ejslex.c

来自「samba最新软件」· C语言 代码 · 共 924 行 · 第 1/2 页

/*
 *	@file 	ejsLex.c
 *	@brief 	EJS Lexical Analyser
 *	@overview EJS lexical analyser. This implementes a lexical analyser 
 *		for a subset of the JavaScript language.
 */
/********************************* Copyright **********************************/
/*
 *	@copy	default.g
 *	
 *	Copyright (c) Mbedthis Software LLC, 2003-2005. All Rights Reserved.
 *	Portions Copyright (c) GoAhead Software, 1995-2000. All Rights Reserved.
 *	
 *	This software is distributed under commercial and open source licenses.
 *	You may use the GPL open source license described below or you may acquire 
 *	a commercial license from Mbedthis Software. You agree to be fully bound 
 *	by the terms of either license. Consult the LICENSE.TXT distributed with 
 *	this software for full details.
 *	
 *	This software is open source; you can redistribute it and/or modify it 
 *	under the terms of the GNU General Public License as published by the 
 *	Free Software Foundation; either version 2 of the License, or (at your 
 *	option) any later version. See the GNU General Public License for more 
 *	details at: http://www.mbedthis.com/downloads/gplLicense.html
 *	
 *	This program is distributed WITHOUT ANY WARRANTY; without even the 
 *	implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. 
 *	
 *	This GPL license does NOT permit incorporating this software into 
 *	proprietary programs. If you are unable to comply with the GPL, you must
 *	acquire a commercial license to use this software. Commercial licenses 
 *	for this software and support services are available from Mbedthis 
 *	Software at http://www.mbedthis.com 
 *	
 *	@end
 */
/********************************** Includes **********************************/

#include	"ejsInternal.h"

#if BLD_FEATURE_EJS

/****************************** Forward Declarations **************************/

static int 		getLexicalToken(Ejs *ep, int state);
static int 		tokenAddChar(Ejs *ep, int c);
static int 		inputGetc(Ejs *ep);
static void		inputPutback(Ejs *ep, int c);
static int		charConvert(Ejs *ep, int base, int maxDig);

/************************************* Code ***********************************/
/*
 *	Open a new input script
 */

int ejsLexOpenScript(Ejs *ep, char *script)
{
	EjsInput	*ip;

	mprAssert(ep);
	mprAssert(script);

	if ((ip = mprMalloc(sizeof(EjsInput))) == NULL) {
		return -1;
	}
	memset(ip, 0, sizeof(*ip));
	ip->next = ep->input;
	ep->input = ip;
	ip->procName = ep->proc?ep->proc->procName:NULL;

/*
 *	Create the parse token buffer and script buffer
 */
	ip->tokbuf = mprMalloc(EJS_PARSE_INCR);
	ip->tokSize = EJS_PARSE_INCR; 
	ip->tokServp = ip->tokbuf;
	ip->tokEndp = ip->tokbuf;

	ip->script = mprStrdup(script);
	ip->scriptSize = strlen(script);
	ip->scriptServp = ip->script;

	ip->lineNumber = 1;
	ip->lineLength = 0;
	ip->lineColumn = 0;
	ip->line = NULL;

	ip->putBackIndex = -1;

	return 0;
}

/******************************************************************************/
/*
 *	Close the input script
 */

void ejsLexCloseScript(Ejs *ep)
{
	EjsInput	*ip;
	int			i;

	mprAssert(ep);

	ip = ep->input;
	mprAssert(ip);
	ep->input = ip->next;

	for (i = 0; i < EJS_TOKEN_STACK; i++) {
		mprFree(ip->putBack[i].token);
		ip->putBack[i].token = 0;
	}

	mprFree(ip->line);
	mprFree(ip->tokbuf);
	mprFree(ip->script);

	mprFree(ip);
}

/******************************************************************************/
/*
 *	Initialize an input state structure
 */

int ejsInitInputState(EjsInput *ip)
{
	mprAssert(ip);

	memset(ip, 0, sizeof(*ip));
	ip->putBackIndex = -1;

	return 0;
}
/******************************************************************************/
/*
 *	Save the input state
 */

void ejsLexSaveInputState(Ejs *ep, EjsInput *state)
{
	EjsInput	*ip;
	int			i;

	mprAssert(ep);

	ip = ep->input;
	mprAssert(ip);

