2 * Copyright 2008 Jacek Caban for CodeWeavers
4 * This library is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU Lesser General Public
6 * License as published by the Free Software Foundation; either
7 * version 2.1 of the License, or (at your option) any later version.
9 * This library is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12 * Lesser General Public License for more details.
14 * You should have received a copy of the GNU Lesser General Public
15 * License along with this library; if not, write to the Free Software
16 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
20 #include "wine/port.h"
29 #include "parser.tab.h"
31 #include "wine/debug.h"
32 #include "wine/unicode.h"
34 WINE_DEFAULT_DEBUG_CHANNEL(jscript);
36 #define LONGLONG_MAX (((LONGLONG)0x7fffffff<<32)|0xffffffff)
38 static const WCHAR breakW[] = {'b','r','e','a','k',0};
39 static const WCHAR caseW[] = {'c','a','s','e',0};
40 static const WCHAR catchW[] = {'c','a','t','c','h',0};
41 static const WCHAR continueW[] = {'c','o','n','t','i','n','u','e',0};
42 static const WCHAR defaultW[] = {'d','e','f','a','u','l','t',0};
43 static const WCHAR deleteW[] = {'d','e','l','e','t','e',0};
44 static const WCHAR doW[] = {'d','o',0};
45 static const WCHAR elseW[] = {'e','l','s','e',0};
46 static const WCHAR falseW[] = {'f','a','l','s','e',0};
47 static const WCHAR finallyW[] = {'f','i','n','a','l','l','y',0};
48 static const WCHAR forW[] = {'f','o','r',0};
49 static const WCHAR functionW[] = {'f','u','n','c','t','i','o','n',0};
50 static const WCHAR ifW[] = {'i','f',0};
51 static const WCHAR inW[] = {'i','n',0};
52 static const WCHAR instanceofW[] = {'i','n','s','t','a','n','c','e','o','f',0};
53 static const WCHAR newW[] = {'n','e','w',0};
54 static const WCHAR nullW[] = {'n','u','l','l',0};
55 static const WCHAR returnW[] = {'r','e','t','u','r','n',0};
56 static const WCHAR switchW[] = {'s','w','i','t','c','h',0};
57 static const WCHAR thisW[] = {'t','h','i','s',0};
58 static const WCHAR throwW[] = {'t','h','r','o','w',0};
59 static const WCHAR trueW[] = {'t','r','u','e',0};
60 static const WCHAR tryW[] = {'t','r','y',0};
61 static const WCHAR typeofW[] = {'t','y','p','e','o','f',0};
62 static const WCHAR undefinedW[] = {'u','n','d','e','f','i','n','e','d',0};
63 static const WCHAR varW[] = {'v','a','r',0};
64 static const WCHAR voidW[] = {'v','o','i','d',0};
65 static const WCHAR whileW[] = {'w','h','i','l','e',0};
66 static const WCHAR withW[] = {'w','i','t','h',0};
73 {breakW, kBREAK, TRUE},
76 {continueW, kCONTINUE, TRUE},
84 {functionW, kFUNCTION},
87 {instanceofW, kINSTANCEOF},
90 {returnW, kRETURN, TRUE},
103 static int lex_error(parser_ctx_t *ctx, HRESULT hres)
106 ctx->lexer_error = TRUE;
110 /* ECMA-262 3rd Edition 7.6 */
111 static BOOL is_identifier_char(WCHAR c)
113 return isalnumW(c) || c == '$' || c == '_' || c == '\\';
116 static int check_keyword(parser_ctx_t *ctx, const WCHAR *word, const WCHAR **lval)
