ejslex.c

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

					}
				}
			}
			/*
			 *	Continue looking for a token, so get the next character
			 */
			if ((c = inputGetc(ep)) < 0) {
				return EJS_TOK_EOF;
			}
			break;

		case '<':									/* < and <= */
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			if (c == '<') {
				tokenAddChar(ep, EJS_EXPR_LSHIFT);
				return EJS_TOK_EXPR;
			} else if (c == '=') {
				tokenAddChar(ep, EJS_EXPR_LESSEQ);
				return EJS_TOK_EXPR;
			}
			tokenAddChar(ep, EJS_EXPR_LESS);
			inputPutback(ep, c);
			return EJS_TOK_EXPR;

		case '>':									/* > and >= */
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			if (c == '>') {
				tokenAddChar(ep, EJS_EXPR_RSHIFT);
				return EJS_TOK_EXPR;
			} else if (c == '=') {
				tokenAddChar(ep, EJS_EXPR_GREATEREQ);
				return EJS_TOK_EXPR;
			}
			tokenAddChar(ep, EJS_EXPR_GREATER);
			inputPutback(ep, c);
			return EJS_TOK_EXPR;

		case '=':									/* "==" */
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			if (c == '=') {
				tokenAddChar(ep, EJS_EXPR_EQ);
				return EJS_TOK_EXPR;
			}
			inputPutback(ep, c);
			return EJS_TOK_ASSIGNMENT;

		case '!':									/* "!=" or "!"*/
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			if (c == '=') {
				tokenAddChar(ep, EJS_EXPR_NOTEQ);
				return EJS_TOK_EXPR;
			}
			inputPutback(ep, c);
			tokenAddChar(ep, EJS_EXPR_BOOL_COMP);
			return EJS_TOK_EXPR;

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

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

		case '|':									/* "||" */
			if ((c = inputGetc(ep)) < 0 || c != '|') {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			tokenAddChar(ep, EJS_COND_OR);
			return EJS_TOK_LOGICAL;

		case '&':									/* "&&" */
			if ((c = inputGetc(ep)) < 0 || c != '&') {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}
			tokenAddChar(ep, EJS_COND_AND);
			return EJS_TOK_LOGICAL;

		case '\"':									/* String quote */
		case '\'':
			quote = c;
			if ((c = inputGetc(ep)) < 0) {
				ejsError(ep, "Syntax Error");
				return EJS_TOK_ERR;
			}

			while (c != quote) {
				/*
				 *	Check for escape sequence characters
				 */
				if (c == '\\') {
					c = inputGetc(ep);

					if (isdigit(c)) {
						/*
						 *	Octal support, \101 maps to 65 = 'A'. Put first 
						 *	char back so converter will work properly.
						 */
						inputPutback(ep, c);
						c = charConvert(ep, 8, 3);

					} else {
						switch (c) {
						case 'n':
							c = '\n'; break;
						case 'b':
							c = '\b'; break;
						case 'f':
							c = '\f'; break;
						case 'r':
							c = '\r'; break;
						case 't':
							c = '\t'; break;
						case 'x':
							/*
							 *	Hex support, \x41 maps to 65 = 'A'
							 */
							c = charConvert(ep, 16, 2);
							break;
						case 'u':
							/*
							 *	Unicode support, \x0401 maps to 65 = 'A'
							 */
							c = charConvert(ep, 16, 2);
							c = c*16 + charConvert(ep, 16, 2);

							break;
						case '\'':
						case '\"':
						case '\\':
							break;
						default:
							ejsError(ep, "Invalid Escape Sequence");
							return EJS_TOK_ERR;
						}
					}
					if (tokenAddChar(ep, c) < 0) {
						return EJS_TOK_ERR;
					}
				} else {
					if (tokenAddChar(ep, c) < 0) {
						return EJS_TOK_ERR;
					}
				}
				if ((c = inputGetc(ep)) < 0) {
					ejsError(ep, "Unmatched Quote");
					return EJS_TOK_ERR;
				}
			}
			return EJS_TOK_LITERAL;

