parser.c

操作系统源代码

 *	quoted.
 * If the token is TREDIR, then we set redirnode to a structure containing
 *	the redirection.
 * In all cases, the variable startlinno is set to the number of the line
 *	on which the token starts.
 *
 * [Change comment:  here documents and internal procedures]
 * [Readtoken shouldn't have any arguments.  Perhaps we should make the
 *  word parsing code into a separate routine.  In this case, readtoken
 *  doesn't need to have any internal procedures, but parseword does.
 *  We could also make parseoperator in essence the main routine, and
 *  have parseword (readtoken1?) handle both words and redirection.]
 */

#define RETURN(token)	return lasttoken = token

STATIC int
xxreadtoken() {
	register c;

	if (tokpushback) {
		tokpushback = 0;
		return lasttoken;
	}
	if (needprompt) {
		putprompt(ps2val());
		needprompt = 0;
	}
	startlinno = plinno;
	for (;;) {	/* until token or start of word found */
		c = pgetc_macro();
		if (c == ' ' || c == '\t')
			continue;		/* quick check for white space first */
		switch (c) {
		case ' ': case '\t':
			continue;
		case '#':
			while ((c = pgetc()) != '\n' && c != PEOF);
			pungetc();
			continue;
		case '\\':
			if (pgetc() == '\n') {
				startlinno = ++plinno;
				if (doprompt)
					putprompt(ps2val());
				continue;
			}
			pungetc();
			goto breakloop;
		case '\n':
			plinno++;
			needprompt = doprompt;
			RETURN(TNL);
		case PEOF:
			RETURN(TEOF);
		case '&':
			if (pgetc() == '&')
				RETURN(TAND);
			pungetc();
			RETURN(TBACKGND);
		case '|':
			if (pgetc() == '|')
				RETURN(TOR);
			pungetc();
			RETURN(TPIPE);
		case ';':
			if (pgetc() == ';')
				RETURN(TENDCASE);
			pungetc();
			RETURN(TSEMI);
		case '(':
			RETURN(TLP);
		case ')':
			RETURN(TRP);
		default:
			goto breakloop;
		}
	}
breakloop:
	return readtoken1(c, BASESYNTAX, (char *)NULL, 0);
#undef RETURN
}



/*
 * If eofmark is NULL, read a word or a redirection symbol.  If eofmark
 * is not NULL, read a here document.  In the latter case, eofmark is the
 * word which marks the end of the document and striptabs is true if
 * leading tabs should be stripped from the document.  The argument firstc
 * is the first character of the input token or document.
 *
 * Because C does not have internal subroutines, I have simulated them
 * using goto's to implement the subroutine linkage.  The following macros
 * will run code that appears at the end of readtoken1.
 */

#define CHECKEND()	{goto checkend; checkend_return:;}
#define PARSEREDIR()	{goto parseredir; parseredir_return:;}
#define PARSESUB()	{goto parsesub; parsesub_return:;}
#define PARSEBACKQOLD()	{oldstyle = 1; goto parsebackq; parsebackq_oldreturn:;}
#define PARSEBACKQNEW()	{oldstyle = 0; goto parsebackq; parsebackq_newreturn:;}

STATIC int
readtoken1(firstc, syntax, eofmark, striptabs)
	int firstc;
	char const *syntax;
	char *eofmark;
	int striptabs;
	{
	register c = firstc;
	register char *out;
	int len;
	char line[EOFMARKLEN + 1];
	struct nodelist *bqlist;
	int quotef;
	int dblquote;
	int varnest;
	int oldstyle;