118 const WCHAR *p1 = ctx->ptr;
119 const WCHAR *p2 = word;
121 while(p1 < ctx->end && *p2) {
128 if(*p2 || (p1 < ctx->end && is_identifier_char(*p1)))
137 /* ECMA-262 3rd Edition 7.3 */
138 static BOOL is_endline(WCHAR c)
140 return c == '\n' || c == '\r' || c == 0x2028 || c == 0x2029;
143 static int hex_to_int(WCHAR c)
145 if('0' <= c && c <= '9')
148 if('a' <= c && c <= 'f')
151 if('A' <= c && c <= 'F')
157 static int check_keywords(parser_ctx_t *ctx, const WCHAR **lval)
159 int min = 0, max = sizeof(keywords)/sizeof(keywords[0])-1, r, i;
164 r = check_keyword(ctx, keywords[i].word, lval);
166 ctx->implicit_nl_semicolon = keywords[i].no_nl;
167 return keywords[i].token;
179 static BOOL skip_html_comment(parser_ctx_t *ctx)
181 const WCHAR html_commentW[] = {'<','!','-','-',0};
183 if(!ctx->is_html || ctx->ptr+3 >= ctx->end ||
184 memcmp(ctx->ptr, html_commentW, sizeof(WCHAR)*4))
188 while(ctx->ptr < ctx->end && !is_endline(*ctx->ptr++));
193 static BOOL skip_comment(parser_ctx_t *ctx)
195 if(ctx->ptr+1 >= ctx->end)
198 if(*ctx->ptr != '/') {
199 if(*ctx->ptr == '@' && ctx->ptr+2 < ctx->end && ctx->ptr[1] == '*' && ctx->ptr[2] == '/') {
207 switch(ctx->ptr[1]) {
210 if(ctx->ptr+2 < ctx->end && *ctx->ptr == '@' && is_identifier_char(ctx->ptr[1]))
212 while(ctx->ptr+1 < ctx->end && (ctx->ptr[0] != '*' || ctx->ptr[1] != '/'))
215 if(ctx->ptr[0] == '*' && ctx->ptr[1] == '/') {
218 WARN("unexpected end of file (missing end of comment)\n");
224 if(ctx->ptr+2 < ctx->end && *ctx->ptr == '@' && is_identifier_char(ctx->ptr[1]))
226 while(ctx->ptr < ctx->end && !is_endline(*ctx->ptr))
236 static BOOL unescape(WCHAR *str)
272 i = hex_to_int(*++p);
277 i = hex_to_int(*++p);
283 i = hex_to_int(*++p);
288 i = hex_to_int(*++p);
293 i = hex_to_int(*++p);
298 i = hex_to_int(*++p);
307 c = c*8 + (*p++ - '0');
309 c = c*8 + (*p++ - '0');
325 static int parse_identifier(parser_ctx_t *ctx, const WCHAR **ret)
327 const WCHAR *ptr = ctx->ptr++;
331 while(ctx->ptr < ctx->end && is_identifier_char(*ctx->ptr))
336 *ret = wstr = parser_alloc(ctx, (len+1)*sizeof(WCHAR));
337 memcpy(wstr, ptr, len*sizeof(WCHAR));
340 /* FIXME: unescape */
344 static int parse_string_literal(parser_ctx_t *ctx, const WCHAR **ret, WCHAR endch)
346 const WCHAR *ptr = ++ctx->ptr;
350 while(ctx->ptr < ctx->end && *ctx->ptr != endch) {
351 if(*ctx->ptr++ == '\\')
355 if(ctx->ptr == ctx->end)
356 return lex_error(ctx, JS_E_UNTERMINATED_STRING);
360 *ret = wstr = parser_alloc(ctx, (len+1)*sizeof(WCHAR));
361 memcpy(wstr, ptr, len*sizeof(WCHAR));
366 if(!unescape(wstr)) {
367 WARN("unescape failed\n");
368 return lex_error(ctx, E_FAIL);
371 return tStringLiteral;
374 static literal_t *new_double_literal(parser_ctx_t *ctx, DOUBLE d)
376 literal_t *ret = parser_alloc(ctx, sizeof(literal_t));
378 ret->type = LT_DOUBLE;
383 literal_t *new_boolean_literal(parser_ctx_t *ctx, BOOL bval)
385 literal_t *ret = parser_alloc(ctx, sizeof(literal_t));