	*state = *ip;

	for (i = 0; i < ip->putBackIndex; i++) {
		state->putBack[i].token = mprStrdup(ip->putBack[i].token);
		state->putBack[i].id = ip->putBack[i].id;
	}
	for (; i < EJS_TOKEN_STACK; i++) {
		state->putBack[i].token = 0;
	}

	state->line = mprMalloc(ip->lineLength);
	mprStrcpy(state->line, ip->lineLength, ip->line);

	state->lineColumn = ip->lineColumn;
	state->lineNumber = ip->lineNumber;
	state->lineLength = ip->lineLength;
}

/******************************************************************************/
/*
 *	Restore the input state
 */

void ejsLexRestoreInputState(Ejs *ep, EjsInput *state)
{
	EjsInput	*ip;
	int			i;

	mprAssert(ep);
	mprAssert(state);

	ip = ep->input;
	mprAssert(ip);

	ip->tokbuf = state->tokbuf;
	ip->tokServp = state->tokServp;
	ip->tokEndp = state->tokEndp;
	ip->tokSize = state->tokSize;

	ip->script = state->script;
	ip->scriptServp = state->scriptServp;
	ip->scriptSize = state->scriptSize;

	ip->putBackIndex = state->putBackIndex;
	for (i = 0; i < ip->putBackIndex; i++) {
		mprFree(ip->putBack[i].token);
		ip->putBack[i].id = state->putBack[i].id;
		ip->putBack[i].token = mprStrdup(state->putBack[i].token);
	}

	mprFree(ip->line);
	ip->line = mprMalloc(state->lineLength);
	mprStrcpy(ip->line, state->lineLength, state->line);

	ip->lineColumn = state->lineColumn;
	ip->lineNumber = state->lineNumber;
	ip->lineLength = state->lineLength;
}

/******************************************************************************/
/*
 *	Free a saved input state
 */

void ejsLexFreeInputState(Ejs *ep, EjsInput *state)
{
	int			i;

	mprAssert(ep);
	mprAssert(state);

	for (i = 0; i < EJS_TOKEN_STACK; i++) {
		mprFree(state->putBack[i].token);
	}
	state->putBackIndex = -1;
	mprFree(state->line);
	state->lineLength = 0;
	state->lineColumn = 0;
}

/******************************************************************************/
/*
 *	Get the next EJS token
 */

int ejsLexGetToken(Ejs *ep, int state)
{
	mprAssert(ep);

	ep->tid = getLexicalToken(ep, state);
	return ep->tid;
}

/******************************************************************************/

/*
 *	Check for reserved words "if", "else", "var", "for", "foreach",
 *	"delete", "function", and "return". "new", "in" and "function" 
 *	done below. "true", "false", "null", "undefined" are handled
 *	as global objects.
 *
 *	Other reserved words not supported:
 *		"break", "case", "catch", "continue", "default", "do", 
 *		"finally", "instanceof", "switch", "this", "throw", "try",
 *		"typeof", "while", "with"
 *
 *	ECMA extensions reserved words (not supported):
 *		"abstract", "boolean", "byte", "char", "class", "const",
 *		"debugger", "double", "enum", "export", "extends",
 *		"final", "float", "goto", "implements", "import", "int",
 *		"interface", "long", "native", "package", "private",
 *		"protected", "public", "short", "static", "super",
 *		"synchronized", "throws", "transient", "volatile"
 */

static int checkReservedWord(Ejs *ep, int state, int c, int tid)
{
	if (state == EJS_STATE_STMT) {
		if (strcmp(ep->token, "if") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_IF;
		} else if (strcmp(ep->token, "else") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_ELSE;
		} else if (strcmp(ep->token, "var") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_VAR;
		} else if (strcmp(ep->token, "for") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_FOR;
		} else if (strcmp(ep->token, "delete") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_DELETE;
		} else if (strcmp(ep->token, "function") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_FUNCTION;
		} else if (strcmp(ep->token, "return") == 0) {
			if ((c == ';') || (c == '(')) {
				inputPutback(ep, c);
			}
			return EJS_TOK_RETURN;
		}
	} else if (state == EJS_STATE_EXPR) {
		if (strcmp(ep->token, "new") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_NEW;
		} else if (strcmp(ep->token, "in") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_IN;
		} else if (strcmp(ep->token, "function") == 0) {
			inputPutback(ep, c);
			return EJS_TOK_FUNCTION;
		}
	}
	return tid;
}