	startlinno = plinno;
	dblquote = 0;
	if (syntax == DQSYNTAX)
		dblquote = 1;
	quotef = 0;
	bqlist = NULL;
	varnest = 0;
	STARTSTACKSTR(out);
	loop: {	/* for each line, until end of word */
#if ATTY
		if (c == '\034' && doprompt
		 && attyset() && ! equal(termval(), "emacs")) {
			attyline();
			if (syntax == BASESYNTAX)
				return readtoken();
			c = pgetc();
			goto loop;
		}
#endif
		CHECKEND();	/* set c to PEOF if at end of here document */
		for (;;) {	/* until end of line or end of word */
			CHECKSTRSPACE(3, out);	/* permit 3 calls to USTPUTC */
			switch(syntax[c]) {
			case CNL:	/* '\n' */
				if (syntax == BASESYNTAX)
					goto endword;	/* exit outer loop */
				USTPUTC(c, out);
				plinno++;
				if (doprompt) {
					putprompt(ps2val());
				}
				c = pgetc();
				goto loop;		/* continue outer loop */
			case CWORD:
				USTPUTC(c, out);
				break;
			case CCTL:
				if (eofmark == NULL || dblquote)
					USTPUTC(CTLESC, out);
				USTPUTC(c, out);
				break;
			case CBACK:	/* backslash */
				c = pgetc();
				if (c == PEOF) {
					USTPUTC('\\', out);
					pungetc();
				} else if (c == '\n') {
					if (doprompt)
						putprompt(ps2val());
				} else {
					if (dblquote && c != '\\' && c != '`' && c != '$'
							 && (c != '"' || eofmark != NULL))
						USTPUTC('\\', out);
					if (SQSYNTAX[c] == CCTL)
						USTPUTC(CTLESC, out);
					USTPUTC(c, out);
					quotef++;
				}
				break;
			case CSQUOTE:
				syntax = SQSYNTAX;
				break;
			case CDQUOTE:
				syntax = DQSYNTAX;
				dblquote = 1;
				break;
			case CENDQUOTE:
				if (eofmark) {
					USTPUTC(c, out);
				} else {
					syntax = BASESYNTAX;
					quotef++;
					dblquote = 0;
				}
				break;
			case CVAR:	/* '$' */
				PARSESUB();		/* parse substitution */
				break;
			case CENDVAR:	/* '}' */
				if (varnest > 0) {
					varnest--;
					USTPUTC(CTLENDVAR, out);
				} else {
					USTPUTC(c, out);
				}
				break;
			case CBQUOTE:	/* '`' */
				PARSEBACKQOLD();
				break;
			case CEOF:
				goto endword;		/* exit outer loop */
			default:
				if (varnest == 0)
					goto endword;	/* exit outer loop */
				USTPUTC(c, out);
			}
			c = pgetc_macro();
		}
	}
endword:
	if (syntax != BASESYNTAX && ! parsebackquote && eofmark == NULL)
		synerror("Unterminated quoted string");
	if (varnest != 0) {
		startlinno = plinno;
		synerror("Missing '}'");
	}
	USTPUTC('\0', out);
	len = out - stackblock();
	out = stackblock();
	if (eofmark == NULL) {
		if ((c == '>' || c == '<')
		 && quotef == 0
		 && len <= 2
		 && (*out == '\0' || is_digit(*out))) {
			PARSEREDIR();
			return lasttoken = TREDIR;
		} else {
			pungetc();
		}
	}
	quoteflag = quotef;
	backquotelist = bqlist;
	grabstackblock(len);
	wordtext = out;
	return lasttoken = TWORD;
/* end of readtoken routine */



/*
 * Check to see whether we are at the end of the here document.  When this
 * is called, c is set to the first character of the next input line.  If
 * we are at the end of the here document, this routine sets the c to PEOF.
 */

checkend: {
	if (eofmark) {
		if (striptabs) {
			while (c == '\t')
				c = pgetc();
		}
		if (c == *eofmark) {
			if (pfgets(line, sizeof line) != NULL) {
				register char *p, *q;

				p = line;
				for (q = eofmark + 1 ; *q && *p == *q ; p++, q++);
				if (*p == '\n' && *q == '\0') {
					c = PEOF;
					plinno++;
					needprompt = doprompt;
				} else {
					ppushback(line, strlen(line));
				}
			}
		}
	}
	goto checkend_return;
}