		case '0': 
			if (tokenAddChar(ep, c) < 0) {
				return EJS_TOK_ERR;
			}
			if ((c = inputGetc(ep)) < 0) {
				break;
			}
			if (tolower(c) == 'x') {
				do {
					if (tokenAddChar(ep, c) < 0) {
						return EJS_TOK_ERR;
					}
					if ((c = inputGetc(ep)) < 0) {
						break;
					}
				} while (isdigit(c) || (tolower(c) >= 'a' && tolower(c) <= 'f'));

				mprDestroyVar(&ep->tokenNumber);
				ep->tokenNumber = mprParseVar(ep->token, type);
				inputPutback(ep, c);
				return EJS_TOK_NUMBER;
			}
			if (! isdigit(c)) {
#if BLD_FEATURE_FLOATING_POINT
				if (c == '.' || tolower(c) == 'e' || c == '+' || c == '-') {
					/* Fall through */
					type = MPR_TYPE_FLOAT;
				} else
#endif
				{
					mprDestroyVar(&ep->tokenNumber);
					ep->tokenNumber = mprParseVar(ep->token, type);
					inputPutback(ep, c);
					return EJS_TOK_NUMBER;
				}
			}
			/* Fall through to get more digits */

		case '1': case '2': case '3': case '4': 
		case '5': case '6': case '7': case '8': case '9':
			do {
				if (tokenAddChar(ep, c) < 0) {
					return EJS_TOK_ERR;
				}
				if ((c = inputGetc(ep)) < 0) {
					break;
				}
#if BLD_FEATURE_FLOATING_POINT
				if (c == '.' || tolower(c) == 'e' || tolower(c) == 'f') {
					type = MPR_TYPE_FLOAT;
				}
			} while (isdigit(c) || c == '.' || tolower(c) == 'e' || tolower(c) == 'f' ||
				((type == MPR_TYPE_FLOAT) && (c == '+' || c == '-')));
#else
			} while (isdigit(c));
#endif

			mprDestroyVar(&ep->tokenNumber);
			ep->tokenNumber = mprParseVar(ep->token, type);
			inputPutback(ep, c);
			return EJS_TOK_NUMBER;

		default:
			/*
			 *	Identifiers or a function names
			 */
			while (1) {
				if (c == '\\') {
					if ((c = inputGetc(ep)) < 0) {
						break;
					}
					if (c == '\n' || c == '\r') {
						break;
					}
				} else if (tokenAddChar(ep, c) < 0) {
						break;
				}
				if ((c = inputGetc(ep)) < 0) {
					break;
				}
				if (!isalnum(c) && c != '$' && c != '_' && c != '\\') {
					break;
				}
			}
			if (*ep->token == '\0') {
				c = inputGetc(ep);
				break;
			}
			if (! isalpha((int) *ep->token) && *ep->token != '$' && 
					*ep->token != '_') {
				ejsError(ep, "Invalid identifier %s", ep->token);
				return EJS_TOK_ERR;
			}

			tid = checkReservedWord(ep, state, c, EJS_TOK_ID);
			if (tid != EJS_TOK_ID) {
				return tid;
			}

			/* 
			 *	Skip white space after token to find out whether this is
			 * 	a function or not.
			 */ 
			while (c == ' ' || c == '\t' || c == '\r' || c == '\n') {
				if ((c = inputGetc(ep)) < 0)
					break;
			}

			tid = EJS_TOK_ID;
			done++;
		}
	}

	/*
	 *	Putback the last extra character for next time
	 */
	inputPutback(ep, c);
	return tid;
}

