1 /* rclex.c -- lexer for Windows rc files parser */
3 /* Copyright 1997, 1998, 1999, 2001, 2002, 2003, 2005, 2006, 2007
4 Free Software Foundation, Inc.
6 Written by Kai Tietz, Onevision.
8 This file is part of GNU Binutils.
10 This program is free software; you can redistribute it and/or modify
11 it under the terms of the GNU General Public License as published by
12 the Free Software Foundation; either version 2 of the License, or
13 (at your option) any later version.
15 This program is distributed in the hope that it will be useful,
16 but WITHOUT ANY WARRANTY; without even the implied warranty of
17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 GNU General Public License for more details.
20 You should have received a copy of the GNU General Public License
21 along with this program; if not, write to the Free Software
22 Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, MA
25 /* This is a lexer used by the Windows rc file parser. It basically
26 just recognized a bunch of keywords. */
31 #include "libiberty.h"
32 #include "safe-ctype.h"
38 /* Whether we are in rcdata mode, in which we returns the lengths of
41 static int rcdata_mode;
43 /* Whether we are supressing lines from cpp (including windows.h or
44 headers from your C sources may bring in externs and typedefs).
45 When active, we return IGNORED_TOKEN, which lets us ignore these
46 outside of resource constructs. Thus, it isn't required to protect
47 all the non-preprocessor lines in your header files with #ifdef
48 RC_INVOKED. It also means your RC file can't include other RC
49 files if they're named "*.h". Sorry. Name them *.rch or whatever. */
51 static int suppress_cpp_data;
53 #define IGNORE_CPP(x) (suppress_cpp_data ? IGNORED_TOKEN : (x))
55 /* The first filename we detect in the cpp output. We use this to
56 tell included files from the original file. */
58 static char *initial_fn;
60 /* List of allocated strings. */
64 struct alloc_string *next;
68 static struct alloc_string *strings;
76 #define K(KEY) { #KEY, KEY }
77 #define KRT(KEY) { #KEY, RT_##KEY }
79 static const struct rclex_keywords keywds[] =
81 K(ACCELERATORS), K(ALT), K(ANICURSOR), K(ANIICON), K(ASCII),
82 K(AUTO3STATE), K(AUTOCHECKBOX), K(AUTORADIOBUTTON),
83 K(BEDIT), { "BEGIN", BEG }, K(BITMAP), K(BLOCK), K(BUTTON),
84 K(CAPTION), K(CHARACTERISTICS), K(CHECKBOX), K(CHECKED),
85 K(CLASS), K(COMBOBOX), K(CONTROL), K(CTEXT), K(CURSOR),
86 K(DEFPUSHBUTTON), K(DIALOG), K(DIALOGEX), K(DISCARDABLE),
87 K(DLGINCLUDE), K(DLGINIT),
88 K(EDITTEXT), K(END), K(EXSTYLE),
89 K(FILEFLAGS), K(FILEFLAGSMASK), K(FILEOS), K(FILESUBTYPE),
90 K(FILETYPE), K(FILEVERSION), K(FIXED), K(FONT), K(FONTDIR),
91 K(GRAYED), KRT(GROUP_CURSOR), KRT(GROUP_ICON), K(GROUPBOX),
92 K(HEDIT), K(HELP), K(HTML),
93 K(ICON), K(IEDIT), K(IMPURE), K(INACTIVE),
94 K(LANGUAGE), K(LISTBOX), K(LOADONCALL), K(LTEXT),
95 K(MANIFEST), K(MENU), K(MENUBARBREAK), K(MENUBREAK),
96 K(MENUEX), K(MENUITEM), K(MESSAGETABLE), K(MOVEABLE),
98 K(PLUGPLAY), K(POPUP), K(PRELOAD), K(PRODUCTVERSION),
99 K(PURE), K(PUSHBOX), K(PUSHBUTTON),
100 K(RADIOBUTTON), K(RCDATA), K(RTEXT),
101 K(SCROLLBAR), K(SEPARATOR), K(SHIFT), K(STATE3),
102 K(STRINGTABLE), K(STYLE),
105 K(VALUE), { "VERSION", VERSIONK }, K(VERSIONINFO),
110 /* External input stream from resrc */
111 extern FILE *cpp_pipe;
113 /* Lexical scanner helpers. */
114 static int rclex_lastch = -1;