393 static int parse_double_literal(parser_ctx_t *ctx, LONG int_part, literal_t **literal)
399 while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr)) {
400 hlp = d*10 + *(ctx->ptr++) - '0';
401 if(d>LONGLONG_MAX/10 || hlp<0) {
408 while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr)) {
413 if(*ctx->ptr == '.') {
416 while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr)) {
417 hlp = d*10 + *(ctx->ptr++) - '0';
418 if(d>LONGLONG_MAX/10 || hlp<0)
424 while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr))
428 if(ctx->ptr < ctx->end && (*ctx->ptr == 'e' || *ctx->ptr == 'E')) {
432 if(ctx->ptr < ctx->end) {
433 if(*ctx->ptr == '+') {
435 }else if(*ctx->ptr == '-') {
438 }else if(!isdigitW(*ctx->ptr)) {
439 WARN("Expected exponent part\n");
440 return lex_error(ctx, E_FAIL);
444 if(ctx->ptr == ctx->end) {
445 WARN("unexpected end of file\n");
446 return lex_error(ctx, E_FAIL);
449 while(ctx->ptr < ctx->end && isdigitW(*ctx->ptr)) {
450 if(e > INT_MAX/10 || (e = e*10 + *ctx->ptr++ - '0')<0)
455 if(exp<0 && e<0 && e+exp>0) exp = INT_MIN;
456 else if(exp>0 && e>0 && e+exp<0) exp = INT_MAX;
460 *literal = new_double_literal(ctx, exp>=0 ? d*pow(10, exp) : d/pow(10, -exp));
461 return tNumericLiteral;
464 static int parse_numeric_literal(parser_ctx_t *ctx, literal_t **literal)
468 l = *ctx->ptr++ - '0';
470 if(*ctx->ptr == 'x' || *ctx->ptr == 'X') {
471 if(++ctx->ptr == ctx->end) {
472 ERR("unexpected end of file\n");
476 while(ctx->ptr < ctx->end && (d = hex_to_int(*ctx->ptr)) != -1) {
481 if(ctx->ptr < ctx->end && is_identifier_char(*ctx->ptr)) {
482 WARN("unexpected identifier char\n");
483 return lex_error(ctx, E_FAIL);
486 *literal = new_double_literal(ctx, l);
487 return tNumericLiteral;
490 if(is_identifier_char(*ctx->ptr)) {
491 WARN("wrong char after zero\n");
492 return lex_error(ctx, E_FAIL);
495 if(isdigitW(*ctx->ptr)) {
496 FIXME("octal literals not implemented\n");
497 return lex_error(ctx, E_NOTIMPL);
501 return parse_double_literal(ctx, l, literal);
504 static int next_token(parser_ctx_t *ctx, void *lval)
507 while(ctx->ptr < ctx->end && isspaceW(*ctx->ptr)) {
508 if(is_endline(*ctx->ptr++))
511 if(ctx->ptr == ctx->end)
513 }while(skip_comment(ctx) || skip_html_comment(ctx));
515 if(ctx->implicit_nl_semicolon) {
518 ctx->implicit_nl_semicolon = FALSE;
521 if(isalphaW(*ctx->ptr)) {
522 int ret = check_keywords(ctx, lval);
526 return parse_identifier(ctx, lval);
529 if(isdigitW(*ctx->ptr))
530 return parse_numeric_literal(ctx, lval);
546 *(const WCHAR**)lval = ctx->ptr++;
550 if(++ctx->ptr < ctx->end && isdigitW(*ctx->ptr))
551 return parse_double_literal(ctx, 0, lval);
555 if(++ctx->ptr == ctx->end) {
556 *(int*)lval = EXPR_LESS;
563 *(int*)lval = EXPR_LESSEQ;
566 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* <<= */
568 *(int*)lval = EXPR_ASSIGNLSHIFT;
571 *(int*)lval = EXPR_LSHIFT;
574 *(int*)lval = EXPR_LESS;