/******************************************************************************/
/*
 *	Get the next EJS token
 */

static int getLexicalToken(Ejs *ep, int state)
{
	MprType		type;
	EjsInput	*ip;
	int			done, tid, c, quote, style, idx;

	mprAssert(ep);
	ip = ep->input;
	mprAssert(ip);

	ep->tid = -1;
	tid = -1;
	type = BLD_FEATURE_NUM_TYPE_ID;

	/*
 	 *	Use a putback tokens first. Don't free strings as caller needs access.
	 */
	if (ip->putBackIndex >= 0) {
		idx = ip->putBackIndex;
		tid = ip->putBack[idx].id;
		ep->token = (char*) ip->putBack[idx].token;
		tid = checkReservedWord(ep, state, 0, tid);
		ip->putBackIndex--;
		return tid;
	}
	ep->token = ip->tokServp = ip->tokEndp = ip->tokbuf;
	*ip->tokServp = '\0';

	if ((c = inputGetc(ep)) < 0) {
		return EJS_TOK_EOF;
	}

	/*
 	 *	Main lexical analyser
	 */
	for (done = 0; !done; ) {
		switch (c) {
		case -1:
			return EJS_TOK_EOF;

		case ' ':
		case '\t':
		case '\r':
			do {
				if ((c = inputGetc(ep)) < 0)
					break;
			} while (c == ' ' || c == '\t' || c == '\r');
			break;

		case '\n':
			return EJS_TOK_NEWLINE;

		case '(':
			tokenAddChar(ep, c);
			return EJS_TOK_LPAREN;

		case ')':
			tokenAddChar(ep, c);
			return EJS_TOK_RPAREN;

		case '[':
			tokenAddChar(ep, c);
			return EJS_TOK_LBRACKET;

		case ']':
			tokenAddChar(ep, c);
			return EJS_TOK_RBRACKET;

		case '.':
			tokenAddChar(ep, c);
			return EJS_TOK_PERIOD;

		case '{':
			tokenAddChar(ep, c);
			return EJS_TOK_LBRACE;

		case '}':
			tokenAddChar(ep, c);
			return EJS_TOK_RBRACE;

		case '+':
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			if (c != '+' ) {
				inputPutback(ep, c);
				tokenAddChar(ep, EJS_EXPR_PLUS);
				return EJS_TOK_EXPR;
			}
			tokenAddChar(ep, EJS_EXPR_INC);
			return EJS_TOK_INC_DEC;

		case '-':
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			if (c != '-' ) {
				inputPutback(ep, c);
				tokenAddChar(ep, EJS_EXPR_MINUS);
				return EJS_TOK_EXPR;
			}
			tokenAddChar(ep, EJS_EXPR_DEC);
			return EJS_TOK_INC_DEC;

		case '*':
			tokenAddChar(ep, EJS_EXPR_MUL);
			return EJS_TOK_EXPR;

		case '%':
			tokenAddChar(ep, EJS_EXPR_MOD);
			return EJS_TOK_EXPR;

		case '/':
			/*
			 *	Handle the division operator and comments
			 */
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			if (c != '*' && c != '/') {
				inputPutback(ep, c);
				tokenAddChar(ep, EJS_EXPR_DIV);
				return EJS_TOK_EXPR;
			}
			style = c;
			/*
			 *	Eat comments. Both C and C++ comment styles are supported.
			 */
			while (1) {
				if ((c = inputGetc(ep)) < 0) {
					if (style == '/') {
						return EJS_TOK_EOF;
					}
					ejsError(ep, "Syntax Error");
					return EJS_TOK_ERR;
				}
				if (c == '\n' && style == '/') {
					break;
				} else if (c == '*') {
					c = inputGetc(ep);
					if (style == '/') {
						if (c == '\n') {
							break;
						}
					} else {
						if (c == '/') {
							break;
						}