115 static size_t rclex_tok_max = 0;
116 static size_t rclex_tok_pos = 0;
117 static char *rclex_tok = NULL;
120 rclex_translatekeyword (const char *key)
122 if (key && ISUPPER (key[0]))
124 const struct rclex_keywords *kw = &keywds[0];
128 if (! strcmp (kw->name, key))
132 while (kw->name != NULL);
137 /* Handle a C preprocessor line. */
142 const char *s = rclex_tok;
150 line = strtol (s, &send, 0);
151 if (*send != '\0' && ! ISSPACE (*send))
154 /* Subtract 1 because we are about to count the newline. */
155 rc_lineno = line - 1;
165 send = strchr (s, '"');
169 fn = xmalloc (send - s + 1);
170 strncpy (fn, s, send - s);
178 initial_fn = xmalloc (strlen (fn) + 1);
179 strcpy (initial_fn, fn);
182 /* Allow the initial file, regardless of name. Suppress all other
183 files if they end in ".h" (this allows included "*.rc"). */
184 if (strcmp (initial_fn, fn) == 0
185 || strcmp (fn + strlen (fn) - 2, ".h") != 0)
186 suppress_cpp_data = 0;
188 suppress_cpp_data = 1;
191 /* Allocate a string of a given length. */
196 struct alloc_string *as;
198 as = xmalloc (sizeof *as);
199 as->s = xmalloc (len);
207 /* Handle a quoted string. The quotes are stripped. A pair of quotes
208 in a string are turned into a single quote. Adjacent strings are
209 merged separated by whitespace are merged, as in C. */
212 handle_quotes (rc_uint_type *len)
214 const char *input = rclex_tok;
220 ret = get_string (strlen (input) + 1);
234 rcparse_warning ("backslash at end of string");
238 rcparse_warning ("use \"\" to put \" in a string");
244 *s++ = ESCAPE_B; /* Strange, but true... */
282 case '0': case '1': case '2': case '3':
283 case '4': case '5': case '6': case '7':
286 if (*t >= '0' && *t <= '7')
288 ch = (ch << 3) | (*t - '0');
290 if (*t >= '0' && *t <= '7')
292 ch = (ch << 3) | (*t - '0');
302 /* We only handle single byte chars here. Make sure
303 we finish an escape sequence like "/xB0ABC" after
304 the first two digits. */
306 while (num_xdigits--)
308 if (*t >= '0' && *t <= '9')
309 ch = (ch << 4) | (*t - '0');
310 else if (*t >= 'a' && *t <= 'f')
311 ch = (ch << 4) | (*t - 'a' + 10);
312 else if (*t >= 'A' && *t <= 'F')
313 ch = (ch << 4) | (*t - 'A' + 10);
322 rcparse_warning ("unrecognized escape sequence");
330 else if (t[1] == '\0')
332 else if (t[1] == '"')
339 rcparse_warning ("unexpected character after '\"'");
341 assert (ISSPACE (*t));
362 /* Allocate a unicode string of a given length. */
365 get_unistring (int len)
367 return (unichar *) get_string (len * sizeof (unichar));
370 /* Handle a quoted unicode string. The quotes are stripped. A pair of quotes
371 in a string are turned into a single quote. Adjacent strings are
372 merged separated by whitespace are merged, as in C. */
375 handle_uniquotes (rc_uint_type *len)
377 const char *input = rclex_tok;
383 ret = get_unistring (strlen (input) + 1);
387 if ((*t == 'L' || *t == 'l') && t[1] == '"')
399 rcparse_warning ("backslash at end of string");
403 rcparse_warning ("use \"\" to put \" in a string");
407 *s++ = ESCAPE_B; /* Strange, but true... */
442 *s++ = (unichar) *t++;
445 case '0': case '1': case '2': case '3':
446 case '4': case '5': case '6': case '7':
449 if (*t >= '0' && *t <= '7')
451 ch = (ch << 3) | (*t - '0');
453 if (*t >= '0' && *t <= '7')
455 ch = (ch << 3) | (*t - '0');
465 /* We only handle two byte chars here. Make sure
466 we finish an escape sequence like "/xB0ABC" after