579 if(++ctx->ptr == ctx->end) { /* > */
580 *(int*)lval = EXPR_GREATER;
587 *(int*)lval = EXPR_GREATEREQ;
590 if(++ctx->ptr < ctx->end) {
591 if(*ctx->ptr == '=') { /* >>= */
593 *(int*)lval = EXPR_ASSIGNRSHIFT;
596 if(*ctx->ptr == '>') { /* >>> */
597 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* >>>= */
599 *(int*)lval = EXPR_ASSIGNRRSHIFT;
602 *(int*)lval = EXPR_RRSHIFT;
606 *(int*)lval = EXPR_RSHIFT;
609 *(int*)lval = EXPR_GREATER;
615 if(ctx->ptr < ctx->end) {
622 *(int*)lval = EXPR_ASSIGNADD;
630 if(ctx->ptr < ctx->end) {
632 case '-': /* -- or --> */
634 if(ctx->is_html && ctx->nl && ctx->ptr < ctx->end && *ctx->ptr == '>') {
641 *(int*)lval = EXPR_ASSIGNSUB;
648 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* *= */
650 *(int*)lval = EXPR_ASSIGNMUL;
656 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* %= */
658 *(int*)lval = EXPR_ASSIGNMOD;
664 if(++ctx->ptr < ctx->end) {
668 *(int*)lval = EXPR_ASSIGNAND;
678 if(++ctx->ptr < ctx->end) {
682 *(int*)lval = EXPR_ASSIGNOR;
692 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* ^= */
694 *(int*)lval = EXPR_ASSIGNXOR;
700 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* != */
701 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* !== */
703 *(int*)lval = EXPR_NOTEQEQ;
706 *(int*)lval = EXPR_NOTEQ;
712 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* == */
713 if(++ctx->ptr < ctx->end && *ctx->ptr == '=') { /* === */
715 *(int*)lval = EXPR_EQEQ;
718 *(int*)lval = EXPR_EQ;
724 if(++ctx->ptr < ctx->end) {
725 if(*ctx->ptr == '=') { /* /= */
727 *(int*)lval = EXPR_ASSIGNDIV;
735 return parse_string_literal(ctx, lval, *ctx->ptr);
739 return parse_identifier(ctx, lval);
745 WARN("unexpected char '%c' %d\n", *ctx->ptr, *ctx->ptr);
755 struct _cc_var_t *next;
760 void release_cc(cc_ctx_t *cc)
762 cc_var_t *iter, *next;
764 for(iter = cc->vars; iter; iter = next) {
772 static BOOL add_cc_var(cc_ctx_t *cc, const WCHAR *name, cc_var_t *v)
779 new_v = heap_alloc(sizeof(cc_var_t) + (len+1)*sizeof(WCHAR));
783 memcpy(new_v, v, sizeof(*v));
784 memcpy(new_v->name, name, (len+1)*sizeof(WCHAR));
785 new_v->name_len = len;
786 new_v->next = cc->vars;
791 static cc_var_t *find_cc_var(cc_ctx_t *cc, const WCHAR *name, unsigned name_len)
795 for(iter = cc->vars; iter; iter = iter->next) {
796 if(iter->name_len == name_len && !memcmp(iter->name, name, name_len*sizeof(WCHAR)))
803 static int init_cc(parser_ctx_t *ctx)
808 static const WCHAR _win32W[] = {'_','w','i','n','3','2',0};
809 static const WCHAR _win64W[] = {'_','w','i','n','6','4',0};
810 static const WCHAR _x86W[] = {'_','x','8','6',0};
811 static const WCHAR _amd64W[] = {'_','a','m','d','6','4',0};
812 static const WCHAR _jscriptW[] = {'_','j','s','c','r','i','p','t',0};
813 static const WCHAR _jscript_buildW[] = {'_','j','s','c','r','i','p','t','_','b','u','i','l','d',0};
814 static const WCHAR _jscript_versionW[] = {'_','j','s','c','r','i','p','t','_','v','e','r','s','i','o','n',0};