/******************************************************************************/
/*
 *	Convert a hex or octal character back to binary, return original char if 
 *	not a hex digit
 */

static int charConvert(Ejs *ep, int base, int maxDig)
{
	int		i, c, lval, convChar;

	lval = 0;
	for (i = 0; i < maxDig; i++) {
		if ((c = inputGetc(ep)) < 0) {
			break;
		}
		/*
		 *	Initialize to out of range value
		 */
		convChar = base;
		if (isdigit(c)) {
			convChar = c - '0';
		} else if (c >= 'a' && c <= 'f') {
			convChar = c - 'a' + 10;
		} else if (c >= 'A' && c <= 'F') {
			convChar = c - 'A' + 10;
		}
		/*
		 *	If unexpected character then return it to buffer.
		 */
		if (convChar >= base) {
			inputPutback(ep, c);
			break;
		}
		lval = (lval * base) + convChar;
	}
	return lval;
}

/******************************************************************************/
/*
 *	Putback the last token read. Accept at most one push back token.
 */

void ejsLexPutbackToken(Ejs *ep, int tid, char *string)
{
	EjsInput	*ip;
	int			idx;

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

	ip->putBackIndex += 1;
	idx = ip->putBackIndex;
	ip->putBack[idx].id = tid;

	if (ip->putBack[idx].token) {
		if (ip->putBack[idx].token == string) {
			return;
		}
		mprFree(ip->putBack[idx].token);
	}
	ip->putBack[idx].token = mprStrdup(string);
}

/******************************************************************************/
/*
 *	Add a character to the token buffer
 */

static int tokenAddChar(Ejs *ep, int c)
{
	EjsInput	*ip;
	uchar		*oldbuf;

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

	if (ip->tokEndp >= &ip->tokbuf[ip->tokSize - 1]) {
		ip->tokSize += EJS_PARSE_INCR;
		oldbuf = ip->tokbuf;
		ip->tokbuf = mprRealloc(ip->tokbuf, ip->tokSize);
		if (ip->tokbuf == 0) {
			ejsError(ep, "Token too big");
			return -1;
		}
		ip->tokEndp += (int) ((uchar*) ip->tokbuf - oldbuf);
		ip->tokServp += (int) ((uchar*) ip->tokbuf - oldbuf);
		ep->token += (int) ((uchar*) ip->tokbuf - oldbuf);
	}
	*ip->tokEndp++ = c;
	*ip->tokEndp = '\0';

	return 0;
}

/******************************************************************************/
/*
 *	Get another input character
 */

static int inputGetc(Ejs *ep)
{
	EjsInput	*ip;
	int			c;

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

	if (ip->scriptSize <= 0) {
		return -1;
	}

	c = (uchar) (*ip->scriptServp++);
	ip->scriptSize--;

	/*
	 *	For debugging, accumulate the line number and the currenly parsed line
	 */
	if (c == '\n') {
#if BLD_DEBUG && 0
		if (ip->lineColumn > 0) {
			printf("PARSED: %s\n", ip->line);
		}
#endif
		ip->lineNumber++;
		ip->lineColumn = 0;
	} else {
		if ((ip->lineColumn + 2) >= ip->lineLength) {
			ip->lineLength += 80;
			ip->line = mprRealloc(ip->line, ip->lineLength * sizeof(char));
		}
		ip->line[ip->lineColumn++] = c;
		ip->line[ip->lineColumn] = '\0';
	}
	return c;
}

/******************************************************************************/
/*
 *	Putback a character onto the input queue
 */

static void inputPutback(Ejs *ep, int c)
{
	EjsInput	*ip;

	mprAssert(ep);

	if (c != 0) {
		ip = ep->input;
		*--ip->scriptServp = c;
		ip->scriptSize++;
		ip->lineColumn--;
		ip->line[ip->lineColumn] = '\0';
	}
}

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

#else
void ejsLexDummy() {}

/******************************************************************************/
#endif /* BLD_FEATURE_EJS */

/*
 * Local variables:
 * tab-width: 4
 * c-basic-offset: 4
 * End:
 * vim:tw=78
 * vim600: sw=4 ts=4 fdm=marker
 * vim<600: sw=4 ts=4
 */