/*
 * Parse a redirection operator.  The variable "out" points to a string
 * specifying the fd to be redirected.  The variable "c" contains the
 * first character of the redirection operator.
 */

parseredir: {
	char fd = *out;
	union node *np;

	np = (union node *)stalloc(sizeof (struct nfile));
	if (c == '>') {
		np->nfile.fd = 1;
		c = pgetc();
		if (c == '>')
			np->type = NAPPEND;
		else if (c == '&')
			np->type = NTOFD;
		else {
			np->type = NTO;
			pungetc();
		}
	} else {	/* c == '<' */
		np->nfile.fd = 0;
		c = pgetc();
		if (c == '<') {
			if (sizeof (struct nfile) != sizeof (struct nhere)) {
				np = (union node *)stalloc(sizeof (struct nhere));
				np->nfile.fd = 0;
			}
			np->type = NHERE;
			heredoc = (struct heredoc *)stalloc(sizeof (struct heredoc));
			heredoc->here = np;
			if ((c = pgetc()) == '-') {
				heredoc->striptabs = 1;
			} else {
				heredoc->striptabs = 0;
				pungetc();
			}
		} else if (c == '&')
			np->type = NFROMFD;
		else {
			np->type = NFROM;
			pungetc();
		}
	}
	if (fd != '\0')
		np->nfile.fd = digit_val(fd);
	redirnode = np;
	goto parseredir_return;
}


/*
 * Parse a substitution.  At this point, we have read the dollar sign
 * and nothing else.
 */

parsesub: {
	int subtype;
	int typeloc;
	int flags;
	char *p;
#ifndef GDB_HACK
	static const char types[] = "}-+?=";
#endif

	c = pgetc();
	if (c != '(' && c != '{' && !is_name(c) && !is_special(c)) {
		USTPUTC('$', out);
		pungetc();
	} else if (c == '(') {	/* $(command) */
		PARSEBACKQNEW();
	} else {
		USTPUTC(CTLVAR, out);
		typeloc = out - stackblock();
		USTPUTC(VSNORMAL, out);
		subtype = VSNORMAL;
		if (c == '{') {
			c = pgetc();
			subtype = 0;
		}
		if (is_name(c)) {
			do {
				STPUTC(c, out);
				c = pgetc();
			} while (is_in_name(c));
		} else {
			if (! is_special(c))
badsub:				synerror("Bad substitution");
			USTPUTC(c, out);
			c = pgetc();
		}
		STPUTC('=', out);
		flags = 0;
		if (subtype == 0) {
			if (c == ':') {
				flags = VSNUL;
				c = pgetc();
			}
			p = strchr(types, c);
			if (p == NULL)
				goto badsub;
			subtype = p - types + VSNORMAL;
		} else {
			pungetc();
		}
		if (dblquote)
			flags |= VSQUOTE;
		*(stackblock() + typeloc) = subtype | flags;
		if (subtype != VSNORMAL)
			varnest++;
	}
	goto parsesub_return;
}


/*
 * Called to parse command substitutions.  Newstyle is set if the command
 * is enclosed inside $(...); nlpp is a pointer to the head of the linked
 * list of commands (passed by reference), and savelen is the number of
 * characters on the top of the stack which must be preserved.
 */

parsebackq: {
	struct nodelist **nlpp;
	int savepbq;
	union node *n;
	char *volatile str;
	struct jmploc jmploc;
	struct jmploc *volatile savehandler;
	int savelen;
	int savedoprompt;

	savepbq = parsebackquote;
	if (setjmp(jmploc.loc)) {
		if (str)
			ckfree(str);
		parsebackquote = 0;
		handler = savehandler;
		longjmp(handler->loc, 1);
	}
	INTOFF;
	str = NULL;
	savelen = out - stackblock();
	if (savelen > 0) {
		str = ckmalloc(savelen);
		bcopy(stackblock(), str, savelen);
	}
	savehandler = handler;
	handler = &jmploc;
	INTON;
	if (oldstyle) {
		/* We must read until the closing backquote, giving special
		   treatment to some slashes, and then push the string and
		   reread it as input, interpreting it normally.  */
		register char *out;
		register c;
		int savelen;
		char *str;