467 the first two digits. */
469 while (num_xdigits--)
471 if (*t >= '0' && *t <= '9')
472 ch = (ch << 4) | (*t - '0');
473 else if (*t >= 'a' && *t <= 'f')
474 ch = (ch << 4) | (*t - 'a' + 10);
475 else if (*t >= 'A' && *t <= 'F')
476 ch = (ch << 4) | (*t - 'A' + 10);
485 rcparse_warning ("unrecognized escape sequence");
487 *s++ = (unichar) *t++;
492 *s++ = (unichar) *t++;
493 else if (t[1] == '\0')
495 else if (t[1] == '"')
503 assert (ISSPACE (*t));
524 /* Discard all the strings we have allocated. The parser calls this
525 when it no longer needs them. */
528 rcparse_discard_strings (void)
530 struct alloc_string *as;
535 struct alloc_string *n;
546 /* Enter rcdata mode. */
548 rcparse_rcdata (void)
553 /* Go back to normal mode from rcdata mode. */
555 rcparse_normal (void)
561 rclex_tok_add_char (int ch)
563 if (! rclex_tok || rclex_tok_max <= rclex_tok_pos)
565 char *h = xmalloc (rclex_tok_max + 9);
571 memcpy (h, rclex_tok, rclex_tok_pos + 1);
580 rclex_tok[rclex_tok_pos++] = (char) ch;
581 rclex_tok[rclex_tok_pos] = 0;
589 if ((r = rclex_lastch) != -1)
596 if (! cpp_pipe || feof (cpp_pipe)
597 || fread (&ch, 1, 1,cpp_pipe) != 1)
599 r = ((int) ch) & 0xff;
601 while (r == 0 || r == '\r');
603 rclex_tok_add_char (r);
612 if ((r = rclex_lastch) == -1)
614 if ((r = rclex_readch ()) != -1)
617 if (rclex_tok_pos > 0)
618 rclex_tok[--rclex_tok_pos] = 0;
629 while ((c = rclex_peekch ()) != -1)
636 if ((c = rclex_peekch ()) == -1 || c == '\n')
640 else if (rclex_readch () == '"')
642 if (rclex_peekch () == '"')
653 rc_uint_type base = 10;
654 rc_uint_type ret, val;
661 switch (rclex_peekch ())
675 ret = (rc_uint_type) (ch - '0');
676 while ((ch = rclex_peekch ()) != -1)
679 val = (rc_uint_type) (ch - '0');
680 else if (ch >= 'a' && ch <= 'f')
681 val = (rc_uint_type) ((ch - 'a') + 10);
682 else if (ch >= 'A' && ch <= 'F')
683 val = (rc_uint_type) ((ch - 'A') + 10);
687 if (! warned && val >= base)
690 rcparse_warning ("digit exceeds base");
698 /* yyparser entry method. */
708 /* Make sure that rclex_tok is initialized. */
710 rclex_tok_add_char (-1);
720 if ((ch = rclex_readch ()) == -1)
730 while ((ch = rclex_peekch ()) != -1 && ch != '\n')
737 ch = IGNORE_CPP (BEG);
741 ch = IGNORE_CPP (END);
744 case '0': case '1': case '2': case '3': case '4':
745 case '5': case '6': case '7': case '8': case '9':
746 yylval.i.val = read_digit (ch);
748 switch (rclex_peekch ())
755 ch = IGNORE_CPP (NUMBER);
759 ch = IGNORE_CPP ((! rcdata_mode ? QUOTEDSTRING : SIZEDSTRING));
760 if (ch == IGNORED_TOKEN)
762 s = handle_quotes (&length);
767 yylval.ss.length = length;
772 if (rclex_peekch () == '"')
776 ch = IGNORE_CPP ((! rcdata_mode ? QUOTEDUNISTRING : SIZEDUNISTRING));
777 if (ch == IGNORED_TOKEN)
779 us = handle_uniquotes (&length);
784 yylval.suni.length = length;
791 if (ISIDST (ch) || ch=='$')
793 while ((ch = rclex_peekch ()) != -1 && (ISIDNUM (ch) || ch == '$' || ch == '.'))
795 ch = IGNORE_CPP (rclex_translatekeyword (rclex_tok));
798 s = get_string (strlen (rclex_tok) + 1);
799 strcpy (s, rclex_tok);
802 else if (ch == BLOCK)
804 const char *hs = NULL;
813 hs = yylval.s = yylval.ss.s;
818 rcparse_warning ("BLOCK expects a string as argument.");
821 else if (! strcmp (hs, "StringFileInfo"))
822 ch = BLOCKSTRINGFILEINFO;
823 else if (! strcmp (hs, "VarFileInfo"))
824 ch = BLOCKVARFILEINFO;
828 ch = IGNORE_CPP (ch);
832 while (ch == IGNORED_TOKEN);