type = T_ANDEQUALS; break;
			default: type = T_AND; reread(); break;
			}
			break;
		case '|':
			switch (nextchar()) {
			case '|': type = T_OROR; break;
			case '=': type = T_OREQUALS; break;
			default: type = T_OR; reread(); break;
			}
			break;
		case '!':
			switch (nextchar()) {
			case '=': type = T_NE; break;
			default: type = T_NOT; reread(); break;
			}
			break;
		case '#':
			switch(nextchar()) {
			case '=': type = T_HASHEQUALS; break;
			case '!':
			case '#': eatline(); break;
			case '\n': reread(); break;
			default: type = T_HASH; reread();
			}
			break;
		case '~':
			switch (nextchar()) {
			case '=': type = T_TILDEEQUALS; break;
			default: type = T_TILDE; reread();
			}
			break;
		case '\\':
			switch (nextchar()) {
			case '\n': setprompt(conf->prompt2); break;
			case '=': type = T_BACKSLASHEQUALS; break;
			default: type = T_BACKSLASH; reread();
			}
			break;
		default:
			if (isletter(ch) || ch == '_') {
				reread();
				type = eatsymbol();
				break;
			}
			if (isdigit(ch) || (ch == '.')) {
				reread();
				type = eatnumber();
				break;
			}
			scanerror(T_NULL, "Unknown token character '%c'", ch);
		}
	}
	curtoken.t_type = (short)type;
	return type;
}


/*
 * Continue to eat up a comment string.
 * The leading slash-asterisk has just been scanned at this point.
 */
static void
eatcomment(void)
{
	int ch;
	setprompt(conf->prompt2);
	for (;;) {
		ch = nextchar();
		if (ch == '*') {
			ch = nextchar();
			if (ch == '/')
				break;
			reread();
		}
		if (ch == EOF || ch == '\0') {
			scanerror(T_NULL, "Unterminated comment");
			reread();
			break;