		STARTSTACKSTR(out);
		while ((c = pgetc ()) != '`') {
			if (c == '\\') {
				c = pgetc ();
				if (c != '\\' && c != '`' && c != '$'
				    && (!dblquote || c != '"'))
					STPUTC('\\', out);
			}
			if (c == '\n') {
				plinno++;
				if (doprompt)
					putprompt(ps2val());
			}
			STPUTC(c, out);
		}
		STPUTC('\0', out);
		savelen = out - stackblock();
		if (savelen > 0) {
			str = ckmalloc(savelen);
			bcopy(stackblock(), str, savelen);
		}
		setinputstring(str, 1);
		savedoprompt = doprompt;
		doprompt = 0;	/* no prompts while rereading string */
	}
	nlpp = &bqlist;
	while (*nlpp)
		nlpp = &(*nlpp)->next;
	*nlpp = (struct nodelist *)stalloc(sizeof (struct nodelist));
	(*nlpp)->next = NULL;
	parsebackquote = oldstyle;
	n = list(0);
	if (!oldstyle && (readtoken() != TRP))
		synexpect(TRP);
	(*nlpp)->n = n;
	/* Start reading from old file again.  */
	if (oldstyle) {
		popfile();
		doprompt = savedoprompt;
	}
	while (stackblocksize() <= savelen)
		growstackblock();
	STARTSTACKSTR(out);
	if (str) {
		bcopy(str, out, savelen);
		STADJUST(savelen, out);
		INTOFF;
		ckfree(str);
		str = NULL;
		INTON;
	}
	parsebackquote = savepbq;
	handler = savehandler;
	USTPUTC(CTLBACKQ + dblquote, out);
	if (oldstyle)
		goto parsebackq_oldreturn;
	else
		goto parsebackq_newreturn;
}

} /* end of readtoken */



#ifdef mkinit
RESET {
	tokpushback = 0;
}
#endif


#if READLINE
/*
 * Remember a prompt for use with readline if input and output is a terminal.
 */

STATIC void
putprompt(s)
	char *s;
	{
	if (editable) {
		r_use_prompt = s;
	} else {
		out2str(s);
	}
}
#endif

#if ATTY
/*
 * Called to process a command generated by atty.  We execute the line,
 * and catch any errors that occur so they don't propagate outside of
 * this routine.
 */

STATIC void
attyline() {
	char line[256];
	struct stackmark smark;
	struct jmploc jmploc;
	struct jmploc *volatile savehandler;

	if (pfgets(line, sizeof line) == NULL)
		return;				/* "can't happen" */
	if (setjmp(jmploc.loc)) {
		if (exception == EXERROR)
			out2str("\033]D\n");
		handler = savehandler;
		longjmp(handler->loc, 1);
	}
	savehandler = handler;
	handler = &jmploc;
	setstackmark(&smark);
	evalstring(line);
	popstackmark(&smark);
	handler = savehandler;
	doprompt = 1;
}


/*
 * Output a prompt for atty.  We output the prompt as part of the
 * appropriate escape sequence.  
 */

STATIC void
putprompt(s)
	char *s;
	{
	register char *p;

	if (attyset() && ! equal(termval(), "emacs")) {
		if (strchr(s, '\7'))
			out2c('\7');
		out2str("\033]P1;");
		for (p = s ; *p ; p++) {
			if ((unsigned)(*p - ' ') <= '~' - ' ')
				out2c(*p);
		}
		out2c('\n');
	} else {
		out2str(s);
	}
}
#endif



/*
 * Returns true if the text contains nothing to expand (no dollar signs
 * or backquotes).
 */

STATIC int
noexpand(text)
	char *text;
	{
	register char *p;
	register char c;

	p = text;
	while ((c = *p++) != '\0') {
		if (c == CTLESC)
			p++;
		else if (BASESYNTAX[c] == CCTL)
			return 0;
	}
	return 1;
}


/*
 * Return true if the argument is a legal variable name (a letter or
 * underscore followed by zero or more letters, underscores, and digits).
 */

int
goodname(name)
	char *name;
	{
	register char *p;

	p = name;
	if (! is_name(*p))
		return 0;
	while (*++p) {
		if (! is_in_name(*p))
			return 0;
	}
	return 1;
}


/*
 * Called when an unexpected token is read during the parse.  The argument
 * is the token that is expected, or -1 if more than one type of token can
 * occur at this point.
 */

STATIC void
synexpect(token) {
	char msg[64];

	if (token >= 0) {
		fmtstr(msg, 64, "%s unexpected (expecting %s)",
			tokname[lasttoken], tokname[token]);
	} else {
		fmtstr(msg, 64, "%s unexpected", tokname[lasttoken]);
	}
	synerror(msg);
}


STATIC void
synerror(msg)
	char *msg;
	{
	if (commandname)
		outfmt(&errout, "%s: %d: ", commandname, startlinno);
	outfmt(&errout, "Syntax error: %s\n", msg);
	error((char *)NULL);
}