819 cc = heap_alloc(sizeof(cc_ctx_t));
821 return lex_error(ctx, E_OUTOFMEMORY);
826 if(!add_cc_var(cc, _jscriptW, &v)
827 || !add_cc_var(cc, sizeof(void*) == 8 ? _win64W : _win32W, &v)
828 || !add_cc_var(cc, sizeof(void*) == 8 ? _amd64W : _x86W, &v)) {
830 return lex_error(ctx, E_OUTOFMEMORY);
834 v.u.n = JSCRIPT_BUILD_VERSION;
835 if(!add_cc_var(cc, _jscript_buildW, &v)) {
837 return lex_error(ctx, E_OUTOFMEMORY);
840 v.u.n = JSCRIPT_MAJOR_VERSION + (DOUBLE)JSCRIPT_MINOR_VERSION/10.0;
841 if(!add_cc_var(cc, _jscript_versionW, &v)) {
843 return lex_error(ctx, E_OUTOFMEMORY);
846 ctx->script->cc = cc;
850 static int cc_token(parser_ctx_t *ctx, void *lval)
855 static const WCHAR cc_onW[] = {'c','c','_','o','n',0};
856 static const WCHAR setW[] = {'s','e','t',0};
857 static const WCHAR elifW[] = {'e','l','i','f',0};
858 static const WCHAR endW[] = {'e','n','d',0};
862 if(!check_keyword(ctx, cc_onW, NULL))
865 if(!check_keyword(ctx, setW, NULL)) {
866 FIXME("@set not implemented\n");
867 return lex_error(ctx, E_NOTIMPL);
870 if(!check_keyword(ctx, ifW, NULL)) {
871 FIXME("@if not implemented\n");
872 return lex_error(ctx, E_NOTIMPL);
875 if(!check_keyword(ctx, elifW, NULL)) {
876 FIXME("@elif not implemented\n");
877 return lex_error(ctx, E_NOTIMPL);
880 if(!check_keyword(ctx, elseW, NULL)) {
881 FIXME("@else not implemented\n");
882 return lex_error(ctx, E_NOTIMPL);
885 if(!check_keyword(ctx, endW, NULL)) {
886 FIXME("@end not implemented\n");
887 return lex_error(ctx, E_NOTIMPL);
891 return lex_error(ctx, JS_E_DISABLED_CC);
893 while(ctx->ptr+id_len < ctx->end && is_identifier_char(ctx->ptr[id_len]))
898 TRACE("var %s\n", debugstr_wn(ctx->ptr, id_len));
900 var = find_cc_var(ctx->script->cc, ctx->ptr, id_len);
902 if(!var || var->is_num) {
903 *(literal_t**)lval = new_double_literal(ctx, var ? var->u.n : NAN);
904 return tNumericLiteral;
907 *(literal_t**)lval = new_boolean_literal(ctx, var->u.b);
908 return tBooleanLiteral;
911 int parser_lex(void *lval, parser_ctx_t *ctx)
915 ctx->nl = ctx->ptr == ctx->begin;
918 ret = next_token(ctx, lval);
919 } while(ret == '@' && !(ret = cc_token(ctx, lval)));
924 literal_t *parse_regexp(parser_ctx_t *ctx)
926 const WCHAR *re, *flags_ptr;
927 BOOL in_class = FALSE;
934 while(*--ctx->ptr != '/');
936 /* Simple regexp pre-parser; '/' if used in char class does not terminate regexp literal */
938 while(ctx->ptr < ctx->end) {
939 if(*ctx->ptr == '\\') {
940 if(++ctx->ptr == ctx->end)
943 if(*ctx->ptr == '\n')
957 if(ctx->ptr == ctx->end || *ctx->ptr != '/') {
958 WARN("pre-parsing failed\n");
962 re_len = ctx->ptr-re;
964 flags_ptr = ++ctx->ptr;
965 while(ctx->ptr < ctx->end && isalnumW(*ctx->ptr))
968 hres = parse_regexp_flags(flags_ptr, ctx->ptr-flags_ptr, &flags);
972 ret = parser_alloc(ctx, sizeof(literal_t));
973 ret->type = LT_REGEXP;
974 ret->u.regexp.str = re;
975 ret->u.regexp.str_len = re_len;
976 ret->u.regexp.flags = flags;